<<

The Elamite Cylinder Seal Corpus, c.3500 – 1000 BC

Volume I, Part III

K. J. Roach

Doctor of Philosophy, (Near Eastern) Archaeology 2008 The University of Sydney

Chapter 5 – Summary of Style Distribution across the Elamite Sites The purpose of this chapter is to detail and outline the specific glyptic style distribution at each site included in the Corpus. This survey has two main objectives. The first is the summation and discussion of the Elamite styles from each site, and thereby the revision and reassessment of the ‘glyptic material’ survey presented for each site in the initial site survey section (Chapter 2), by detailing the site glyptic material in the terms of the new Elamite stylistic paradigm here presented. The second intention is to provide some of the background information and data, be it contextual, stylistic and chronological, regarding the function of various glyptic items at each site and across , thereby enabling the following discussion on glyptic function (Chapter 6). The style distribution (how many styles and in what proportions) of each site will be presented, and thereby the basic chronological distribution of the glyptic material, with any necessary discussion where this information strongly contradicts the established chronological periodisation of a site, will be outlined. The glyptic material types (seals/sealings) and the specific materials will be presented, as will any information regarding seal function from provenance (that is, grave or temple context etc.) or type (sealing type especially). For the most part, this information may be presented and detailed in graphs, figures and tables.

5.1 As already mentioned and explained, Susa has contributed by far the most items to the Corpus. Table and Graph 5.1 illustrate the distribution of the two thousand seven hundred and fifty-five Susian items across the Elamite styles. As is evidenced, and as has already been mentioned in regards to the articulation of the styles, the only true style not represented at Susa is the ‘Anshanite Style’ (AS) (the ‘Not Illustrated’ classification is also not represented at Susa, though this is more a publication phenomenon, than a question of stylistic distribution). The absence of the AS style at Susa is indicative, and indeed characteristic, of the style, and aids in its definition as ‘Anshanite’ rather than generally Susian or Elamite. Susa is the only site in the Corpus that was continuously occupied throughout the entire span of this study (see Chapter 2 for details). This archaeological reality concurs with the glyptic style distribution, as all periods are represented in the styles of Susa. A comparison between Graph 5.1 and summary Graph 4.81 indicates that the Susian glyptic style proportions generally accord with the total Corpus proportions (the most notable

ECS Corpus, Volume I, Part III 577 exception being the EME style, the total large number of which may be accounted for by the extraordinary contribution of Haft Tepe to that particular corpus), both as one may expect from the majority contributing site of Susa, and as indicates the central role of Susa in the Elamite Corpus.

Style Distribution STS 375 PEU 156 JNRS 252 OBRS 88 CPE 296 PEO 43 GS 255 EME 46 AGD 283 KRS 13 STF 284 LME 35 SF 109 LPS 44 LSF 20 LGD 34 ARS 145 No Image 21 PEA 75 Miscellaneous 8 UTRS 116 Unclassifiable 57 2755 Table 5.1. Susa style distribution.

380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

S S S O E e PE GS TF SF EA RS us ST C S LSF ARS P PEU BR PE EME K LM LPS LGD JNRS AGD UTR O mag neo o I la N el isc M Graph 5.1. Susa style distribution.

Table and Graph 5.2 indicate the glyptic type (seal/sealing) proportions of the Susian corpus. The supremacy of seals over sealings (seals accounts for over 57% of the Susian corpus) again almost exactly replicates the seal/sealing division for the

ECS Corpus, Volume I, Part III 578 entire Corpus (see Graph 4.84), indicating the general dominance of Susa within the Corpus.

Glyptic Type Seals 1574 Sealings 1181 Table 5.2. Susa glyptic types.

Seals

Sealings

Graph 5.2. Susa glyptic types.

Table 5.3 lists the material types of the Susian Corpus and Graph 5.3 illustrates the proportions of the significant material type contributors (those with five or more items). The fact that represents the largest single group of Susian glyptic materials concurs both with the distribution pattern of the wider Elamite Corpus (Graph 4.83), and with the general age and quality or approach of the majority Susian publications (that is, before the importance, and indeed the methods, of sealing type recognition were known). It is anticipated that with further (physical) study, many of these items may be identified as jar sealings, door lock sealings and so on, and thereby fill the general void of these items in the Susian corpus. The significant number of at Susa (the second largest group) also conforms with the general Corpus pattern, and indicates that seals were regularly used to seal tablets at Susa. The significant number of at Susa should also be noted, and illustrates a glyptic administrative function associated with these items (in at least the earlier periods) at Susa, a point returned to below (Chapter 6). The reduced (but still relatively large body) of seals at Susa in comparison with the wider Corpus is of note (faience forms the largest seal material group in the Corpus, while it is only the fourth largest seal group at Susa, after , , and ). The general dominance of in the Corpus may be attributed to the extraordinarily large faience corpora

ECS Corpus, Volume I, Part III 579 of Choga Zanbil and Surkh Dum-i-Luri (both further detailed below). It is of note that this ‘artificial’ material is in great dominance at two sites that shared a specific, and exceptional, votive function only (for the glyptic corpora), as will be returned to below in the function discussion (Chapter 6). In this regard it is also pertinent to note the total absence of any seals at Susa (in light of their significant numbers at Choga Zanbil, thus indicating that glass may be a specific Choga Zanbil, or votive [or both] material). The general dominance of accords with the overall material distribution of the Corpus. The material is the second largest seal material in the Susian corpus, and as demonstrated by Graph 4.83, the third largest material in the total Corpus. The significant contribution of the total corpus is in fact Susian (Susa contributes nearly 95% of the total corpus, or 259/274 items). Thus more (or perhaps, less) than being a uniquely Elamite material, may be described as essentially Susian (this concurs with the above cited study [Connan & Deschesne 1996] and the discussion of this still elusive material [Chapter 1]). Aside from these standout materials, the general variability in the Susian material corpus (indeed, the greatest variability of any of the sites under discussion here) reflects the great size of the data set, and concurs with what one might expect of such a large group.

Glyptic Materials Limestone 303 Aragonite 18 Terra cotta 97 Marble 60 Heulandite 31 Bone 1 Sandstone 4 Hematite 53 Shell 113 Lapis lazuli 22 Serpentine 34 Copper 3 Basalt 8 Schist 50 Bitumen aggregate 259 Steatite 82 Black stone 1 Faience 121 Rock crystal 4 Black rock 2 Glazed steatite 149 Carnelian 2 Brown stone 1 Unknown cylinder 58 Milky quartz 2 Green stone 1 Ball of clay 1 Flint 1 Grey rock 1 Sealed bulla(e) 213 Amethyst 1 Grey stone 7 Sealing(s) 588 Chalcedony 1 Greyish stone 1 Door lock sealing 3 Agate 2 Green volcanic rock 1 Jar sealing 50 Jasper 22 White stone 4 Wall lock sealing 3 Sealed envelope 2 Alabaster/gypsum 45 Clay cylinder 9 Sealed tablet(s) 320 Table 5.3. Susa glyptic materials.

ECS Corpus, Volume I, Part III 580 600 550 500 450 400 350 300 250 200 150 100 50 0

e t e e t r ) le it m t te ne b ni tine his de ell (e) ng ston asal ndi sto Sh gate ali et(s e B Jasperypsu Sc ylin re ulla m Mar Steat g rago Faience i eula Hematite rey c ar se tabl L A H Serpen G Terra cotta ed b Sealing(s)J ed Lapis lazuli ster/ n agg al a Clay eal me Glazed steatiteSe S itu Alab B Unknown cylinder Graph 5.3. Susa glyptic materials (with five or more examples).

Several different functional interpretations may be seen in the Susian and other glyptic corpora, on the basis both of provenance and type. As already outlined above (Chapter 1), and later returned to in greater detail (Chapter 6), two glyptic functions may be identified through provenance or archaeological context. Provenance in a tomb or grave context provides a funerary interpretation for a seal; provenance in a temple (or other cult installation) allows for a votive function to be attributed to a seal. It should be noted that neither of these two functions, funerary nor votive, excludes the possibility that the seal originally held a more traditional administrative function. The details of this discussion, and the possibility of discerning such dual (or more) functional seals, will be returned to below (Chapter 6). In terms of type, a sealing of any type (jar sealing, door lock sealing, bulla, sealed tablet, pure undifferentiated sealings and so on) provides evidence that the seal used to make the impression had an administrative function (again, the possibility that seals so used also had a more ethereal, symbolic function, and a later funerary or votive depositional function cannot similarly be discounted). Sealings may be further divided according to participation in a writing-centred (or literate) administrative system (that

ECS Corpus, Volume I, Part III 581 is, seal impressed tablets), or a non-writing administrative (other sealings) usage. In these reconstructions, it has been assumed that all are not tablets, but other types of sealings, such as jar, door, basket etc., and so have been associated with the ‘non-writing’ administrative sphere. This is based on the strongly held belief that sealed tablets are easily, and immediately, recognisable as a form, and have been listed as such in even the earliest glyptic publications (before much consideration, if any, was given to other forms of sealings), and even when not so listed, can be readily discerned and defined from images. It is therefore hypothesised that almost all tablets within the Corpus have been noted and so listed already. Of course it is possible that some tablets have been missed, and would be better recognised in a physical study, though any addition to the number of sealed tablets in the Corpus is deemed most probably to be slight if not negligible. Thus, all non-tablet sealings (including the category) are subsumed under the ‘non- writing’ administrative sphere in the function reconstructions presented here. It may be assumed that most of the seals found within a context other than cultic (the votive seals) or funerary also had an administrative function. This is however, only assumption, and cannot in fact be proven, and so such items have not been included in the following function surveys (except in specific circumstances where a seal was found in a discreet, clear and obvious administrative context, such as the ‘archive’ seal of Chogha Gavaneh, see below). Table and Graph 5.4 illustrate the reconstruction of glyptic function for the Susian corpus as is currently available. The first, and most striking, point of this reconstruction is the fact that for most of the Susian corpus (over 55%) no function can be proposed (the category). This is due to the general lack of decent stratigraphic or provenance information for much of the Susian material (especially those unearthed in the earlier Susa excavations), as has already been detailed elsewhere, and limitations in some of the publications (that is, provenance information may be known, but not detailed in the publications). Thus only fourteen Susian items can be ascribed a ‘funerary’ context, though it seems likely (though merely hypothetical) that more seals were found in the significant number of tombs and gravesites excavated over the years at Susa.

ECS Corpus, Volume I, Part III 582 Glyptic Function Non-writing administration 861 Temple/votive 36 Funerary 14 Writing administration 320 Unknown 1524 Table 5.4. Susa glyptic function.

1600

1500

1400

1300

1200

1100

1000

900

800

700

600

500

400

300

200

100

0 Non-writing Writing Temple/Votive Funerary Unknown administration administration Function

Graph 5.4. Susa glyptic function. Entries coloured purple indicate an administrative function; those coloured orange a symbolic (non-administrative) function; green indicates items of ‘unknown’ function.

Table 5.5 further illustrates the style proportions of the fourteen Susa funerary seals. As is demonstrated, the Susian funerary seals belong to a number of stylistic groups and periods, with no significant majority. It should be recognised however that the earliest Susian funerary seals belong to an ‘Akkadian and ’ period style (three ARS items), meaning that no funerary seals are earlier than an ‘Old Elamite’

ECS Corpus, Volume I, Part III 583 period style. The significance of this chronological distribution (if indeed any significance may be inferred from such a small sample) is unknown.

‘Funerary’ Seals ARS 3 KRS 1 UTRS 3 LME 2 OBRS 2 LGD 1 PEO 1 Unclassifiable 1 14 Table 5.5. Style distribution of the Susa ‘funerary’ seals.

A group of thirty-six seals from Susa, all published in Delaporte’s presentation of the Susa seals excavated by de Morgan, were provided with the provenance “1904, temple de Shoushinak” (Delaporte 1920: 55). It is assumed that this refers to the on the Acropole (not the Apadana temple of the same name), excavated by de Morgan and indicated on Figure 2.2. Due to the apparent cultic designation of this context, these items have been given a votive (or at least temple) function here, though given the reasonably scant details regarding this area and the nature of the glyptic deposition, this definition is less concrete than the similar identifications made for the Choga Zanbil and Surkh Dum-i-Luri temple/votive deposits. Table 5.6 illustrates the stylistic distribution of these thirty-six items. Two interesting points may be ascertained from this distribution pattern. The first is the fact that a wide chronological time frame is represented by these items, from the Susa III period JNRS style through to the late Middle Elamite LME style. The other important fact is that many (indeed the vast majority) of these items predate the Middle Elamite period flourishing of the votive seal function. It has already been seen, and will be further detailed below, that so-called ‘heirloom’ seals were deposited in the Surkh Dum sanctuary, and a lesser number, but none the less significant, amount of ‘old’ seals were deposited at Choga Zanbil. As both these sites were founded in the Middle Elamite period however, it may be assumed that these items were antique or ‘heirlooms’, kept by their owners (or successive owners) and then deposited in the Middle Elamite period (or beyond in the case of Surkh Dum) with the vast majority of contemporary or ‘modern’ seals. In both the Surkh Dum and Zanbil instances the old or ‘heirloom’ seals are not evidence of a long, continual depositional use, but rather indicate already ancient seals deposited simultaneously with new. On the basis of this evidence (the Surkh Dum and Choga Zanbil material) it may be inferred that the practice of votive offering or deposition of a seal was

ECS Corpus, Volume I, Part III 584 restricted to a Middle Elamite date. However, the evidence from Susa, as elusive and ill-formed as it may be, may indicate that the offering of a seal was a practice of greater antiquity, and practiced at least in the ‘Old Elamite’ period. Unfortunately, the evidence from the temple at Susa is frustratingly fragmentary, and as such, whether these seals were true votive offerings or kept in the temple for some other reason (for instance, as administrative seals for use in the economic realm of the temple and its personnel), and indeed whether these items were provenanced in a single contextual unit (stratigraphical chronological) or different levels (and therefore implying different chronological deposition) is unclear. The possibility for the antiquity of the ‘votive seal’ function is intriguing however, and will be further commented on below.

Temple/votive Seals JNRS 1 PEU 8 ARS 2 OBRS 9 PEA 7 PEO 7 UTRS 1 LME 1 36 Table 5.6. Style distribution of the Susa temple/votive seals.

The overall dominance of administrative functions (both non-writing and writing associated) concurs with the general, and traditional, functional interpretation of glyptic material. However, both the existence of funerary and temple/votive items, and the significant number of unattributable function items should be noted, and will be further discussed below (Chapter 6).

5.2 Chogha Mish Chogha Mish contributed one hundred and forty-nine glyptic items to the Corpus. This may be described as a medium sized group in the wider Corpus, and the fifth largest of the sixteen contributing sites. Table 5.7 and Graph 5.5 illustrate the style distribution of Chogha Mish. The stylistic proportions of Chogha Mish generally concur with the chronological and occupational information from that site. Thus the vast majority of the Chogha Mish glyptic items belong to the Susa II period STS style, in accordance with the ‘Protoliterate’ occupation of the site. The small, but not insignificant number of JNRS Susa III period items may indicate that occupation at Chogha Mish continued to some (limited) degree into this subsequent period.

ECS Corpus, Volume I, Part III 585 Alternatively, this may imply that part of the JNRS style belongs in the earlier Susa II period. Both possibilities are feasible, neither is currently assured however.

Style Distribution STS 134 AGD 3 PEU 1 JNRS 10 LPS 1 149 Table 5.7. Chogha Mish style distribution.

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0 STS JNRS AGD PEU LPS Graph 5.5. Chogha Mish style distribution.

While there is evidence at Chogha Mish for Sukkalmah and (beyond the actual realms of this study) Achaemenid period occupation, the later (than the ‘Protoliterate’ contemporary items) two seals were not associated with these contexts but were surface finds. The Chogha Mish seal classified as PEU (2606) may be associated with

ECS Corpus, Volume I, Part III 586 the Sukkalmah period (‘Old Elamite’) occupation at the site therefore, though this is uncertain. If so, this may indicate either a later chronological classification for (part) of the PEU style (in accordance with the already hypothesised extension of the PEU into the early Sukkalmah period outlined above), or that this particular item was an old or ‘heirloom’ seal kept from the previous period into the Sukkalmah (as may be possible given the weathered, somewhat ‘ancient’ appearance of this particular item). The LPS item (3300) may be associated with the first millennium BC occupation of the site, and thus confirms the LPS extension into this era.

Glyptic Type Seals 11 Sealings 138 Table 5.8. Chogha Mish glyptic types.

Seals

Sealings

Graph 5.6. Chogha Mish glyptic types.

Table 5.9 and Graph 5.7 illustrate the glyptic material distribution of the Chogha Mish corpus. As is demonstrated, the largest material type is (with over 34%) of the corpus, followed closely by (over 32%). Bullae, like sealed tablets discussed above, are generally easily recognised and usually so noted. Furthermore, the Chogha Mish excavation report (Delougaz & Kantor 1996) paid special interest and attention to the bullae from Chogha Mish (see Chapter 2), and thus it may assumed that all bullae have been identified and no further items would be added to this distribution pattern. Similarly, the more recent publication of the Chogha Mish excavation report (relative to the older publications, such as Delaporte 1920) means that the sealings were more expertly identified than previous publications (as evidenced by the number and details of the other sealings in this corpus shown in Table 5.9, such as door, jar and bale sealings), implying that those items listed purely as ‘sealing’ evaded identification

ECS Corpus, Volume I, Part III 587 (presumably as they were too fragmentary or non-diagnostic; the fact that both the primary authors were unable to finish the Chogha Mish publication, and that this task fell to Alizadeh, who lacked access to all the material, may mean that these items had not been studied and classified, but with physical examination some may be classifiable however).

Glyptic Materials Black stone 3 Sealing(s) 51 Buff coloured stone 1 Door sealing(s) 15 Stone 1 Jar sealing(s) 6 Tan coloured stone 1 Jar stopper 2 White stone 4 Bottle stopper 2 Faience 1 Bale sealing 3 Flat(tish) sealing 2 Bulla(e) 48 Sealed tablet 9 Table 5.9. Chogha Mish glyptic materials.

Another symptom of the more recent publication of the Chogha Mish material is the fact that most of the seals have not been allocated a material type, but merely supplied with a colour description (the exception being the identification of the single seal). As already outlined above (Chapter 1), this is in accordance with the preferred description of material type in the absence of a satisfactory scientific/petrographic study to facilitate accurate designation. This does however limit the comparative and correlative potential of this material.

55 50 45 40 35 30 25 20 15 10 5 0

e e e e ) s) et ne ne n c (e (s) ( per l ton o to g g(s) p ston sto S s in ing tab k aien ulla l sealing sealing d F B ealin eal r stopper lac red hite S s u r sea Ja ale sh) eale B W ottle sto B ti S oor Ja B colo D lat( f F uf an coloured st B T Graph 5.7. Chogha Mish glyptic materials.

ECS Corpus, Volume I, Part III 588 Table 5.10 and Graph 5.8 illustrate the glyptic function division of the Chogha Mish material. As the function of most of the excavated Chogha Mish architecture and evidence remains unknown, or at least uncertain (see Chapter 2), and little contextual information for the seals is available, all the cylinder seals from Chogha Mish have been given a basic ‘unknown function’ classification. The sealing function has been divided, as outlined above, according to items associated with writing (tablets, and thus in the Chogha Mish case, more correctly proto-writing [proto- ] tablets) and those not, and therefore associated with other areas of administration or control. The division and proportions of these items concurs with the pattern one would expect from a ‘Protoliterate’ site. The reliance on bullae at Chogha Mish and the corresponding chronological position of the site will be further detailed below in relation to the preponderance of this administrative device in this (Chapter 6).

Glyptic Function Non-writing administration 129 Writing administration 9 Unknown function 11 Table 5.10. Chogha Mish glyptic function.

130 120 110 100 90 80 70 60 50 40 30 20 10 0 Non-writing administration Writing administration Unknown Function Graph 5.8. Chogha Mish glyptic function.

ECS Corpus, Volume I, Part III 589 5.3 Haft Tepe The Khuzistan site, Haft Tepe, contributed two hundred and ten glyptic items to the Corpus, the second largest contribution of a single site, second only to the extraordinary Susa. The site of Haft Tepe is indeed itself extraordinary in terms of item numbers in light of the fact that essentially this site was only occupied for a single period (if one conflates the late Sukkalmah and early Middle Elamite period into one occupation phase of Haft Tepe, see Chapter 2). Table 5.11 and Graph 5.9 illustrate the style distribution at Haft Tepe. As is clearly demonstrated, the vast majority (over 90%) of the Haft Tepe corpus belong to the EME style. This distribution concurs both with the periodisation of Haft Tepe, and the definition of the EME style. The single ARS item from Haft Tepe (2252) may be considered an ‘heirloom’ or re-used seal. The OBRS item (2702) may either also be an ‘heirloom’ seal, or a late OBRS item from the late Sukkalmah period occupation at Haft Tepe. The presence of two KRS items (3028, 3043) at Haft Tepe, both already discussed in reference to this style, may indicate the early appearance of this style (in the ME I, rather than the ME II flourishing proposed here). The items may be considered too fragmentary in the first case to be given much credence or notice, or may be considered a true (First) Kassite item in the second instance, thus limiting this possible KRS reassessment. It should also be noted that the two ‘Miscellaneous’ Haft Tepe items (3483, 3484) are ‘Mittanian’, and thus conform with the general chronological definition of Haft Tepe (as this civilisation flourished around the middle of the second millennium BC), and indicates a degree of interaction between the northern Mesopotamian civilisation and Haft Tepe.

Style Distribution ARS 1 No Image 1 OBRS 1 Miscellaneous 2 EME 190 Unclassifiable 7 KRS 2 Not Illustrated 6 210 Table 5.11. Haft Tepe style distribution.

Table 5.12 and Graph 5.10 illustrate the seal/sealing division of the Haft Tepe material. As is evidenced, and as already mentioned, sealings are significantly more numerous than seals at Haft Tepe (over 96%). This is probably, at least partially, due to an extraordinary area of Haft Tepe (an ‘archive’ or ‘craft production’ area) that was excavated.

ECS Corpus, Volume I, Part III 590 200

190

180

170

160

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0

E us RS M RS o ble ted A BRS K a ra O E o Image lust N Il iscellane nclassifi ot M U N Graph 5.9. Haft Tepe style distribution.

Glyptic Type Seals 7 Sealings 203 Table 5.12. Haft Tepe glyptic types.

ECS Corpus, Volume I, Part III 591 Seals

Sealings

Graph 5.10. Haft Tepe glyptic types.

Table 5.13 and Graph 5.11 illustrate the glyptic material of Haft Tepe. Like the Chogha Mish seal corpus, only faience seals have been identified in the Haft Tepe material (indeed, as already mentioned, Negahban originally classified these seals as , which has been updated to here, according to the description of the material given by Negahban [see Chapter 1]), with the other seals only listed according to colour, a phenomenon that is scientifically preferable, but limits comparative potential. As can be demonstrated, is by far the largest material type (over 82%). As is evidenced by the list of sealing types included in Table 5.13, no identification of jar, door, bale (and so on) sealings have been made for the Haft Tepe material. It may be assumed that all the available tablets (and envelopes) have been recognised (according to the above outlined easily identifiable nature of these items). The presumably included amongst it various jar, bottle, bale, door and wall sealings (as well as non- diagnostic sealings). The exact function of the two (a unique Haft Tepe, or probably more correctly, Negahban, category) is unknown, and cannot be reconstructed from the published image alone. The two should also be noted, though whether these items were deliberately fired for preservation purposes, or accidentally heated in some unidentified conflagration is unknown, and limits the consideration of these items.

Glyptic Materials Black stone 1 Sealing(s) 173 Grey stone 1 Sealed clay cone 2 Dark grey stone 1 Fired clay sealing 2 Sealed envelope 7 Faience 4 Sealed tablet(s) 19 Table 5.13. Haft Tepe glyptic materials.

ECS Corpus, Volume I, Part III 592 175 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

e e c (s) ng (s) one g i et ston st aien eal l k F s rey stone ealin envelope tab lac G S clay cone lay d d B d c ark grey eale eale D eale ired S S S F

Graph 5.11. Haft Tepe glyptic materials.

Table 5.14 and Graph 5.12 illustrate the proposed glyptic function of the Haft Tepe glyptic items. The sealed envelopes were conflated with the sealed tablets in the administrative material associated with writing, as these items are generally themselves inscribed with writing, or encased inscribed tablets, see below for details. As with the Chogha Mish seals, no detailed provenance or contextual information was

ECS Corpus, Volume I, Part III 593 available for the Haft Tepe seals, and so all have been conflated under an rubric. As can be seen, the vast majority of the Haft Tepe glyptic items can be shown to have a non-writing administration function. In the context of Haft Tepe, this function may be especially identified as a craft production (and raw material distribution) function (as detailed in Chapter 2, and further discussed below).

Glyptic Function Non-writing administration 177 Writing administration 26 Unknown function 7 Table 5.14. Haft Tepe glyptic function.

130 120 110 100 90 80 70 60 50 40 30 20 10 0 Non-writing administration Writing administration Unknown Function Graph 5.12. Haft Tepe glyptic function.

5.4 Choga Zanbil Choga Zanbil contributed one hundred and seventy-six items to the Elamite Cylinder Seal Corpus, the fourth largest site corpus (after Susa, Haft Tepe and Surkh Dum-i- Luri; see Graphs 4.85, 4.86). Table 5.15 and Graph 5.13 illustrate the style distribution of the Choga Zanbil corpus. As is clearly demonstrated, and as already mentioned, the majority of Choga Zanbil items belong to the LME style (over 53%). This distribution conforms to the late Middle Elamite (post ME I) date for Choga Zanbil. The significant contribution of other contemporary late Middle Elamite styles (KRS, AS, LPS and LGD) also confirms and conforms to this date. All told, some one hundred and thirty-two (or 75%) of the Choga Zanbil corpus has a late Middle Elamite stylistic date, in accordance with the general later Middle Elamite period apogée of the site. The twenty-one ‘Not Illustrated’ Choga Zanbil items also

ECS Corpus, Volume I, Part III 594 reportedly conform stylistically to what would be labelled, in the current study KRS and LGD (see Chapter 4 for the explanation of the rejection of the stylistic designation of these items), further adding to the general late Middle Elamite date of the Choga Zanbil corpus. The nine Choga Zanbil items that possess an earlier stylistic date (one AGD item, one OBRS item, four PEO items and three EME items; just over 5% of the Choga Zanbil corpus) may be considered ‘heirloom’ or ancient seals that were deposited at Choga Zanbil (due to the established foundation date of Choga Zanbil, it may be assumed that the actual act of deposition occurred in the [late] Middle Elamite period). The presence of these ‘ancient style’ seals at Choga Zanbil implies that not all seals so deposited at the site were necessarily made for this express purpose, a point that will be detailed further below (Chapter 6).

Style Distribution AGD 1 LME 94 OBRS 1 AS 1 PEO 4 LPS 4 EME 3 LGD 15 Unclassifiable 14 KRS 18 Not Illustrated 21 176 Table 5.15. Choga Zanbil style distribution.

95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

E E S EO M RS M A PS GD ble ted BRS P K L L a ra O E L lust Il nclassifi ot U N Graph 5.13. Choga Zanbil style distribution.

ECS Corpus, Volume I, Part III 595 As already mentioned, no sealings whatsoever were uncovered at Choga Zanbil. This is generally considered indicative of the glyptic and, indeed the overall site, function. Table 5.16 and Graph 5.14 demonstrate the materials of the Choga Zanbil seals. As is clearly indicated, the vast majority of the Choga Zanbil materials are faience items (over 72%). Faience is a significant material in the whole context, and can be considered characteristic of the Elamite Corpus generally, not just Choga Zanbil (see Graph 4.83). However, the significant Choga Zanbil faience corpus may indicate the presence of a faience production industry at the site (possibly located in the annex of the Ishni-qarab Temple [see Chapter 2]). The second most numerous material at Choga Zanbil is (albeit at over 15%, a significantly smaller group than ). Glass is, like faience, a quartz-based ‘artificial’ material, that is related to, but distinct from, faience (see Chapter 1). Glass seals are only found at Choga Zanbil in the Corpus, and thus this material may be considered characteristic of this site. Quartz-based ‘artificial’ materials (that is, faience and glass) thus account for the significant majority of the Choga Zanbil corpus (one hundred and fifty-four items, or over 87%). The variety of other, smaller seal materials indicates that traditional stone seals (and not only ‘artificial’ quartz-based materials) were also used at Choga Zanbil.

Glyptic Materials Limestone 2 Grey stone 2 Marble 4 Terra cotta 1 Steatite 1 Bronze 1 Rose quartz 1 Bitumen aggregate 4 Alabaster/gypsum 1 Faience 127 Apricot (coloured) stone 2 Glass 27 Brown stone 1 Unknown cylinder 2 Table 5.16. Choga Zanbil glyptic materials.

As no sealings were found at Choga Zanbil, any functional definition of the Choga Zanbil material was based entirely on provenance and context. Table 5.17 illustrates the various archaeological contexts in which seals were found at Choga Zanbil. As is illustrated, most of the cylinder seals from Choga Zanbil were found in cultic or votive contexts, namely in the various ‘Chapels’ and ‘Temples’, including the itself (see Chapter 2 for details of these areas). These areas were not the only contexts in which seals were found however, including one in a tomb, and several in the ‘palace’ area (see Figures 2.8 and 2.9). Table 5.18 and Graph 5.15 illustrate the glyptic function of the Choga Zanbil seals. The items from the various

ECS Corpus, Volume I, Part III 596 ‘Chapels’, the Ziggurat and the temples have all been given a votive designation. This designation is certain for the ‘Chapels’, and particularly Chapels III and IV that line the southwestern façade of the Ziggurat (see Chapter 2 for details, and Figure 2.9), but is also generally assumed for the other ‘cultic’ areas.

130

120

110

100

90

80

70

60

50

40

30

20

10

0

e e e s r l e m n ze te c e one n a las arb cotta st teatit sto ro aien G M S B greg F ylind ime wn rey stone g c L ose quartz erra ro G T wn R B o labaster/gypsu itumen a nkn A B U

Graph 5.14. Choga Zanbil glyptic materials.

Glyptic Context Ziggurat 3 Chapel E 2 Tomb 1 Palace 16 Chapel I 2 Temple 5 Chapel II 4 Chamber M 4 Chapel III 46 Chamber 1 1 Chapel IV 72 Other context 20 Table 5.17. Choga Zanbil glyptic context.

Glyptic Function Funerary 1 Palace 16 Votive 134 Unknown Function 25 Table 5.18. Choga Zanbil glyptic function.

ECS Corpus, Volume I, Part III 597 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Funerary Votive Palace Unknown Function

Graph 5.15. Choga Zanbil glyptic function.

The ‘funerary’ function of the single Choga Zanbil seal (2733) from a tomb is obviously assured (the fact that this item is here classified as PEO, and therefore has a Sukkalmah period stylistic date, and is thus an ‘heirloom’ seal, should be noted). The sixteen seals provenanced from the Palace area in the eastern sector of Choga Zanbil (see Chapter 2 and Figure 2.8) are of interest and have been classified separately from the other context items of unknown function. The precise function of this ‘palatial’ area (and the accuracy of this designation) is unknown, though some monumentality and occupation may be discerned (see Chapter 2). The exact function of these ‘Palace’

ECS Corpus, Volume I, Part III 598 items are therefore similarly unclear, though it is possible that some of these seals at least may have had a generally, traditional, administrative function. The fact that no items had a clear or definite ‘administrative’ function at Choga Zanbil warrants reiteration, and is linked to the absolute absence of sealings at Choga Zanbil. Indeed, the entire site of Choga Zanbil, not just the glyptic realm, had an unique, and specifically cultic or votive, rather than administrative function.

5.5. Tepe Sharafabad Tepe Sharafabad contributed only five items to the Corpus, and thus represents the smallest of the excavated Khuzistan corpora (excluding the single surface find from Deh-i Now). Table 5.19 and Graph 5.16 illustrate the style distribution of the Tepe Sharafabad corpus. As outlined above, two distinct phases of occupation at Tepe Sharafabad are of relevance to this study, the first is labelled ‘’ by the excavators, and corresponds to the current Susa II period, the second is labelled ‘Elamite’ and roughly corresponds with the later Sukkalmah and early Middle Elamite (ME I) period of this study (see Chapter 2 for details). The three STS items from Tepe Sharafabad were sourced form the ‘Uruk’ Sharafabad context, and thus conform with this chronological designation. The single PEO item from Sharafabad was found in the ‘Elamite’ context, and similarly accords chronologically with this classification. Indeed, the inscription on the PEO seal, possibly a ‘dated seal’ with reference to ‘Tata’ a sukkal (see above for details), further confirms the Sukkalmah period designation for this item, and therefore is in agreement with its excavated context. The PEA item is somewhat more problematic. It was found in the ‘Elamite’ context, though stylistically dates to an earlier ‘Akkadian and Awan’ phase. It must be assumed therefore that this item is an old, antique or ‘heirloom’ seal in the Sukkalmah/Middle Elamite ‘Elamite’ Sharafabad context.

Style Distribution STS 3 PEA 1 PEO 1 5 Table 5.19. Tepe Sharafabad style distribution.

ECS Corpus, Volume I, Part III 599 3

2

1

0 STS PEA PEO

Graph 5.16. Tepe Sharafabad style distribution.

Table 5.20 and Graph 5.17 illustrate the glyptic type division of the Tepe Sharafabad corpus. This division (two seals, three sealings) conforms exactly to the chronological division of the Sharafabad corpus (that is, the three sealings date to the Susa II [STS] style, the two seals belong to the later ‘Old Elamite’ period). Rather than a strict functional explanation for this division, a contextual justification may be proposed. The three Susa II sealings were uncovered in a dump or trash pit, and thus can be described as discarded rubbish items. The two seals were found in occupational contexts. Thus the seal/sealing division appears to be contextual and archaeologically explained, and not functional.

Glyptic Type Seals 2 Sealings 3 Table 5.20. Tepe Sharafabad glyptic types.

Seals

Sealings

Graph 5.17. Tepe Sharafabad glyptic types.

Table 5.21 illustrates the glyptic material proportions of the Sharafabad corpus. As each material type demonstrates the same numerical number (that is, each has only one example), a graphical representation of this division has not been included as this is deemed to be redundant. As can be seen, the PEA seal (2339)

ECS Corpus, Volume I, Part III 600 provides one of the few non-Susian bitumen aggregate seals of the Corpus. The presence of this isolated seal indicates both that this material was not strictly limited to Susa (contra Conan & Deschesne 1996, Chapter 1), and the close links between neighbouring sites of Susa and Sharafabad. The fact that no were recorded in the Sharafabad corpus indicates the excellent and detailed publication of these items. It should be noted that none of the sealed items can be associated with writing, and thus demonstrate a ‘non-writing’ administrative function at Sharafabad. It is also of interest that none of these items are sealed bullae, a material type later discussed as particularly an early (that is Susa II) phenomenon. Given the extremely limited numbers of the available Sharafabad corpus, and the limited excavation and exposure at the site, both the absence of bullae and of sealed tablets should not be treated as any genuine evidence for an alternate function or serious reassessment of the chronological scheme, but rather a genuine accident and circumstance of discovery. Indeed, for the later Sukkalmah period ‘Elamite’ occupation at Sharafabad, the presence of an inscribed seal, and tablets (albeit reportedly unsealed) indicates some literate administrative action at Sharafabad. No real conclusion regarding the seal function can be drawn, though the lack of a tomb or temple context for both items would seem to imply a general administrative function for each, though without any concrete evidence, the simple disallowing of a votive or funerary context cannot be taken as proof of administrative control function.

Glyptic Materials Steatite 1 Door lock sealing 1 Bale sealings 1 Bitumen aggregate 1 Jar sealing 1 Table 5.21. Tepe Sharafabad glyptic materials.

5.6. Deh-i Now Little conclusions or analysis may be provided for the single surface find seal from Deh-i Now. This single seal (2890) clearly forms the smallest site corpora, and was included for the sake of completeness, despite its uncertain context and surface collection nature. Little may be said regarding the material of this seal without physical (petrographic) analysis, as it was reportedly cut from (Amiet 1972: 262), a term that, as already discussed, does not in fact accurately indicate a known stone or rock, but rather, a stone classification (see Chapter 1 for details). This single seal is given an EME (6) stylistic classification, and therefore an early Middle

ECS Corpus, Volume I, Part III 601 Elamite stylistic date. This conforms generally with the Middle Elamite proposed occupation of the site (as evidenced by other surface finds, most clearly inscribed bricks name Middle Elamite kings), though not strictly (as the EME style is believed not to have continued in any great degree beyond the ME I period, and the Deh-i Now bricks speak of ME II and III period officials). Due to the general fragmentary nature of this information, and the lack of any excavation and exposure at the site to indicate extent or duration of occupation, it may be argued however that the chronological information provided by the seal and the inscribed bricks are generally coincident. Of course, given the lack of excavated context and any excavated remains at Deh-i Now, no comment on the nature or function of this item, or glyptic material generally at this site may be made here.

5.7. Surkh Dum-i-Luri Surkh Dum-i-Luri contributed the third largest site corpus to the study (after Susa and Haft Tepe), and the significant majority of the Luristan material. Table 5.22 and Graph 5.18 illustrate the style distribution of the Surkh Dum glyptic corpus. As can be seen, Surkh Dum contains a great variety of stylistic items, indeed, only the continuously occupied Susa demonstrates a more thorough stylistic survey. The styles not represented at Surkh Dum include the PEU and AS, as well as three of the four miscellany categories (‘No Image’, ‘Miscellaneous’ and ‘Not Illustrated’, the last of which may be considered a publication factor, not an archaeological issue). As already outlined above (Chapter 2), the different selection criteria for the Surkh Dum corpus, in which items were included not purely on the basis of their provenance within the ‘Elamite’ realms of this study, but by stylistic classification, caused the lack of ‘Miscellaneous’ styles included in the Surkh Dum corpus. Table 2.18 illustrates the Surkh Dum items from other (non-Elamite) contemporary styles, which from any other site would have been included in this category (see Chapter 2 for the justification of their non-inclusion). Thus at least one true Kassite seal, and several Middle and Neo-Assyrian items were found at Surkh Dum but not included in the Corpus, there presence should be noted however. The Surkh Dum-i-Luri ‘sanctuary’ was founded in the Middle Elamite (II) period, with its true flourishing in the later first millennium ‘rebuilding’ (see Chapter 2 for details). Thus, all the items with stylistic dates before the Middle Elamite period may be considered ancient or ‘heirloom’ seals. These include the single JNRS, CPE,

ECS Corpus, Volume I, Part III 602 GS, AGD, STF, LSF, ARS and PEA items, as well as the seven SF, two UTRS, six OBRS, two PEO and the five EME items (the last of which, may be considered only fractionally older than the first Surkh Dum occupation period). Thus thirty of the Surkh Dum items (or slightly over 16%) may be considered chronologically misplaced, or more correctly, stylistically ancient at the time of their deposition. A similar process of deposition as seen at the other ‘votive’ site of Choga Zanbil may be discerned here, though the fact that a greater proportion of ‘heirloom’ seals were deposited at Surkh Dum than at this site should be noted (just over 3% of the Choga Zanbil corpus may be considered ‘heirloom’ seals, compared to over 16% of the Surkh Dum corpus). It should also be noted that the majority of the Surkh Dum ‘ancient’ seals belonged to Mesopotamian ‘Related’ styles, rather than Elamite classifications (in the Surkh Dum corpus, JNRS, GS [a ‘Shared’ style in fact], SF, ARS, UTRS and OBRS, eighteen items in total, or nearly two-thirds of the ancient seals; the two KRS, contemporary occupation items should also be noted in this survey). This may aid in the general reconstruction of the cultural constitution of Surkh Dum, or at least demonstrate its location along the Elamite-Mesopotamian border lands, beyond the true (variously Khuzistan and Fars) Elamite ‘heartland’. The presence of ‘Elamite’ styles at Surkh Dum (the CPE, LSF, PEA, PEO, EME, and the contemporary LME) should also be noted however, and indicates that the site was indeed at least partially within the ‘Elamite’ realm.

Style Distribution JNRS 1 UTRS 2 CPE 1 OBRS 6 GS 1 PEO 2 AGD 1 EME 5 STF 1 KRS 2 SF 7 LME 4 LSF 1 LPS 118 ARS 1 LGD 24 PEA 1 Unclassifiable 3 181 Table 5.22 Surkh Dum-i-Luri style distribution.

ECS Corpus, Volume I, Part III 603 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

S S F E E R PE G GD TF S SF RS EA EO M RS M PS GD ble C S L A P TRS BRS P K L L a JN A U O E L

nclassifi U

Graph 5.18. Surkh Dum-i-Luri style distribution.

The styles represented at Surkh Dum-i-Luri that may be considered contemporaneous with the main occupation of the site include the KRS (two items), the LME (four), the LPS (one hundred and eighteen) and the LGD (twenty-four), a total of one hundred and forty-eight seals. As one would expect, and indeed as is a defining factor, the significant majority of these items, and the wider Surkh Dum corpus, are classified as LPS. This style demonstrates stylistic links with both Elamite and Mesopotamian styles (see Chapter 4 for details), but is classified as ‘Elamite’ in the sense of ‘Iranian’ here, as it is deemed to be indigenous to (specifically Luristan). The glyptic material thus demonstrates that the Surkh Dum society had links with both ‘Elam’ proper (Khuzistan and Fars), and , as demonstrated in the distribution of the glyptic items, but generally was the location of

ECS Corpus, Volume I, Part III 604 a distinct, though related, Luristan culture (represented by the majority LPS items), that was neither purely Mesopotamian, nor truly Elamite. Like Choga Zanbil, and as already mentioned, all the Surkh Dum glyptic items are seals, with no sealings whatsoever uncovered at the site. Table 5.23 and Graph 5.19 illustrate the glyptic materials of the Surkh Dum seals. The great variety of material types illustrated by these figures reflects the general stylistic variability of the corpus. The significant majority contribution of should be noted (over 46%). Elsewhere, this material has been described as characteristically ‘Elamite’ and significantly represented in the middle and second half of the second millennium (that is the late Sukkalmah and Middle Elamite periods of this study; it should be noted that faience is by no means restricted to this era alone however, and is well known throughout the chronological span of this study). In the Surkh Dum context the large number of faience items may be considered evidence both for an ‘Elamite’ influence/element at this site, and for the general chronological placement of the site.

Glyptic Materials Limestone 7 Buff stone 1 Marble 1 Green stone 1 Sandstone 1 Grey stone 1 Lapis lazuli 1 Red stone 1 Steatite 1 White stone 3 Rock crystal 1 Baked clay cylinder 1 Carnelian 1 Bone 5 Agate 1 Shell 1 Alabaster/gypsum 13 Fossilised coral 1 Chalk 1 Ivory 1 Hematite 9 Bronze 1 Serpentine 15 Copper 9 Black stone 5 Bitumen aggregate 8 Brown stone 6 Faience 84 Table 5.23. Surkh Dum-i-Luri glyptic materials.

ECS Corpus, Volume I, Part III 605 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

e e n m ite er ce o at tone n est psu s stone Bon opp regate gy n te stone C Faie im Hem L er/ erpentine agg S Black row ast B Whi en itum Alab B Graph 5.19. Surkh Dum-i-Luri glyptic materials, with three or more items.

5.8. Kamtarlan Kamtarlan (I and II) contributed eight glyptic items to the Corpus, a reasonably small, but not insignificant, and generally average, number in terms of the majority of the Luristan corpora (excluding the exceedingly large Surkh Dum-i-Luri). Table 5.24 and Graph 5.20 illustrate the style proportions of the Kamtarlan items. The stylistic classification here presented essentially concurs with the occupation chronology of Kamtarlan (see Chapter 2 for details of Kamtarlan). The three JNRS (732 – 734) items were all found in Kamtarlan I, Plot P, in a ‘Jemdet Nasr’, Susa III context, and thus concur with the Susa III stylistic date for the JNRS. The lone LSF item (2111) was found in a Susa IV context of Kamtarlan II (Plot F), and thus similarly concords to a (late) Susa IV stylistic date for this item. Three of the four LPS items (3343, 3344, 3378) were found in terminal second millennium/early first millennium contexts at Kamtarlan I (Plot M), and thus conform with the general stylistic date for this style. The last Kamtarlan item (3377), is somewhat more problematic, as it was uncovered in Plot F of Kamtarlan II, an area above described as a Susa IV context (and thus too early by a significant margin to include a true LPS item). This misplaced chronological item may be explained in one of two ways. Firstly, it is possible that the

ECS Corpus, Volume I, Part III 606 context of this item at Kamtarlan has been mislabelled or wrongly attributed. While the already cited LSF item (2111) would appear to confirm the Susa IV date for this context, the general scarcity and stratigraphical/archaeological problems and difficulties of the Kamtarlan evidence (detailed above, Chapter 2) may allow for such a misdating (indeed, an LSF item in an LPS context is not impossible, as this item may be considered an ‘heirloom’ seal, while the reverse, an LPS item in an LSF context, cannot be adequately explained). The use of seals to (re-)date archaeological contexts and levels has already been rejected however, and thus this is merely a suggestion, made possible by the general poor and fragmentary available Kamtarlan information. The other possibility is that the LPS item from Kamtarlan has been wrongly allocated. The fragmentary, incomplete, nature of this item (3377) makes this explanation possible, but this fragmentary appearance also prevents any adequate resolution to this problem. Regardless of the explanation of this misplaced Kamtarlan item however, the general picture of the Kamtarlan corpus is one of agreement between the archaeological (occupational) and stylistic chronology. The dominance of the Luristan ‘indigenous’ style (LPS), demonstrate this constitution of Kamtarlan, as does the presence of three items in a Mesopotamian ‘Related’ style (JNRS), and one in an ‘Elamite’ style (LSF). The small number of this corpus prevents any real cultural designations however.

Style Distribution JNRS 3 LSF 1 LPS 4 8 Table 5.24. Kamtarlan style distribution.

4

3

2

1

0 JNRS LSF LPS

Graph 5.20. Kamtarlan style distribution.

Table 5.25 and Graph 5.21 illustrate the seal/sealing division of the Kamtarlan corpus. As is illustrated, a single sealing alone was uncovered in the Kamtarlan

ECS Corpus, Volume I, Part III 607 excavations. The relatively small scale exposure at this site, and the age of these excavations (that is, before an era when sealings were regularly discerned and noted in excavations) limits the trustworthiness and use of this division however. The fact that this single sealing is of (as indicated by Table 5.26 and Graph 5.22), as opposed to the more regular unbaked (or fired) clay sealings may further explain why this item was discerned in the 1930’s era excavations of Kamtarlan. Whether this item was deliberately baked for preservation purposes or as the result of an accidental firing cannot be known, especially in light of the absence of any other sealings to act as a control. The number of faience items should also be noted, and, as with Surkh Dum above, may indicate an ‘Elamite’ element in this Luristan corpus.

Glyptic Type Seals 7 Sealings 1 Table 5.25. Kamtarlan glyptic types.

Seals

Sealings

Graph 5.21. Kamtarlan glyptic types.

Glyptic Materials Limestone 1 Faience 3 Marble 1 Clay cylinder 1 Alabaster/gypsum 1 Baked clay sealing 1 Table 5.26. Kamtarlan glyptic materials.

3 2 1 0

le r g b sum estone p linde Mar Faience cy Lim er/gy Clay clay sealin ed labast A Bak

Graph 5.22. Kamtarlan glyptic materials.

ECS Corpus, Volume I, Part III 608 The single sealing from Kamtarlan may be given a ‘non-writing’ associated administration function, as illustrated by Table 5.27 and Graph 5.23. The general lack of any significant contextual information (or interpretation) at Kamtarlan prevents any functional allocation to any of the seals from the site. Again, due to the dearth of information, and the small Kamtarlan data set, no real glyptic function reconstruction can be proposed here. The fact that cylinder seals were apparently used (in at least one recorded case here) to seal items should be noted however (especially in light of many of the other Luristan sites [Surkh Dum-i-Luri, Chigha Sabz, Kalleh Nisar, Bani Surmah, Tepe Djamshidi], that lack any such administrative glyptic function evidence).

Glyptic Function Non-writing administration 1 Unknown Function 7 Table 5.27. Kamtarlan glyptic function.

7 6 5 4 3 2 1 0 Non-writing administration Unknown Function Graph 5.23. Kamtarlan glyptic function.

5.9. Chigha Sabz Like the site of Kamtarlan discussed above, the nine item corpus from Chigha Sabz is small by comparison to the large corpora (from Susa, Chogha Mish, Haft Tepe, Choga Zanbil and Surkh Dum-i-Luri, and to a lesser degree, Tal-i Malyan), though reasonably sized in the Luristan context. Table 5.28 and Graph 5.24 illustrate the style distribution of Chigha Sabz. As can be seen, the majority of the Sabz items belong to the LPS style, with another (LGD) belonging to a similar chronological time frame. The single GS and AGD items belong to an earlier chronological period, though the general difficulty in allocating a date to the AGD classified items should be recalled. In this regard, only the GS item is certainly misplaced chronologically. Despite the fact that Chigha Sabz was reportedly occupied for several phases, from before the

ECS Corpus, Volume I, Part III 609 earliest period of this study, though to its end, all the Chigha Sabz seals were found in apparent hoards, dug into these earlier levels of occupation (see Chapter 2 for details). These hoards have been associated with the Iron Age III ‘camp site’ occupation of Chigha Sabz, and while a general first millennium date may be proposed for these hoards on the basis of correlative material (as well as the seals, ceramics and metals were also found in these hoards), the exact chronological classification cannot be agreed (due to scarcity of information). In this regard, the LPS items may be shown to have been discovered in a context generally coincident with their stylistic date, while the GS (and the difficult to ascribe AGD and, possibly also the LGD) items may be considered ‘heirloom’ seals despotised in these hoards. The fact that three seals from Chigha Sabz, deemed not to belong to any style of the Corpus, but to be (pure, rather than partly as the LPS style is) first millennium in date should also be noted (see Table 2.21). It is also interesting to note that all the non-LPS or non-first millennium Chigha Sabz items (that is the GS, AGD and LGD items), all belong to geometric styles. A possible functional or typological (that is, bead rather than seal) proposal may be advocated for some of these items, but cannot currently be proven, but only hypothesised.

Style Distribution GS 1 LPS 6 AGD 1 LGD 1 9 Table 5.28. Chigha Sabz style distribution.

6 5 4 3 2 1 0 GS AGD LPS LGD Graph 5.24. Chigha Sabz style distribution.

Table 5.29 and Graph 5.25 illustrate the seal materials of the Chigha Sabz corpus. As already mentioned, all the Chigha Sabz items are seals, with no sealings found in the hoards, or elsewhere in the excavated areas of the site. The significant contribution of as a material, as with the Surkh Dum corpus, indicates that

ECS Corpus, Volume I, Part III 610 the Chigha Sabz items both generally date to the later second and early first millennium BC (a period of significant exploitation of this material type), and were at least partially within the ‘Elamite’ sphere of influence (or perhaps more correctly in this instance, the LPS style which was favoured at Chigha Sabz was an ‘Elamite’ influenced style). It should also be remembered that the material is not an accurate designation of a seal stone or mineral (see Chapter 1 for details), but is merely perpetuated here as no physical (or petrographic) analysis has been undertaken to rectify this poorly attributed material.

Glyptic Materials Chlorite 1 Steatite 1 Faience 5 Porphyry 2 Table 5.29. Chigha Sabz glyptic materials.

5

4

3

2

1

0 Chlorite Faience Steatite Porphyry Graph 5.25. Chigha Sabz glyptic materials.

As already detailed, all the Chigha Sabz seals were found in, presumably Iron Age, hoards dug into old, abandoned Chigha Sabz occupation levels. This process may be related to the hoard deposition of Surkh Dum-i-Luri discussed above, though no votive or cultic designation may be inferred from this practice. Rather, the Chigha Sabz hoards appear to have been dug as a means of safe-keeping or protection, either for general purposes of storage, or against some ravaging foe. The Chigha Sabz items thus have a specific, and unique within the Corpus, function. The possible administrative or other functional life of these items cannot be ascertained on the current available information, nor can the absence of sealings at Sabz be taken as absolute evidence for a non-administrative function at the site, given the era of excavation of the site, and its limited exposure.

ECS Corpus, Volume I, Part III 611 5.10. Bani Surmah Nine cylinder seals from Bani Surmah were included in the Corpus, and thus the same comments regarding distribution and relative numbers as stated regarding Kamtarlan and Chigha Sabz apply here. As already noted (Chapter 2), some of the graves at Bani Surmah were not excavated, though it is assumed that these non-explored tombs had previously been raided and emptied by looters. Thus, unlike many other sites within the Corpus (including Chogha Mish, Haft Tepe, Tepe Sharafabad, Surkh Dum-i-Luri, Kamtarlan, Chigha Sabz, and indeed Susa to a certain degree), the total remaining and preserved number of seals may be deemed to have been discovered from Bani Surmah. This implies that cylinder seals were generally not common as grave items at Bani Surmah, as can be judged by the proliferation of other artefact groups (see Chapter 2 and Haerinck & Overlaet 2006), and that the Bani Surmah corpus was finitely small within the Elamite Corpus as a whole. Table 5.30 and Graph 5.26 illustrate the style distribution of the Bani Surmah Corpus. The majority of these items belong to the SF style, and therefore may be given the stylistic date within the Susa IV (or Early Dynastic) period. The single JNRS item manifests a Susa III stylistic date, and the two AGD items may be considered generally ‘early’ in date (that is pre-Akkadian). This chronological pattern concurs with the dates and classifications of the other Bani Surmah grave goods (see Chapter 2), and thus the general date for the inhumations. It may be hypothesised therefore that these items were deposited generally within or around the period of their general use (that is, they were not ancient or ‘heirloom’ seals when deposited, as seen with some of the votive items at Choga Zanbil and Surkh Dum).

Style Distribution JNRS 1 AGD 2 SF 6 9 Table 5.30. Bani Surmah style distribution.

ECS Corpus, Volume I, Part III 612 6 5 4 3 2 1 0 JNRS AGD SF

Graph 5.26. Bani Surmah style distribution.

The excavators believe that all the Bani Surmah seals were most likely of Mesopotamian manufacture (Haerinck & Overlaet 2006: 51), a claim reinforced by the classification of these seals by Haerinck and Overlaet (2006) and Tourovets (1996; see Table 2.22, and the Catalogue entries of the individual Bani Surmah seals). As illustrated by Table 5.30, apart from the two AGD1 classified seals, all the Bani Surmah cylinder seals do indeed belong to styles classified as ‘Mesopotamian Related’ (JNRS and SF). This thus, to some degree, conforms to the proposal of Haernick and Overlaet that the seals were of a Mesopotamian nature. However, as discussed in Chapter 4, the Mesopotamian designs and characteristics of the ‘Mesopotamian Related Styles’ do not automatically allow for a Mesopotamian-origin designation for the individual seals. It is possible that the Bani Surmah Mesopotamian-Related cylinder seals were created by local (that is ‘Elamite’ or ‘Iranian’) craftsmen who were operating within a Mesopotamian-Related lexicon, and do not represent true Mesopotamian imports. Of course the precise location of the area of manufacture for the Bani Surmah material, whether in an as yet unrecognised Luristan occupation area, a (semi-)nomadic camp or from another area of the ‘Elamite’ realm (such as Khuzistan), cannot be ascertained at the moment, especially in light of the fact that the identity of the people who used the Bani Surmah graves is unknown. Rather, it should only be noted that the Mesopotamian-like nature of the cylinder seals of Bani Surmah conform to the noted groups of ‘Mesopotamian Related’ Elamite Styles, and as such belong to a wider stylistic phenomenon in

1As the AGD classification is one of geometric designs (see Chapter 4), it lacks a specific Mesopotamian, Elamite or ‘Shared Mesopotamian-Elamite’ nature, as there is no specific marker or depiction that would relegate it to any one group. As such the AGD seals from Bani Surmah could equally have been of Elamite or Mesopotamian origin. Thus these notoriously difficult to classify geometric designs are not included in the brief discussion that follows. The possibility, in this grave context especially, that AGD items may be beads (see Chapter 4), can also not be disregarded here.

ECS Corpus, Volume I, Part III 613 Elamite glyptic development. Thus, it cannot be ascertained whether the Bani Surmah glyptic items were indeed true Mesopotamian-origin imports, were manufactured locally (within the general vicinity of Bani Surmah) by ‘Elamites’ (used here only in accordance with our above wide definition of the area under study here, perhaps a better term may be ‘Iranians’) working in a ‘Mesopotamian Related’ lexicon, or were created by other Elamites (such as those from Khuzistan [Susa]) working in this same lexicon and imported to Luristan. As already mentioned, and as one would expect from a site that only contained graves, all the Bani Surmah glyptic items are seals, with no sealings found (indeed, the author knows of no sealings associated with a tomb or grave any where, in Iran certainly). Table 5.31 and Graph 5.27 illustrate the material proportions of the Bani Surmah seals. There is nothing of particular note or exception in this distribution pattern, as these materials conform to what one might expect of the period and styles represented at Bani Surmah, including the reliance on ‘soft’ materials in this early period.

Glyptic Materials Limestone 5 Alabaster/gypsum 2 Steatite 2 Table 5.31. Bani Surmah glyptic materials.

5

4

3

2

1

0 Limestone Alabaster/gypsum Steatite

Graph 5.27. Bani Surmah glyptic materials.

Due to the grave/tomb provenance of these items, quite obviously a ‘funerary’ function may be assumed for all the Bani Surmah seals. The process of their deposition, including their possibly earlier function as true ‘administrative’ devices or seals cannot be adequately judged in light of the fact that the identity and location of the Bani Surmah buried and buriers is unknown. The funerary nature of cylinder seals in ‘Elam’ will be further addressed below however. Table 5.32 illustrates the specific

ECS Corpus, Volume I, Part III 614 tomb context distribution of the Bani Surmah seals. It should be remembered in this regard that all the Bani Surmah tombs here represented were multiple (possibly generational) inhumations, meaning that the four seals from A14 for instance should not all be considered to belong to a single individual.

Glyptic Context A2 2 A14 4 B6 2 B13 1 Table 5.32. Bani Surmah glyptic context.

5.11. Kalleh Nisar Kalleh Nisar contributed seven items to the Corpus, and thus once again the same proportional comments as already discussed regarding the other (not Surkh Dum-i- Luri) Luristan corpora apply. Like Bani Surmah, all the Kalleh Nisar glyptic items are seals, from grave/tomb contexts, with no sealings nor seals found in association with the extremely limited occupational architecture found in Areas C and D of Kalleh Nisar (see Chapter 2). Also like Bani Surmah, all the known and remaining (not looted) seals from Kalleh Nisar have been published, and thus this corpus can be considered representative. Table 5.33 and Graph 5.28 illustrate the style distribution proportions of the Kalleh Nisar corpus. The chronological distribution of these styles (from the earliest JNRS, Susa III period style through to the OBRS/PEO, Sukkalmah period styles) corresponds adequately to the periods of use of the Kalleh Nisar graves, as gleaned from correlative evidence, detailed above (Chapter 2; from the Susa III through to Sukkalmah [Old Babylonian] periods, inclusive of the Susa IV [Early Dynastic], ‘Akkadian and Awan’ [Akkadian and post-Akkadian] and ‘ III and Shimashki’ [Neo-Sumerian] periods). Generally, like the Bani Surmah material, most of the Kalleh Nisar seal styles belong to ‘Mesopotamian Related’ styles (the single items each from the JNRS, SF and ARS styles, and the two OBRS items). However, two of these seals belong to styles characterised as ‘Elamite’ (the STF and PEO, seals 728 and 2771 respectively). The apparent Elamite focused nature (rather than Mesopotamian) of the styles of these two seals further raises questions of the nature of the apparent Mesopotamian source of the stylistic influences (and indeed the origin of the seals themselves) on the seals of Kalleh Nisar (and the related Bani Surmah), and

ECS Corpus, Volume I, Part III 615 indicates some interaction with the true ‘Elamite’ realm for these Luristan sites (further discussed below [Chapter 7]).

Style Distribution JNRS 1 ARS 1 STF 1 OBRS 2 SF 1 PEO 1 7 Table 5.30. Kalleh Nisar style distribution.

3

2

1

0 JNRS STF SF ARS OBRS PEO

Graph 5.28. Kalleh Nisar style distribution.

Table 5.31 and Graph 5.29 illustrate the seal material proportions of the Kalleh Nisar corpus. This distribution patterns concurs with the expected pattern for the stylistic distribution as suggested by the styles represented at Kalleh Nisar, including the fact that the three hematite seals (2710, 2711, 2774) belong to Sukkalmah (Old Babylonian) styles (OBRS and PEO).

Glyptic Materials Limestone 1 Hematite 3 Alabaster/gypsum 1 Serpentine 2 Table 5.34. Kalleh Nisar glyptic materials.

3

2

1

0 Limestone Alabaster/gypsum Hematite Serpentine

Graph 5.29. Kalleh Nisar glyptic materials.

ECS Corpus, Volume I, Part III 616 Table 5.35 demonstrates the context of the Kalleh Nisar seals. The only tomb with more than one seal (C12, with two items), is a collective tomb, and so this may be similarly explained as the multiple seal tombs of Bani Surmah. Also like Bani Surmah, and as already mentioned, all the Kalleh Nisar seals come from a grave/tomb context, and so can be considered ‘funerary’ seals in function.

Glyptic Context A1 1 C12 2 A7 1 C13 1 A341D2 1 Table 5.35. Kalleh Nisar glyptic context.

5.12. Godin Tepe The ten Godin Tepe items (along with the ten from Tepe Giyan) provide the largest of the Luristan corpora (excepting the extraordinary Surkh Dum-i-Luri corpus). Table 5.36 and Graph 5.30 illustrate the style distribution of the Godin Tepe glyptic items. Despite the fact that Godin Tepe was occupied from the earliest periods of our study (Susa II) until the end of the Sukkalmah period, all the glyptic material from Godin Tepe dates to the first two phases generally, the Susa II (STS) and Susa III (JNRS, CPE, GS [with a possible extension into the following Susa IV period]) periods specifically. Indeed, as already mentioned, it has been claimed by the excavators that no glyptic material was found at Godin Tepe outside of the (purportedly Susa II) ‘Oval Enclosure’ (Young 1986: 217; see Chapter 2). However, one sealing at least (1365) was found in a Godin III context. This item is the sole GS style item in the Godin Tepe corpus, and is therefore slightly chronologically misplaced in the Godin III (later Susa IV) context in which it was found. This may be explained either as evidence for a continuation of this style into the later Susa IV period, or for the continued use of the GS seal that made this impression (that is, the seal’s re- [or continued] use), or the antiquity of the individual sealing itself. It should be noted that all these explanations are possible, and that the context of this item only marginally exceeds that which one would expect for a GS sealing.

Style Distribution STS 2 CPE 6 JNRS 1 GS 1 10 Table 5.36. Godin Tepe style distribution.

ECS Corpus, Volume I, Part III 617 6 5 4 3 2 1 0 STS JNRS CPE GS Graph 5.30. Godin Tepe style distribution.

The other Godin Tepe (the two STS, one JNRS and six CPE) items do accord with the above cited statement of Young regarding the restriction of glyptic material at Godin to the ‘Oval Enclosure’ (though it should be remembered that seal 999 was found within a mudbrick, in the ‘Oval Enclosure’ area, but in a Godin IV context). These items are only a selection of the total ‘Oval Enclosure’ Godin corpus, with the omission here of the not included stamp seals, and other items not previously published, and so cannot be necessarily considered an accurate sample or reflection of the glyptic reality. Furthermore, although it is striking that the vast majority of the Godin glyptic material came from this one area, the relative small exposure, and indeed the focus, at least in the published literature, on this sector as a field of study should be taken into account, and limits the overall accuracy and sample quality of the Godin material. The nine Godin Tepe items (aside from the lone GS sealing), were specifically published with the expressed purpose of illustrating the Susa II period date of the ‘Oval Enclosure’ (Weiss & Young 1975: 6 – 13). However, as is demonstrated in Table 5.36 and Graph 5.30, this purpose has in fact not adequately been achieved by this material. The two STS items (102, 442) do indeed concur with a Susa II date, and the single JNRS sealing (571) may be argued to date to a late Susa II phase, though less satisfactorily than a Susa III date (see Chapter 4 for details of the chronology of the JNRS style). However, the single largest stylistic group at Godin Tepe contains six items of clear CPE classification, which therefore must be given a Susa III, not Susa II date. Despite the use of these items to argue for a Susa II period date for the ‘Oval Enclosure’, Young acknowledges the fact that many of the Godin Tepe seals and sealings should be accorded a stylistic ‘Jemdet Nasr’ date (that is, Susa III) (Young 1986: 217). While this information is described as “significant”, no

ECS Corpus, Volume I, Part III 618 reassessment of the general Susa II (Uruk) date for the ‘Oval Enclosure’ has been proposed by the excavators (Young 1986: 217). The presence of not one, but several (and indeed, in terms of the nine ‘Oval Enclosure’ items [admittedly an extremely small sample], the fact that seven of the nine may be considered to have a post-Susa II date is indeed significant) glyptic items that belong stylistically to a later period (and therefore, cannot purely be dismissed as ‘heirloom’ or misplaced, re-used items as may be the case if the reverse was true [that these items belonged to an earlier period than proposed]) cannot be ignored however. While one is hesitant to redate the entire ‘Oval Enclosure’ phenomenon on the basis of glyptic stylistic information alone, the strong presence of Susa III period styles (both the CPE and JNRS) in this context must argue for some reassessment of the chronological span of this occupation. It may be hypothesised therefore, that at least part of the ‘Oval Enclosure’ intrusion should date to the ‘Proto-Elamite’ (Susa III) and not the ‘Uruk’ (Susa II) Expansion, on the evidence of the glyptic material. This proposal must await further information from Godin Tepe, and analysis of the other artefact materials from this area before further comment is made. Table 5.37 and Graph 5.31 illustrate the glyptic types (seals/sealings) of the Godin Tepe corpus. As is illustrated the majority of the items are sealings, with only two seals. Given the limited sample of the corpus, little discussion regarding the nature of this division may be articulated here.

Glyptic Type Seals 2 Sealings 8 Table 5.37. Godin Tepe glyptic types.

Seals

Sealings

Graph 5.31. Godin Tepe glyptic types.

ECS Corpus, Volume I, Part III 619 Table 5.38 and Graph 5.32 illustrate the glyptic material proportions of the Godin Tepe corpus. The modern, scientific approach to the publication of the Godin material is evidenced by this survey, as the seals are only listed in terms of stone appearance rather than a material designation attempted purely by macroscopic inspection, and the sealings are generally differentiated and detailed. As such it may be assumed that the single ‘undifferentiated’ sealing in this survey was a non- diagnostic item, that no amount of physical examination could discern its original type. The significant dominance of in this small corpus should be noted.

Glyptic Materials Grey-green stone 1 Jar stopper 1 Black stone 1 Jar sealing 1 Sealing 1 Sealed tablet 5 Table 5.38. Godin Tepe glyptic materials.

5

4

3

2

1

0 Grey-green Black stone Sealing Jar stopper Jar sealing Sealed tablet stone Graph 5.32. Godin Tepe glyptic materials.

Table 5.39 and Graph 5.33 illustrate the discernable glyptic function of the Godin Tepe material. The two seals may be demonstrated to have neither a funerary nor votive function, though any further information remains unknown, and are thus classified as ‘unknown function’ (the secondary context of item 999, which was found within a mudbrick, as already mentioned, should be noted). The five sealed tablets may be allocated a writing-based administrative function, though it is more correct to describe the ‘writing’ of this early (Susa II and Susa III periods) as proto-writing or proto-cuneiform rather than fully fledged true writing. The available evidence (for sealing 946 and 1107, the tablet type on the other items cannot be discerned from the published image) would seem to indicate that these tablets are numerical, rather than script based. The other sealing items in the Godin corpus were classified as evidence

ECS Corpus, Volume I, Part III 620 for a non-writing associated administrative function. The evidence thus provided for an administrative function for glyptic materials in Luristan should be noted.

Glyptic Function Non-writing administration 3 Writing administration 5 Unknown Function 2 Table 5.39. Godin Tepe glyptic function.

5

4

3

2

1

0 Non-writing administration Writing administration Unknown Function Graph 5.33. Godin Tepe glyptic functions.

5.13. Tepe Djamshidi The two seals from Tepe Djamshidi form the (equal with Chogha Gavaneh) smallest excavated site corpora in the entire Corpus (excluding the single, surface find seal from Deh-i Now). The two items are both seals, and are classified as AGD (1408) and OBRS (2712). The OBRS item is cut from hematite, and thus conforms to the general material type of this style, and implies a Sukkalmah (Old Babylonian) stylistic date. The AGD seal is made of terra cotta, and thus similarly concurs with a typical material of such items. The AGD item cannot be adequately dated, but a general ‘early’ period may be proposed on the basis of general comparison. The Tepe Djamshidi tombs (which provide the only information at the site) date from the middle Susa II period through to the end of the Sukkalmah period, and thus both the AGD and OBRS stylistic dates fit within this scheme. However, both the Djamshidi items were found in the one tomb (Tomb 3). This assures the general ‘funerary’ function (like those from Bani Surmah and Kalleh Nisar) of these items, though causes an interesting chronological conundrum. Unlike the graves from Bani Surmah, and most of those from Kalleh Nisar, the Tepe Djamshidi tombs are generally single inhumation burials, and therefore the chronological discrepancy between the AGD and OBRS items cannot be explained through multiple,

ECS Corpus, Volume I, Part III 621 generational, inhumations. It may be assumed, on the basis of the other Tomb 3 grave goods, the presence of the OBRS seal and the wider chronological use of Tepe Djamshidi that a general Sukkalmah (Old Babylonian) date can be proposed for this context. Two possible explanations, each equally probable and related one to another, may be postulated to explicate the appearance of the earlier AGD item in this context. The AGD item may either be wrongly allocated an AGD classification, and more correctly belong to a later chronological period, or the item may not be a seal at all, but rather a cylindrical bead with a geometric design, and thus any chronological placement ill-founded. 1408 was primarily classified as AGD on the basis of its similarity with several items from Susa, most notably 1402 and 1406 (the latter of which is a sealing, and thus indicates that this particular item at least was a seal, and not a bead). The line between bead and seal is notoriously difficult and blurred however, as is the allocation of chronological date on stylistic grounds for geometric designs. Thus each possibility is likely, though neither can be assumed. A third possibility is that the AGD item is a true ‘early’ period seal and was deposited in the Tomb 3 grave as an ‘heirloom’ in the Sukkalmah period. Any of these explanations is plausible, but none is assured or favoured, and little more time need be spent debating possible explanations for this isolated item of little consequence. The fact that the other Tepe Djamshidi item belongs to a ‘Mesopotamian Related’ style (OBRS) should be noted, and may imply a Mesopotamian (rather than Elamite) oriented constitution of the site. Given the extremely limited evidence of this one seal however, little weight should be taken from this.

5.14. Tepe Giyan As already noted, the ten Tepe Giyan seals (for all the glyptic material items from Giyan are seals) form the equal (with Godin Tepe) largest Luristan corpus, excluding the extraordinary Surkh Dum-i-Luri; though in the terms of the wider Corpus, this is a smaller site corpus. Table 5.40 and Graph 5.34 illustrate the style distribution of the Tepe Giyan glyptic items. Tepe Giyan was occupied from the Susa III period through to the middle of the Middle Elamite period (the end of the ME III phase), and thus the stylistic dates of these items conform generally to this pattern. Four Giyan items may be given a Susa III date (two JNRS items, and one item each of CPE and GS style). Another two items (the two AGD seals) also can be given an ‘early’ date. It is striking that six of the ten Giyan seals therefore belong to the earliest phase (or phases) of

ECS Corpus, Volume I, Part III 622 Giyan occupation. Given the general fragmentary and unclear state of the Giyan remains little functional or proportional conclusions may be drawn from this. The fact that ‘Mesopotamian Related’ (the two JNRS seals), ‘Elamite’ (the CPE item) and ‘Shared Mesopotamian-Elamite’ (the GS seal) classified styles all occur at Tepe Giyan indicates that this particular Luristan site was influenced both by southern, true Elamite styles, and western, Mesopotamian, forces. The two ‘Miscellaneous’ items are ‘Mittanian’ seals, and thus indicate a Tepe Giyan – Mittanian axis around the middle of the second millennium BC. The two ‘Unclassifiable’ items may belong to a specific Tepe Giyan school or style, but as there are only two such items, this cannot be sufficiently evidenced.

Style Distribution JNRS 2 AGD 2 CPE 1 Miscellaneous 2 GS 1 Unclassifiable 2 10 Table 5.40. Tepe Giyan style distribution.

2

1

0 JNRS CPE GS AGD Miscellaneous Unclassifiable

Graph 5.34. Tepe Giyan style distribution.

As already mentioned, all the glyptic items from Tepe Giyan are seals. Table 5.41 and Graph 5.35 illustrate the material types of this corpus. Little comment can be made regarding this survey, especially in light of the fact that most of the materials are classified by appearance only. The presence of and items should be noted however.

Glyptic Materials Serpentine 1 Soft stone 1 Grey stone 4 Soft yellow stone 1 Terra cotta 1 Dark grey stone 1 Faience 1 Table 5.41. Tepe Giyan glyptic materials.

ECS Corpus, Volume I, Part III 623 4

3

2

1

0 Serpentine Grey stone Dark grey Soft stone Soft yellow Terra cotta Faience stone stone

Graph 5.35. Tepe Giyan glyptic materials.

Three seals in the Giyan corpus were found in tombs (1025 in Tomb 52, 3486 in Tomb 68 and 3569 in Tomb 102), and thus may be allocated a funerary function. No contextual information is known for the other Giyan items, and thus these are classified as ‘unknown function’ in Table 5.42 and Graph 5.36.

Glyptic Function Funerary 3 Unknown Function 7 Table 5.42. Tepe Giyan glyptic function.

7 6 5 4 3 2 1 0 Funerary Unknown Function

Graph 5.36. Tepe Giyan glyptic function.

5.15 Chogha Gavaneh Two items only were provenanced from Chogha Gavaneh, thus comprising the equal smallest of the excavated corpora (equal with Tepe Djamshidi, excluding the surface find of Deh-i Now). One seal (of unknown material) and five tablets sealed by the same seal, both of OBRS (sub-group 2) type comprise this corpus. The seal, and the five tablets impressed by one seal, were all provenanced from the so-called Old Babylonian ‘archive’ of Room B15, thus conforming with the general Sukkalmah (Old Babylonian) stylistic date of the items. As the seal was found in the ‘archive room’, in close association with the sealed tablets (and the other, non-sealed tablets,

ECS Corpus, Volume I, Part III 624 see Chapter 2 and Abdi & Beckman 2007), a general administrative, and specifically writing-associated administration, function may be assumed for this item (and, obviously, the sealed tablets). This is the only seal in the Corpus for which such a function appears confirmed by context. The fact that both these items belong to the OBRS ‘Related Mesopotamian’ style, and are associated with Old Babylonian texts and tablets may imply that Chogha Gavaneh was more within the Mesopotamian, rather than Elamite, sphere of influence. The limited glyptic evidence, and indeed the general fragmentary and disjointed reconstruction of ancient Chogha Gavaneh prevent any real certainty in this regard however.

5.15. Tal-i Malyan The sixty-three items from Tal-i Malyan comprises a middle range corpus, that is significantly less than the larger corpora such as Susa, Chogha Mish, Haft Tepe, Choga Zanbil and Surkh Dum-i-Luri, though significantly larger than the smaller site corpora (Tepe Sharafabad, Deh-i Now, Kamtarlan, Chigha Sabz, Bani Surmah, Kalleh Nisar, Godin Tepe, Tepe Djamshidi, Tepe Giyan and Chogha Gavaneh). Table 5.43 and Graph 5.37 illustrate the style distribution of the Tal-i Malyan corpus. Table 5.44 illustrates the style distribution of the Tal-i Malyan corpus according to the specific sectors/operations of Tal-i Malyan. The pattern of distribution conforms generally with the chronological pattern of occupation at the site.

Style Distribution STS 2 PEA 4 JNRS 1 PEU 7 CPE 23 LME 1 GS 7 AS 10 LPS 2 AGD 5 Unclassifiable 1 63 Table 5.43. Tal-i Malyan style distribution.

ECS Corpus, Volume I, Part III 625 24 20 16 12 8 4 0

U GS AS STS CPE GD E LPS able JNRS A PEA P LME i assif ncl U Graph 5.37. Tal-i Malyan style distribution.

Style Distribution by Operations (Malyan sectors) Style ABC TUV EDD Unprovenanced STS – – 2 – JNRS – 1 – – CPE 16 5 2 – GS – 7 – – AGD 1 4 – – PEA 4 – – – PEU 5 – – 2 LME – – 1 – AS – – 10 – LPS – – 2 – Unclass. – – – 1 Table 5.44. Tal-i Malyan style distribution by Operations (Malyan sectors).

The two STS items have a stylistic date in the Susa II period, and thus slightly pre-date the initial period of occupation at Malyan. As demonstrated by Table 5.44, both these items were provenanced from EDD, an area of Tal-i Malyan that was apparently not occupied in the Susa II or Susa III periods, and so these may be considered ‘heirloom’ seals. The fact that these items are both sealings indicates that ‘heirloom’ seals were not merely kept as keepsakes or votive/funerary items, but were also used as administrative seals. The single JNRS item (from Operation TUV), the twenty-three CPE items (sixteen from ABC, five from TUV and two from EDD) and seven GS items (from Operation TUV) combine to produce thirty-one Susa III period items. Most of these items were found in Operations ABC and TUV each of which have Banesh (and therefore Susa III) occupation, and thus the occupational date and the stylistic dates generally concur. Like the two STS Operation EDD items, the two CPE items from this Operation may be considered ‘heirloom’ or ancient items. These two items were also sealings, again indicating that such ancient seals were still used

ECS Corpus, Volume I, Part III 626 as administrative devices. It should be noted that three of these ill-fitting EDD items (362, 948, 1009) come from an earlier (than the main Middle Elamite/Qaleh occupation) level (EDD Building Level IIA), and thus may not be so removed from the earlier date of these items (see Chapter 2 for details of the EDD occupation). The five AGD items were found in the two early occupation Operations (one from ABC, four from TUV), in accordance with the general ‘early’ (in terms of this study) chronological designation. The four PEA items, and five PEU items were excavated in Kaftari occupations of ABC (two PEU items have no known provenance), conforming with the general chronological distribution of these items. The single LME item, ten AS and single LPS items all have a late Middle Elamite (post ME I) date, and were all provenanced from the EDD Qaleh (Middle Elamite) occupation, in accordance with the stylistic date of these items. Table 5.45 and Graph 5.38 illustrate the glyptic types of the Malyan corpus. As is illustrated, the majority of the items were sealings (over 71%). This demonstrates both the general glyptic functional difference of Malyan (as opposed to that at Choga Zanbil and Surkh Dum [votive], and Bani Surmah, Kalleh Nisar and Tepe Djamshidi [funerary]), and the careful, scientific recent excavations at Tal-i Malyan (meaning that sealings were recognised, and recorded, in greater, more accurate numbers).

Glyptic Type Seals 18 Sealings 45 Table 5.45. Tal-i Malyan glyptic types.

Seals

Sealings

Graph 5.38. Tal-i Malyan glyptic types.

Table 5.46 and Graph 5.39 illustrate the glyptic materials of the Tal-i Malyan corpus. The modern approach undertaken by the Malyan excavators and publishers is

ECS Corpus, Volume I, Part III 627 also demonstrated in this survey, as the seals are generally identified by appearance only (except for the three items), and the sealings have been described in greater detail. The fact that and are the two most numerous groups in the Malyan corpus is however attributable to the fragmentary publication of most of the Malyan Operations (GHI, FX106, GGX98, BY8, see Chapter 2 for details). Some of the ‘sealings’ may be non-diagnostic items however. The survey also demonstrates the need for a thorough physical and petrographic study of this material.

Glyptic Materials White stone or plaster cylinder 1 Clay tag sealing 5 Soft white stone 1 Sealing(s) 26 Claystone or ceramic 1 Jar sealing(s) 6 Faience 3 White plaster sealing 1 Door lock sealing 4 Unknown cylinder 12 Sealed tablet 3 Table 5.46. Tal-i Malyan glyptic materials.

26 24 22 20 18 16 14 12 10 8 6 4 2 0

r e c ) g et on ing (s laste ami inder al aling er l ing e ite st Faience ed tabl or p ag se Seal or c t lock s one Jar sealing(s)laster sealin Seal oft wh one p st S st Clay ite Unknown cy hite Door Wh Clay W

Graph 5.39. Tal-i Malyan glyptic materials.

Table 5.47 and Graph 5.40 illustrate the glyptic function of the Tal-i Malyan corpus. The are the only items associated with writing in an

ECS Corpus, Volume I, Part III 628 administrative system. All the other sealings were grouped in the non-writing administrative group. A single seal, of the LPS style (3277) was found in a tomb in Operation EDD and so can be considered to have a ‘funerary’ function. Apart from this item, no function can be discerned for the rest of the Malyan seals from contextual evidence.

Glyptic Function Non-writing administration 42 Writing administration 3 Funerary 1 Unknown Function 17 Table 5.47. Tal-i Malyan glyptic function.

45 40 35 30 25 20 15 10 5 0 Non-writing Writing administration Funerary Unknown Function administration Graph 5.40. Tal-i Malyan glyptic function.

The articulation and summation of the Elamite Cylinder Seal Corpus Styles contained in the previous two chapters (Chapters 4 and 5), according to chronological development (Chapter 4) and geographical distribution (Chapter 5) concludes the data presentation. The following chapters will contain two discussions regarding particular aspects of the material here presented. The first of these discussions, presented in Chapter 6, regards the functional use of Elamite Cylinder Seals, as discerned from the evidence presented here. The second discussion will detail the constitution of ancient Elam, with specific reference to previous reconstructive studies, and Mesopotamian interactions, as may be evidenced through the glyptic data presented here.

ECS Corpus, Volume I, Part III 629 Chapter 6 – Glyptic Function in Elam The two distinct (though interrelated) glyptic function divisions (the administrative and the symbolic) have already been discussed and outlined in the course of the preceding study (Chapters 1 and 2, Chapter 4 and 5). As already discussed, both the administrative and the symbolic functions may be divided into two different types; the administrative into that element not associated with writing and that that is, and the symbolic into the ‘votive’ and ‘funerary’ functions. Each of these four divisions will be discussed in detail below, as will a fifth functional classification, that contains two miscellaneous groups (the ‘Palatial’ provenance seals from Choga Zanbil and the non- votive hoard seals from Chigha Sabz), as well as the major function, which is, unfortunately, simply labelled ‘unknown’. The reasons for this large elusive classification, and the methods of function allocation in the other cases will be detailed. This Chapter will discuss and define Elamite glyptic function in terms of the chronological, stylistic and archaeological glyptic paradigm outlined above. Thus, this survey is intended as a discussion of glyptic function within ‘Elam’ (or more correctly, the area under discussion here and defined broadly as ‘Elam’) alone, and within the stated chronological parameters. No functional inference beyond these geographical or chronological limits should be taken therefore. As the stylistic and chronological development of the Elamite glyptic paradigm has now been established (Chapters 4 and 5), this discussion may be faithfully and accurately undertaken. Each glyptic function will thus be discussed in terms of its stylistic, chronological and geographic distribution, with any functional or interpretative implications of this distribution addressed. A final note regarding one function, already discussed above (Chapter 1), but not represented in the following survey, should be made. This function is the ‘amuletic’ function of seals. As already detailed, seals were known to have been used (at least as can be gleaned from Mesopotamian sources for this region) as amulets, or protective charms or devices used to ward off evil or guard the bearer. This function can be discerned both from texts that prescribe the use of seals in spells and rituals, and by the prayers and benedictions contained within the seal inscriptions themselves (Gelb 1997: 112; Hallo 1977: 58; Black & Green 1992: 300; Matthews 1990: 11; Collon 2005: 62; 1997: 20; van Buren 1954: 33, 119, 131). The material of a seal, especially when made out of a stone or mineral imbued with perceived ‘magical’

ECS Corpus, Volume I, Part III 630 powers (such as the well known lapis lazuli) also adds to the amuletic function of seal (Collon 2005: 62, 100; 1997: 19), a facet of a seal’s function with is linked to beads (Collon 1997: 19; Moorey 1994: 103; Gorelick & Gwinnett 1981; 1990). The ‘amuletic’ function of a seal is also derived from the function of a seal as an identifier of the individual (and is thus linked to the administrative role of seals discussed below). The order of implied function (that is, whether the amuletic function as individual identifier [or as imbued with the essence of the individual] led to the administrative function of seals, or whether the administrative function as identifier enabled the amuletic function to be perceived) is unknown and currently unknowable (on the basis of available information), though it seems most likely that each function and facet influenced the other to greater and lesser degrees across time and space. The definition of a seal as an identifier, and as an object imbued in some degree with the ‘persona’ of the owner(s) also enabled the votive function of seals, as will be detailed below. The amuletic function, more than any other, is difficult, if not impossible, to discern in the archaeological record, but is probably present, in greater and lesser degrees, for all seals. For these two reasons, the lack of an appropriate test or study with which these items may be identified, and as all seals most probably held this function to some degree, the ‘amuletic’ function of seals has not been identified or detailed in the following functional survey, though the continual presence of this function within the Corpus should be noted. The geometric items classified in the ‘early’ (Archaic, AGD) and ‘late’ (LGD) classifications may more appropriately be described as amuletic devices than true seals. These items highlight the fact that seals and amuletic beads belong along a continuum of similar, related objects. The geometric decorated cylindrical devices that were carried as items of personal decoration and as protective amulets may be better described as beads, while ‘seal’ may apply to those similarly constructed and formed items that were (occasionally or regularly) employed to actually seal a device. When such items became seals or beads, and the frequency or otherwise of their use as such, makes a division into a dichotomy difficult, if not unattainable. For this reason the geometric designs have been classified separately from the other stylistically classified items. The fact, as already demonstrated, that at least some of these geometric design items were used as true seals (that is, there are impressions of

ECS Corpus, Volume I, Part III 631 these items, see Chapter 4, and Graph 5.1) indicates that not all the geometric items may be considered amuletic beads. Table and Graph 6.1 demonstrate the functional distribution of the geometric items (the AGD and LGD classifications, see below for further details of these functions and their divisions). These illustrate that geometric seals were used as administrative devices (sealings), in significant numbers, but were also employed as symbolic devices (funerary and votive seals). The vast majority of these items are classified however, and any number of these items may have been beads rather than true seals. The significant number of geometric design items that may be classified as votive devices (thirty-four of the thirty-eight ‘symbolic’ items) should be noted, and is also probably linked to the amuletic, bead-like aspect of these items. Thus, the ‘amuletic’ function of seals is linked to beads that also had a similar function, they are linked along a continuum of type therefore. As one cannot place a break along a continuum, the geometric items have been classified separately from other items, as discerning whether these items were true seals, with a bead like aspect, or beads that were occasionally employed as seals cannot be discerned from the current position. Similarly, as ‘amuletic’ function may be assumed for all of the items in the Corpus (to greater and lesser degrees), and indeed, as this facet influenced and/or enabled other functions, no attempt at discerning amuletic and non-amuletic devices in the Corpus has been undertaken. It may be assumed that some of the geometric design items classified in AGD and LGD were in fact amuletic beads rather than seals, though as these two types are linked along the continuum, this division is essentially one of semantics and emphasis, and so all such cylindrical shaped items that bear a geometric design, that may potentially have been employed as a cylinder seal, have been included here. Due to this functional discussion regarding the geometric items, and the linked difficulty in discerning chronological placement for these items, the AGD and LGD classified glyptic material will not be included in the stylistic and chronological discussions of the functional distribution patterns, though their number will be included in the total for each type (thus the ten AGD sealed bullae will be counted amongst the bullae total below). Table and Graph 6.1 present the functional types of these items for reference and the sake of completeness.

ECS Corpus, Volume I, Part III 632 ‘Geometric Design’ Functional Types Sealed bulla(e) 10 Funerary 4 Sealing(s) 52 Palace 4 Sealed tablet 3 Non-votive Hoard 2 Votive 34 Unknown Function 264 Table 6.1. Functional classifications of the ‘Geometric Design’ (AGD and LGD) items.

270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

n e e n o iv ard o ot alac V unerary P uncti F F tive Ho ministrati ministration o wn d d o g a on-v in ing a N nkn rit U W on-writ N

Graph 6.1. Functional classification of the ‘Geometric Design’ (AGD and LGD) items.

6.1 Administrative The function defined as ‘administrative’ may also be labelled the traditional or standard glyptic function, or indeed, the definitional function of a true ‘seal’. The cylindrical shaped devices known as cylinder seals were used, in this study, to seal

ECS Corpus, Volume I, Part III 633 clay tablets and envelopes, and pieces of clay affixed to vessels (ceramic containers, baskets and bales) and doors (and walls). All these functions may be linked to the control of a society, and therefore may be described as ‘administrative’. This need not infer a written or bureaucratic aspect however, as will become apparent in this survey. This action of ‘sealing’ a device provides the true definition of a seal, and its appellation. In this regard, the control or administrative function of a seal may be described as the true function of a seal. The relationship between this function, and the symbolic function discussed below will be returned to. The ‘administrative’ function of seals may be classified into two separate groups, that associated with writing, and that which is not. Both divisions are involved in the regulation or control of a society, and thus warrant the appellation ‘administrative’. The precise function and method of sealing and use varies between these classifications, and within them, and will be detailed below.

6.1.1 Non-writing administration Quite obviously, and as the name would suggest, ‘non-writing administration’ entails sealing practices and actions that are not associated with writing or a script. It should be noted that this classification does not, necessarily, imply the absence of writing within the wider cultural context, nor a writing-associated administration, but is merely a classificatory difference. Whether associated with writing or not, the action of sealing involves a shared method and type of information processing, transference or recording, though between these two types (writing and non-writing associated) and within these classifications, different pieces of information may be recorded or transmitted.

6.1.1.1 Non-writing associated sealings Several non-writing associated sealing types can be discerned in the Elamite corpus (other types, known from other regions and periods need not concern us here). These include the impression of a piece of clay placed over a rim or mouth of a jar or vessel (variously characterised as , or , depending on the placement of the clay, and the nature of the vessel [an open or closed form]), or a piece of clay used to lock a door () or seal a wall ( or sealings), or pieces of clay that marked other items or containers of objects, such as bales (), baskets (),

ECS Corpus, Volume I, Part III 634 and the clay tags, that were presumably used to mark such items (). Each of these items may be discerned through physical examination, generally of the reverse of the sealing, by criteria such as shape (the shape of the item upon or against which a piece of clay was impressed will reflect the shape and form of the sealings, thus a flat, roughly rectangular sealing may be assumed to have been impressed against a flat, steady surface, and thus be described as a wall sealing, while a sealing which may appear to mirror the rim-shape of a vessel, may be assumed to have been impressed against such an item) and markings (the presence of wood grain remains or impression on the reverse indicates that the sealing was pressed against a wooden object, possibly the bolt of a door lock, or a wall or door jamb, similarly, the traces of rope or twine markings indicates that the clay was pressed against rope, which may have itself formed a part of a jar, bale or jar sealing; the presence of plaster traces on the reverse may indicate that the impression was placed against a plastered wall, and thus be a wall lock sealing, and traces of matting, basketry or weaving may indicate that the item sealed a basket or similar container; for further details regarding the criteria for sealing type identification see [Ferioli & Fiandra 1979; 1983; 1994; Fiandra 1979; Zettler 1987]). As only the extremely limited and fragmentary ‘Tehran Sealings’ (of which only tablets were visually identified due to the non-diagnostic nature of the items) were the only sealings in this study to undergo physical examination by the author (as detailed above, Chapter 3), a thorough discussion of identification method is unwarranted here. For the most part, any description of specific sealing type in the Corpus has relied upon identification made in the original publications (that is, by those who were able to make a physical study of the material). As already detailed in full (Chapters 2 and 5), the various ages and approaches of these publications led to variation in the quality and execution of such classification analyses. In many cases, no (or extremely limited) attempt at sealing type identification was made, and thus no such information was recorded here. Also as already outlined, while for writing associated sealings (specifically tablets and clay envelopes) and the functional class (further outlined immediately below), non-writing sealing types (jar sealings, door lock sealings etc.) cannot generally be identified from image alone, but require a physical examination. This is due to the fact that, as outlined above, the (three-dimensional) shape and the remains of the sealings reverse (a view rarely published in photographs and line drawings) are the factors that hold the key in such identifications, but are generally undiscernible from published

ECS Corpus, Volume I, Part III 635 images. Thus, due to the fact that the vast majority of the items in this study were not appropriately published so as to discern sealing type from image, and most of these publications were undertaken before the practice of sealing type identification was common, the overwhelmingly majority of non-writing associated pieces are listed simply as (as demonstrated by Graph 6.2, where only one hundred and six of a total nine hundred and fifty-three non-administration sealings in the Corpus have been classified by type, of these most [fifty-eight], were classified as jar sealings).

Classified

Unclassified

Graph 6.2. Division of non-writing associated sealings according to classification.

It should be remembered that a significant, un-specified and unknown, number of these items may be non-diagnostic sealings. That is, sealings that are too fragmentary, partial, damaged or ill-preserved to enable type classification, regardless of physical analysis or not. Until a thorough, complete, physical survey of all the sealings in the Corpus has been completed, the proportion of non-diagnostic pieces will remain unknown however. As already mentioned, and detailed below, other sealing types, specifically bullae and writing-associated items (tablets and sealed envelopes) have generally been discerned in even the oldest, less-thorough publications. Furthermore, even in the rare cases that such items are not so listed, bullae, tablets and envelopes are easily and readily recognisable from photographs and images. It is therefore concluded that all (or virtually all) bullae, tablets and envelopes in the Corpus have been discerned, recognised and noted as such. It therefore must hold true that all the undifferentiated sealings are either various non-writing associated sealing types (such as jar sealings, door lock sealings and so on), or non-diagnostic pieces that were originally such types (the possibility that extremely fragmentary bullae, tablets and envelopes would occur amongst the non-diagnostic pieces is acknowledged, though it is anticipated that this number would be a relatively small amount, as even in extremely fragmentary cases,

ECS Corpus, Volume I, Part III 636 the distinctive form and shape of these items would be partially preserved). Thus, the relatively few items that have been allocated a non-writing administrative type in the Corpus, are joined with the undifferentiated sealings, to form a basic, non-writing associated administration group. Due to the general fragmentary and disjointed nature of this corpus (and the minority portion that contains any type classification), few suggestions regarding type proportions (how many jar sealings relative to bale sealings and so on) and thus glyptic function may be proposed. The significant number of such items (only the equally unsatisfying classification contains more single examples than the non-writing associated administrative group, as will be demonstrated below) does however indicate that seals were regularly, and routinely used, across ‘Elam’ (as defined here) to seal non-writing administrative items. Table 6.2 and Graph 6.3 illustrate the distribution of non-writing associated administrative sealings (excluding , see below for details of this sub-type) across the sites of the Corpus. As is indicated, all the ‘Elamite’ sites that produced sealings produced this classification of non-writing associated sealings (thus the only sites not represented are those that produced seals only), except for Chogha Gavaneh, the extremely limited corpus of which only contributed one sealing item to the Corpus, a sealed tablet (and thus an ‘associated with writing’ sealing). The relative proportions across the sites is of little meaningful comparison, as the site corpora differed greatly in size and type (for example, the massive number of glyptic items uncovered at Susa in the century long major excavations of the site, compared with the limited exposure of mere weeks at Kamtarlan), and thus little can be taken from this.

Non-writing associated sealings, site distribution Susa 646 Tepe Sharafabad 3 Chogha Mish 81 Kamtarlan 1 Godin Tepe 3 Haft Tepe 177 Tal-i Malyan 42 Table 6.2. Site distribution of the non-writing associated sealings (excluding bullae).

ECS Corpus, Volume I, Part III 637 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 Susa Chogha Haft Tepe Tepe Kamtarlan Godin Tal-i Mish Sharafabad Tepe Malyan

Graph 6.3. Site distribution of the non-writing associated sealings (excluding bullae).

ECS Corpus, Volume I, Part III 638 Table 6.3 and Graph 6.4 illustrate the stylistic and classification distribution of the non-writing associated sealings. In accordance with the above statement, this classification does indeed occur across all the styles of the paradigm here established, with only the ‘No Image’ classification absent from the survey presented in the table and graph. The significant proportion of STS sealings may have a functional cause, and will be further addressed below. The significant number of EME undifferentiated sealings is linked to the primary source of this style, Haft Tepe (that produced a great number of sealing items, poorly published [and thus undifferentiated in the main]), and may have a functional cause (that is, the possible craft production function of the area excavated at Haft Tepe [see Chapter 2] may have increased the number of non- writing associated sealings in this area, as attached to ‘shipments’ of raw materials in sealed vessels and containers).

Non-writing associated sealings, style distribution STS 253 OBRS 20 JNRS 30 PEO 3 CPE 92 EME 176 GS 63 KRS 13 AGD 50 LME 19 STF 125 AS 8 SF 13 LPS 2 LSF 2 LGD 2 ARS 8 Miscellaneous 2 PEA 1 Unclassifiable 18 UTRS 36 Not Illustrated 6 PEU 11 Table 6.3. Style distribution of the non-writing associated sealings (excluding bullae).

ECS Corpus, Volume I, Part III 639 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

S S F E E S us TS R PE G GD TF S SF RS EA EU EO M RS M A PS GD o ble S C S L A P TRS P BRS P K L L a JN A U O E L

iscellanenclassifi M U Graph 6.4. Style distribution of the non-writing associated sealings (excluding bullae).

Finally, Table 6.4 and Graph 6.5 demonstrate the chronological division of the classification. As outlined above, the geometric design items (AGD and LGD) for which it is notoriously difficult to ascertain an accurate date, have not been included in this (nor any of the other chronological) survey. Following the stylistic paradigm established here, the STS style is deemed to belong to the Susa II period, the JNRS, CPE and GS styles are listed as Susa III; the STF style is placed in a joint Susa III/IV period (for the purposes of comparison here only); the SF and LSF styles date to the Susa IV period; the ARS and PEA represent the ‘Akkadian and Awan’ period; similarly the UTRS and PEU date to the ‘Ur III and Shimashki’ period, and the OBRS

ECS Corpus, Volume I, Part III 640 and PEO the Sukkalmah; the EME style is placed in a separate early Middle Elamite phase, though the extension of this style into the late Sukkalmah period should be recalled; finally, the KRS, LME, AS and LPS are all subsumed into the late Middle Elamite period, though it should also be recalled that some of the LPS items more correctly date to the first millennium BC. This same chronological paradigm will be employed in all the functional category discussions to follow.

Non-writing associated sealings, chronological distribution Susa II 253 Akkadian & Awan 9 Susa III 185 Ur III & Shimashki 47 Susa III/IV 125 Sukkalmah 23 Early Middle Elamite 176 Susa IV 15 Late Middle Elamite 42 Table 6.4. Chronological distribution of the non-writing associated sealings.

260 240 220 200 180 160 140 120 100 80 60 40 20 0

I n I IV ite te sa I/ shki m mi u la S Susa III E Susa IV & Awa Susa II Sukkalmah le dd adian i k III & Shima Middle Ela M y te Ak Ur a Earl L Graph 6.5. Chronological distribution of the non-writing associated sealings.

The significant number of non-writing associated sealings in the earlier periods (Susa II and III) should be noted, and will be further discussed below in relation to the other functional proportions from this period. Like the stylistic contribution of the EME style, the early Middle Elamite period preponderance for such a functional class is linked to the exceptional and significant contribution of Haft

ECS Corpus, Volume I, Part III 641 Tepe. The extent to which Haft Tepe, and the area of that site that underwent excavation and thus yielded this material, can be considered typical of an Elamite site in this period is debatable (especially in light of the fact that there is no excavated contemporary site with which to compare and test the Haft Tepe results [the site of Susa was occupied in this period, though the extent of this occupation is currently unclear, and the differentiation of the actual early Middle Elamite occupation at Susa is difficult, and the material fragmentary, making such a correlation untenable]), and must impact on the extent to which the early Middle Elamite break with the general trend (of reduced non-writing administration usage) may be accepted as an accurate reflection of the ancient reality, or an anomaly caused by the Haft Tepe excavations. Thus, it may be said that ‘Elamite’ cylinder seals were used across the ‘Elamite’ realm (and across the three ‘provinces’, Khuzistan, Luristan and Fars), and throughout the chronological span of this study. The fact that a significant number of the sites discussed in this study did not produce non-writing associated sealings (or indeed, as will be demonstrated, any sealing material) is an interesting functional phenomenon, that will be addressed in the concluding remarks of this survey. The fact that the majority of the sealings classified in this division are simple undifferentiated and unclassified items is lamentable, and prevents any further functional definition or discussion here. Only the most cursory of functional conclusions may be drawn. The classification of these items as separate from a writing-based administration is significant, and implies that much of the sealing practice in Elam was not concerned with the ‘signing’ of tablets in lieu of the modern signature, as is often most simply assumed. The sealing of jars, containers and items that may be described as ‘mobile storage units’, may have had a different functional application than the items which sealed the ‘immobile storage units’ (that is, the door and wall lock sealings that presumably ‘sealed’ rooms and areas of buildings). The placement of a piece of clay on a jar or other mobile storage vessel (including baskets and bales) presumably marked the goods in some manner. The information transferred by this marking is open to debate and cannot really be ascertained currently; it may describe or label the goods themselves, indicate from whence (or by whom) the goods were packed or sent, or to the eventual destination (or receiving individual) of the items; the sealing may have authorised or authenticated the items so sealed; or the sealing may have acted as a bookkeeping device, a point of reference or reminder of one or more of the just cited pieces of information. All these possibilities are plausible, but

ECS Corpus, Volume I, Part III 642 none can currently be favoured or proven. It is indeed possible that some or all of these functions operated separately or concurrently over the period of glyptic use, especially in light of the significant (two and half thousand years) chronological span of this study (indeed, one can not anticipate that bookkeeping and control mechanisms of a modern, digital-era stock exchange to be the same as a Medieval British market, two points in time closer to one another than the Susa II period is to the Middle Elamite). The act of sealing an immobile storage device (a room), through a wall, door lock or wall sealing may have a related, though different function (or information transference mechanism). The two types are themselves differentiated, the door lock seal appears to have been impressed on a device that actually closes and secured a room, in the manner of a lock1 (the ‘wall lock’ sealing is in some instances a synonymous term for a door lock sealing, but in other cases refers to a piece of clay placed over a closed door so that part of the clay covers the door, part the wall or door jamb, which was then sealed, this is essentially the same as a door lock sealing, without the presence of an actual lock device, with the same result, that in order to open the door the seal must be broken, while the wall sealing may not have necessary locked or secured a room, but merely be a piece of clay placed on a wall and impressed with a seal, possibly as a sort of label or record; the ‘clay tag sealings’ of Tal-i Malyan may have functioned in this way [Sumner 2003], or have been similarly impressed as a label on a mobile storage device). The information transferred from the door lock sealing may have been an identification of the individual who did the sealing, and was thus the last person to ‘lock’ the room, or to have removed or added items to the ‘store’; or the information transferred may have been date or time, rather than individual involved; or the door lock, like the wall lock, could have acted as a label and indication of what was in the room. Like the mobile storage devices, the general fragmentary nature of this corpus prevents any further analysis, or favoured conclusions to be advocated.

1See Johnston et al. 2001 for a discussion of whether such sealings were safe from tampering, it should be noted here that the security or efficacy of these devices in keeping out unwanted individuals is not of importance here, the fact that elements of a society kept items in a room that was sealed with a device or marker is.

ECS Corpus, Volume I, Part III 643 6.1.1.2 Bullae Bullae are a sealing type with a distinct and obvious form. A bulla is a ball of clay, generally hollow and filled with tokens, the outside of which is sealed with one, or more seals, that generally cover most, if not all of the surface area. Bullae are here classified as non-writing administrative devices, following the commanding division, as these items are indeed not impressed upon tablets or envelopes that bear cuneiform (or proto-cuneiform) text. However, the proposed causal and developmental link between bullae and tokens and the development of numerical tablets and eventually writing should be noted (see Damerow & Englund 1989; Friberg 1978; 1979; Le Brun & Vallat1978; Matthews 1993; Nissen 1986; Schmandt-Besserat 1978; 1979; 1981; 1986; 1994; Vallat 1986), and therefore links these items with the earlier manifestation of writing administration discussed below (indeed, in this regard, the bullae and numerical tablets may belong along a continuum, with the break here being the presence or absence of writing). Presumably, the clay tokens held some value, either as a mnemonic device for record-keeping purposes (that is, a reminder to the individual that the three tokens there contained related to three units), or as a method of information transference (that is, the tokens that represent the three units may have been ‘read’ and understood to relate to the particular units by any individual familiar with the system), with the action of encasing these objects within a sealed container (for thus essentially is the bulla) presumably safekeeping, authorising or authenticating the information there contained (Collon 2005: 113). The relationship of these items with item transference (intra- or inter-site movement of goods and objects), and therefore an analogy with shipping receipts, may be proposed, but is currently unproven. It is not needed, nor entirely relevant, to detail the discussion and debate surrounding the function of bullae here, see the above volumes for the various arguments and proposals. The above cited function is the basic description adhered to here. Table 6.5 and Graph 6.6 illustrate the number of in the Corpus, according to site distribution. It is abundantly clear that the is a form limited only to Susa and Chogha Mish. Furthermore, Table 6.6 and Graph 6.7 illustrate the stylistic distribution of the ‘Elamite’ bullae, and Table 6.7 and Graph 6.8 the chronological distribution. As is clearly evidenced, the STS style contributes by far the most bullae, and by extension, bullae are an, almost uniquely, Susa II phenomenon (this indeed fits with the distribution of these items in other Susa II

ECS Corpus, Volume I, Part III 644 contemporary [that is, ‘Uruk Expansion’] sites beyond ‘Elam’ [see below for details]). Indeed, all the Chogha Mish bullae belong to the STS/Susa II classification, with all the other bullae bearing non-STS items provenanced from Susa. There actually appears to be two distinct groups of bullae (as may be discerned from the publication images, it must be remembered that a physical analysis of this material has not been undertaken). The ‘true’ bullae, those items that contained tokens, appear to be restricted to the Susa II period (STS style). The non-STS Susian ‘bullae’ do not contain tokens, and are thus not true ‘bullae’, but rather are balls or lumps of clay that bear a seal impression. The ‘true’ bullae are those that acted as a receipt or record of goods, and contained tokens, and thus functioned according to the above outlined practice. The other, non-STS, ‘bullae’ or clay balls may have sealed knots of rope tied around containers or vessels, and thus both be more synonymous with modern customs or postage bullae and linked with the jar and other sealings discussed above. The description of these items here as ‘bullae’ follows their designation in their original publication (for example Amiet 1972: 262), and is adhered to as a physical study was not undertaken, though it is hypothesised that this description is one of confused nomenclature, where the true, protoliterate, token-containing bullae, and the non-STS balls of clay were described by the same French term bulle, but not according to like function or form. Thus, one hundred and eighty-four items may be considered true, token-bearing bullae (forty-eight from Chogha Mish, one hundred and thirty-six from Susa), all impressed with STS style designs, and another seventy- seven sealings, all from Susa, appear to be balls or lumps of clay, that functioned essentially as sealings, and did not bear tokens, and will be labelled bulle in the following discussion.

Bullae/bulles Site Distribution Bullae Bulle Total Susa 136 77 213 Chogha Mish 48 – 48 261 Table 6.5. Site distribution of bullae/bulles.

ECS Corpus, Volume I, Part III 645 140

130

120

110

100

90

80

70 Bullae 60 Bulles 50

40

30

20

10

0 Susa Chogha Mish

Graph 6.6. Site distribution of bullae/bulles.

Bullae/bulles style distribution STS 184 SF 9 JNRS 10 ARS 5 CPE 8 UTRS 4 GS 11 PEU 1 AGD 10 OBRS 1 EME 4 STF 13 Unclassifiable1 Table 6.6. Style distribution of bullae/bulles; all the non-STS items are bulles.

Bullae chronological distribution Susa II 184 Akkadian & Awan 5 Susa III 29 Ur III & Shimashki 5 Susa III/IV 9 Sukkalmah 1 Susa IV 9 Early Middle Elamite 4 Table 6.7. Chronological distribution of bullae/bulles; all post-Susa II items are bulles.

ECS Corpus, Volume I, Part III 646 185 180 175 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

S S F U S le G SF RS b STS NR CPE ST A PE EME J AGD UTRS OBR assifia Uncl Graph 6.7. Style distribution of bullae/bulles, all the non-STS items are bulles.

ECS Corpus, Volume I, Part III 647 190

180

170

160

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0

II h III IV ite usa am S usa III/IV usa Awan l S S & imashki usa h ukkalma S ian S S d II & Middle E kka r I A U arly E Graph 6.8. Chronological distribution of bullae/bulles; all post-Susa II items are bulles.

It should be noted that the figure one hundred and eighty-four does not in fact refer to individual bullae, but glyptic items. That is, there are one hundred and eighty- four individual glyptic images within the Corpus that were impressed upon bullae. This conforms with the above outlined classification system, where image items were considered, not individual sealing units (thus, ‘duplicate’ sealings, that is, two or more sealings all impressed by the same seal were considered one unit, and conversely, sealings that bore two [or more] impressions were counted as per image, and thus two

ECS Corpus, Volume I, Part III 648 items [Catalogue entries] were recorded for the one object [bullae]). A particular facet of bullae use (and one that no doubt held a particular functional or methodological purpose) is the fact that it was common for one or more separate seals to be impressed upon the one bulla. Thus in the Corpus, twenty-eight bullae from Susa (and a single non-STS bulle, that bore image items 2801 and 2866, both of EME style), and ten from Chogha Mish bore the impressions of two, or more, separate seals. The thirty multiple sealing bullae are listed in Table 6.8. As is demonstrated, the majority of these items bore two separate sealing images (twenty-nine bullae), with another sub- set that bore three (eight items), and an extraordinary bulla from Chogha Mish that bore no less than six impressions made by separate seals. The number in parenthesis in the Table indicates the STS sub-style of the image. The fact that in very few instances does the image on a bullae belong to the same subject matter sub-group is interesting, and may have some functional cause (thus Chogha Mish bullae 1, 117, 118 all bore two STS [6] images, the six-seal bulla from Chogha Mish two STS [16] items, and Susa bullae 176, 370, 376 also bore two STS [16] images). A thorough analysis of associated image types (not undertaken here due to time and relevance constraints) may provide useful information in regards to image type and association. As an aside, the fact that no multiple sealings bullae cross stylistic borders (that is, all these items are STS style, regardless of image types) demonstrates the internal consistency and accuracy of the STS sub-groups as a single style. The precise reason(s) for multiple sealing by different seals is unknown, but no doubt is linked to the, still generally enigmatic, function of these items. If the bullae were objects of identification (that is, they marked, or labelled the goods or items with which they were associated), with the seal images contributing part of the identifying information (a seal of a specific image type may have meant ‘cattle’ and so on), then multiple sealings on the one bullae may have simply implied multiple goods or types of objects (that is, ‘cattle’ and ‘caprids’ for instance). If the bullae marked or recorded a transaction, and the seals acted as identifiers of individuals involved in this transaction (perhaps as a record or contract of goods ‘shipped’ and received), then the multiple sealings may have denoted multiple witnesses to, or participants in, the transaction or agreement. Again, a thorough, specific analysis of the seal images may aid in the resolution and reconstruction of this phenomenon.

ECS Corpus, Volume I, Part III 649 Site Multiple Sealings (STS Site Multiple Sealings (STS sub-style) sub-style) Susa 3 (1) 187 (8) Mish 116 162 196 (6) (7) (8) Susa 5 (1) 28 (2) Susa 125 179 (7) (8) Susa 8 (1) 29 (2) Susa 127 300 (7) (14) Mish 12 (1) 117 (6) 118 (6) Susa 144 373 (7) (16) Mish 13 (1) 285 Mish 161 352 395 396 423 477 (13) (7) (15) (16) (16) (17) (21) Mish 36 (2) 52 (3) 115 (6) Susa 174 212 (8) (10) Susa 39 (3) 313 Susa 176 370 376 (14) (8) (16) (16) Susa 40 (3) 61 (4) Susa 177 308 (8) (14) Susa 41 (3) 269 299 Susa 181 267 (13) (14) (8) (13) Susa 43 (3) 443 Susa 183 271 (20) (8) (13) Susa 53 (4) 292 Susa 184 277 (14) (8) (13) Susa 57 (4) 203 (9) Susa 189 201 314 (8) (9) (14) Susa 59 (4) 272 Mish 197 207 398 (13) (8) (9) (16) Susa 60 (4) stamp sealing Susa 266 444 not listed (13) (20) Susa 66 (4) 224 461 Susa 306 377 (10) (21) (14) (16) Mish 75 (4) 476 Susa 312 385 (21) (14) (16) Susa 89 (5) 142 (7) Mish 353 other seal image not published (15) Mish 101 122 (6) Susa 372 stamp sealing not listed (5) (16) Susa 111 319 Mish 422 474 (6) (15) (17) (21) Table 6.8. The multiple seal-impressed bullae from Susa and Chogha Mish. The number in parentheses indicates STS sub-group allocation.

In summary, two different types of artefacts are classified broadly as bullae, the true, token-bearing, protoliterate (Susa II, STS) round clay envelopes called bullae, and the lumps or balls of clay, labelled bulles. The bulles may be considered another form of non-writing sealing, as jar sealings and others discussed above. The bullae are a more specific group, functionally and in form, but are also non-writing associated administrative devices. Bullae appear to be a purely STS (Susa II) Khuzistan phenomenon (that is, within the Corpus, the presence of bullae in

ECS Corpus, Volume I, Part III 650 Mesopotamian Uruk sites should be noted). The precise function of bullae, and the reasons for their proliferation in the Susa II period, and then absence, is related to the overall early glyptic function, including that associated with writing (or proto- writing), and as such will be returned to in the articulation of these glyptic types (that is, sealed tablets).

6.1.1.3 Summation and Conclusion There are thus several categories of sealing types that are here classified as non- writing associated administrative sealings. All sealings in the Corpus are administrative, in that they all function as a control mechanism. The items classified in this group are those not associated with writing. The items labelled are classified in this group as they are not associated with writing (indeed, their use pre- dates the period of true writing), though they do form a distinct group, functionally and in form. The various other sealings, both classified and undifferentiated, were apparently used to mark or identify goods, or identify individuals or associations (groups of individuals) who claimed ownership over the goods so marked. The bullae presumably recorded or transferred information, and thus may have been functionally related to, and subsequently superseded by tablets (both sealed and otherwise, as the written text of an unsealed tablet may have conveyed more or less the same information as the seal impression on a bulla). Table and Graph 6.9 illustrate the sealing types within the non-writing associated administrative sealings group. As is demonstrated the largest single group in the classification are simple undifferentiated sealings, the fact that these items are undifferentiated, that is, not classified according to sealing type (jar sealing, wall lock sealing etc.) is a symptom of the age and approach of the publications that formed the majority source of this study. With a physical examination it is hoped that many (but not all, as one must anticipate some non-diagnostic items amongst the corpus) of these items would be classifiable to any number of the specific sealing type classifications (though probably not bullae, as it is held that bullae are easily recognisable, and have been discerned in many older publications, or when there missed, may be identified from a published image). Bullae as a group form the largest of the classified sealing types, though it should be recalled that one hundred and eighty-four are not the total number of actual bullae, but of seal impressions contained on bullae, with thirty-eight bullae that bore two or more sealings. The dominance of bullae as a form is even

ECS Corpus, Volume I, Part III 651 more startlingly given the fact that this form was limited to one period only (the Susa II period). However, apart from the dominance of bullae in this period, little more may be commented on regarding these proportions given the fact that the majority of items have not been classified.

Non-writing associated administrative sealings type distribution Ball of clay 1 Bale sealing 4 Door lock sealing 7 Sealed clay cone 2 Door sealing 15 Clay tag sealing 5 Wall lock sealing 3 White plaster sealing 1 Flat(tish) sealing 2 Fired clay sealing 3 Jar stopper 3 Bulla(e) 184 Bottle stopper 2 Bulles 77 Jar sealing 58 Unclassified sealing 847 1214 Table 6.9. Type distribution of the non-writing associated administrative sealings.

Table and Graph 6.10 illustrate the site distribution of the non-writing associated administrative devices. As is demonstrated, non-writing associated sealings occur at all the sites of the Corpus that produced sealings (thus excluding the seal only sites, Choga Zanbil, Surkh Dum-i-Luri, Chigha Sabz, Bani Surmah, Kalleh Nisar and Tepe Djamshidi), except for the extremely limited contributing source Chogha Gavaneh (as this site only produced two glyptic items, each from the same context, the absence of non-writing associated sealings from this site need not subtract from the general distribution pattern). Thus non-writing associated administrative sealings occurred across the area defined as ‘Elam’, and the three historical provinces. The general proportions of this distribution conform to the general sizes of the site corpora, and thus little comparative data may be sourced from this avenue of investigation. The large contribution from Chogha Mish and Haft Tepe, essentially single period sites, should be noted however, as both may indicate specific glyptic functions at each site, further detailed below following the articulation of the other glyptic function.

Non-writing associated administrative sealings site distribution Susa 859 Tepe Sharafabad 3 Chogha Mish 129 Kamtarlan 1 Godin Tepe 3 Haft Tepe 177 Tal-i Malyan 42 Table 6.10. Site distribution of the non-writing associated administrative sealings.

ECS Corpus, Volume I, Part III 652 860 840 820 800 780 760 740 720 700 680 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

g g r g le lay n in ng ng ing ing in l ed l l a(e) ifi eali al al ll Bu of c s seali stoppe seali sea sea Bu all or se k te stopperar ag se class B o Jar t J ale t ster n D l loc Bo B la U or lock al o Flat(tish) sealing Sealed clayClay conete p D W hi Fired clay sealing W

Graph 6.9. Type distribution of the non-writing associated administrative sealings.

ECS Corpus, Volume I, Part III 653 860 840 820 800 780 760 740 720 700 680 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 Susa Chogha Haft Tepe Tepe Kamtarlan Godin Tepe Tal-i Mish Sharafabad Malyan Graph 6.10. Site distribution of the non-writing associated administrative sealings.

ECS Corpus, Volume I, Part III 654 Table and Graph 6.11 illustrate the stylistic distribution of the non-writing associated sealings, and Table and Graph 6.12 the chronological distribution of the items. As is demonstrated, the STS and therefore the Susa II period strongly dominates this survey. Even if one removes the significant contribution of the sealed bullae to the Susa II corpus, it is still the largest chronological division (with two hundred and fifty-three items). It is also of note that the earlier periods dominate the distribution, with a general pattern or trend of reduced sealing use over time. The anomaly of the EME style and early Middle Elamite period may be attributed to the specific function of the excavated area at Haft Tepe (that is, a craft production associated area, or an area associated with the reception of raw, or semi-worked, materials). The reason for the early dominance of non-writing associated glyptic material will be further detailed below, but no doubt it is linked to the fact that ‘writing’ in its true, functional form, was in fact not invented or developed in the Susa II period certainly, and only in a ‘proto’ or early sense in the Susa III.

Non-writing associated administrative sealings style distribution STS 437 PEU 12 JNRS 40 OBRS 21 CPE 100 PEO 3 GS 74 EME 180 AGD 60 KRS 13 STF 138 LME 19 SF 22 LPS 2 LSF 2 LGD 2 ARS 13 Miscellaneous 2 PEA 1 Unclassifiable 19 UTRS 40 Not Illustrated 6 Table 6.11. Style distribution of the non-writing associated administrative sealings.

Non-writing associated administrative sealings chronological distribution Susa II 437 Akkadian & Awan 14 Susa III 214 Ur III & Shimashki 52 Susa III/IV 138 Sukkalmah 24 Early Middle Elamite 180 Susa IV 24 Late Middle Elamite 42 Table 6.12. Chronological distribution of the non-writing associated administrative sealings.

ECS Corpus, Volume I, Part III 655 440 430 420 410 400 390 380 370 360 350 340 330 320 310 300 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

S S F E E S us TS R PE G GD TF S SF RS EA EU EO M RS M A PS GD o ble ted S C S L A P TRS P BRS P K L L a ra JN A U O E L lust Il iscellanenclassifiot M U N Graph 6.11. Style distribution of the non-writing associated administrative sealings.

ECS Corpus, Volume I, Part III 656 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

II h III IV ite ite usa am m S usa III/IV usa Awan l S S & imashki usa h ukkalma le Ela S ian S S d idd II & Middle E M kka r I A U arly ate E L Graph 6.12. Chronological distribution of the non-writing associated administrative sealings.

6.1.2 Writing administration As one would expect, this classification also concerns sealings as artefacts of administration, though these items are classified according to their association with written texts. For purposes of classification, proto-cuneiform, ‘Proto-Elamite’ and numerical texts are subsumed under the rubric ‘writing’. See Damerow & Englund (1989), Friberg (1978; 1979) and Le Brun and Vallat (1978) for details of the development of writing and the classificatory differences between these types. Two particular, though related, sealing types are associated with writing, sealed tablets and sealed envelopes.

6.1.2.1 Tablets The act of sealing tablets is perhaps the most commonly cited and described seal practice (Gibson & Biggs [eds.] 1977; Hallo & Winter [eds.] 2001; Kelly-Buccellati

ECS Corpus, Volume I, Part III 657 [ed.] 1986; Steinkeller 1977; Teissier 1984; Collon 2005: 113 – 119), though, as will become apparent in the following survey, sealed tablets, or sealed items associated with writing generally, are not the most dominant sealing form in the Elamite Corpus. The act of sealing a tablet may be described as synonymous with a modern signature, or corporate letter-head, and presumably indicated either, the participation in, or witness to, the contract or transaction detailed in the text. This authorisation and indication may have been under the authority of an individual (that is, the seal belonged to, and was an indicator of, an individual person), or a group, corporation or association. Table and Graph 6.13 demonstrate the distribution of sealed tablets across the Elamite sites. The distribution is wide across ‘Elam’, though slightly diminished on that of the non-writing administration discussed above. Thus, Tepe Sharafabad and Kamtarlan did not produce sealed tablets. The early date of the Sharafabad corpus (that is, Susa II period, and therefore before the development of true writing), and the small size of both this and the Kamtarlan corpus may describe this absence. The single item from Chogha Gavaneh (it should be remembered that this single item is in fact five tablets impressed by the same seal) should also be noted as it differs from the non-writing sealing distribution. The fact that tablets occur across the ‘Elamite’ realm (and the three historical provinces) is important. The striking dominance of Susa as a contributing source is linked to the extraordinary general size of this corpus, but may also be caused by the fact that Susa was a cultural, and administrative, capital for much of its existence.

Sealed Tablets Site Distribution Susa 320 Godin Tepe 5 Chogha Mish 9 Chogha Gavaneh 1 Haft Tepe 19 Tal-i Malyan 3 357 Table 6.13. Site distribution of sealed tablets.

ECS Corpus, Volume I, Part III 658 320 310 300 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Susa Chogha Mish Haft Tepe Godin Tepe Chogha Tal-i Malyan Gavaneh Graph 6.13. Site distribution of sealed tablets.

ECS Corpus, Volume I, Part III 659 Table and Graph 6.14 illustrate the stylistic distribution of the ‘Elamite’ sealed tablets, and Table and Graph 6.15 the implied chronological distribution. Several facets of this survey warrant discussion. Firstly, the complete absence of the ARS and PEA style items, and thus any ‘Akkadian and Awan’ period sealed tablets is striking, and may be linked to a different administrative and control system during this period in Elam. It should be noted that non-writing associated sealings do occur in the ‘Akkadian and Awan’ period, though in limited numbers (see Table and Graph 6.12). This may imply a general limited use of sealing and seals during this period. This is further indicated by the fact that the ‘Akkadian and Awan’ period is the least represented in terms of seals also (see below for details), further implying a reduced glyptic employment in this period. This may be linked to the ‘political’ structure of ‘Elam’ in this period, where the area was dominated by a foreign power in the earlier period (the Akkadians, who, as discussed above, did not impose an administrative system on the conquered area like that imposed by the Ur III state), and a non- centralised ‘Awanite’ system in the later. The development of the writing system known as ‘Linear Elamite’ towards the end of this period may indicate that the administrative and control systems of ‘Elam’ in this period was different from the Mesopotamian-type glyptic based system during this period. The increased (from the preceding pattern) number of EME (and therefore early Middle Elamite period) items may indicate an increased sealed tablet use in this period, but it may also be linked to the extraordinary contribution of Haft Tepe (already discussed above).

Sealed Tablets Style Distribution STS 77 STF 11 PEO 1 JNRS 9 SF 1 EME 32 CPE 168 UTRS 14 AS 2 GS 6 PEU 1 No Image 21 AGD 3 OBRS 4 Unclassifiable 7 Table 6.14. Style distribution of sealed tablets.

ECS Corpus, Volume I, Part III 660 170

160

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0

S S F E S TS R PE G TF S EU EO A ble S C GD S TRS P BRS P M a JN A U O E o Image N nclassifi U Graph 6.14. Style distribution of sealed tablets.

Sealed Tablets Chronological Distribution Susa II 77 Ur III & Shimashki 15 Susa III 183 Sukkalmah 5 Susa III/IV 11 Early Middle Elamite 32 Susa IV 1 Late Middle Elamite 2 Table 6.15. Chronological distribution of sealed tablets.

ECS Corpus, Volume I, Part III 661 185 180 175 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0 Susa II Susa III Susa III/IV Susa IV Ur III & Sukkalmah Early Late Shimashki Middle Middle Elamtie Elamite Graph 6.15. Chronological distribution of sealed tablets.

Finally, the large number of CPE items, and those of the STS warrants some discussion. Firstly, the significant number of Susa II period sealed tablets is notable as these items are not inscribed with a true, or established writing, but are so-called ‘numerical tablets’, or pre-writing (or proto-cuneiform) texts, that appear to be primarily numerical in nature and record-keeping devices in function (Damerow &

ECS Corpus, Volume I, Part III 662 Englund 1989; Friberg 1978; 1979; Schmandt-Besserat 1978; 1981; Englund & Grégoire 1991). The markings on these tablets may not be considered true written language as they lack a grammar or system to transfer abstract ideas or pieces of information (Schmandt-Besserat 1981). Essentially all that may be conveyed in these tablets are records of numbers and amounts of objects (Friberg 1978; 1979; Damerow & Englund 1989; Schmandt-Besserat 1978; 1981). These tablets are an early step in the development of writing, and indeed led to the ultimate formulation of a written language, and thus are classified here under the rubric ‘writing’. The dominance of sealed tablets in the Susa III period (with the combination of the JNRS, CPE and GS style items) is of note. This may be attributed in part to the disproportionate study and publication of ‘Proto-Elamite’ tablets, the tablet type upon which the Susa III items are generally impressed. As has already been detailed, the ‘Proto-Elamite’ texts form a distinct, and uniquely ‘Elamite’ (perhaps the term ‘Iranian’ should be preferred) assemblage, that has been the subject of extended and continual study since their discovery in the late 1800’s. This has led to a significant number of published ‘Proto-Elamite’ tablets, and in turn therefore a significant number of these items available for inclusion here. The non-‘Proto-Elamite’ texts, that is, those inscribed with true cuneiform script and texts (generally containing various Akkadian dialects and Elamite, but more rarely also Sumerian) have mostly been discovered at Susa (as have the ‘Proto-Elamite’ texts) but have been published in a different manner. That is, generally a sealing on a ‘Proto-Elamite’ tablet was noted, and indeed, at least one volume (Legrain 1921) was essentially devoted to these items from Susa, while the same cannot be said for the other (non-‘Proto-Elamite’) texts, where often a transliteration or copied version of the text was supplied, without reference to or replication of a sealing (or note as to whether such an item was indeed present). Indeed, many tablets found at Susa over the more than one hundred years of excavation have yet to be published or systematically studied, raising the possibility that any number of these items may have been sealed, a fact that may, on recognition, alter the relative distribution proportions significantly of ‘Proto-Elamite’ and other texts tablets, and therefore the chronological distribution of sealed tablets. There is however evidence that many of the published and known cuneiform tablets did not bear an impression of any kind. This may imply, though the dispropionate nature of the evidence must cause some hesitation in the application of any implication, a different glyptic function between the early, proto-script writing

ECS Corpus, Volume I, Part III 663 systems (the numerical Susa II texts and the ‘Proto-Elamite’ Susa III texts) and the later true writing cuneiform texts. The functional difference may be one of information transference. That is, if only basic numerical and type information may have been transferred by the ‘numerical’ and ‘Proto-Elamite’ (as the ‘Proto-Elamite’ texts remain essentially undeciphered, the exact amount of information transferable in this script is currently unknown, but they may be related to the contemporary proto- cuneiform texts from Mesopotamia in form and content [Damerow & Englund 1989; Englund & Grégoire 1991; Friberg 1978; 1979], and thus may be similarly considered pre-writing in structure) texts (and for that matter the ‘bullae’ discussed above), the use of a cylinder (or any other) seal may have provided further, more detailed information, such as origin and destination of objects, the identity of individuals or groups involved in a given transaction, or the time and place of the action of recording. Thus the act of sealing was an important, and necessary, part of the information transference of the sealing, as they provided important elements of information. Following the full development of writing, where such abstract ideas and pieces of information may have been imparted by the script itself, the use of seal impressions to provide such information became redundant, and the function of sealing a tablet may have implied something different altogether. This may explain both the large number of early, proto-literate sealed tablets (of both the Susa II and Susa III periods; the sealed bullae of the Susa II period may also be explained this way), and the generally reduced and limited function of sealed tablets in the true literate periods. Thus, in the earlier periods, the function of a sealing on a tablet may impart an important piece of information (whether identity, location, origin, destination, time, place or any or all of these), required to understand fully the transaction or contract detailed. In the later literate periods, a sealing on a tablet may have concurred with the general, and traditional functional interpretation, that of identification of a witness or authorisation of a contract or transaction. This functional difference between early and later sealed tablets is hypothetical, and based upon the chronological distribution difference between the early and later periods, and the classificatory difference between ‘proto-’ and ‘true’ writing. The further acceptance of this proposal must await further testing and analysis, including philological study of the sealed texts in association with the glyptic styles (a study type that has not often been undertaken).

ECS Corpus, Volume I, Part III 664 6.1.2.2 Sealed Envelopes Sealed clay envelopes form a small and specific group of writing associated administrative devices. These items are essentially clay containers that contained inscribed tablets. They are therefore functionally related to bullae, though the envelopes contain inscribed tablets (themselves sometimes also sealed), not tokens, and were often not only sealed with sealings, but inscribed with a summary or explanation of the text inside (Collon 2005: 113 – 116; and therefore, are associated with writing, unlike bullae). The precise function and reason for the extra layer of authentication and safeguarding provided by the encasing and sealing of a tablet (already itself sealed and presumably therefore authenticated) in an envelope, that was also sealed and contained the same information, is unclear (Collon 2005: 113 – 116). Presumably, if a dispute as to the record as contained on the envelope arose, it would be broken and the tablet consulted (Collon 2005: 113 – 116), though the efficacy and usefulness of this practice is unclear if, presumably, the testimony of the envelope was thus not trusted in a dispute. Table and Graph 6.16 illustrate the site distribution of ‘Elamite’ sealed envelopes. As is demonstrated this small group is primarily represented at Haft Tepe, with two examples from Susa. This sealing type is thus restricted to Khuzistan within ‘Elam’ (and indeed, within Khuzistan, only in the specific Susiana plains regions in which Haft Tepe is also placed). Table and Graph 6.17 further demonstrate the stylistic division of the sealed envelopes and Table and Graph 6.18 the chronological distribution. As is evidenced, all but one of the sealed envelopes are impressed with EME items, and therefore date to the early Middle Elamite (and the late Sukkalmah) period. On the basis of this, admittedly very limited survey of details, it may be argued that ‘sealed envelopes’ were essentially an early Middle Elamite, Susiana region phenomenon.

Sealed Envelopes Site Distribution Susa 2 Haft Tepe 7 9 Table 6.16. Site distribution of sealed envelopes.

ECS Corpus, Volume I, Part III 665 7 6 5 4 3 2 1 0 Susa Haft Tepe Graph 6.16. Site distribution of sealed envelopes.

Sealed Envelopes Style Distribution LSF 1 EME 8 Table 6.17. Style distribution of sealed envelopes.

8 7 6 5 4 3 2 1 0 LSF EME Graph 6.17. Style distribution of sealed envelopes.

Sealed Envelopes Chronological Distribution Susa IV 1 Early Middle Elamite 8 Table 6.18. Chronological distribution of sealed envelopes.

8 7 6 5 4 3 2 1 0 Susa IV Early Middle Elamite Graph 6.18. Chronological distribution of sealed envelopes.

ECS Corpus, Volume I, Part III 666 6.1.2.3 Summation and Conclusion The second of the two administration classifications of seal function is that which is associated with writing. As well as inscribed tablets, the most obviously recognised and understood sealing form, inscribed sealed envelopes, and a single seal identified as administrative in function can be included in this group, as is demonstrated by Table and Graph 6.19. The administrative function is allocated to this single seal from Chogha Gavaneh as it was found in a clear administrative context, in the Old Babylonian (Sukkalmah period) ‘archive’ at the site (already discussed above, Chapters 2 and 5).

Writing associated administrative glyptic type distribution Sealed tablet 357 Sealed envelope 9 Administration seal 1 367 Table 6.19. Type distribution of the writing associated administrative items.

Table and Graph 6.20 summaries the site distribution of the ‘Elamite’ administrative items associated with writing. Table and Graph 6.21 illustrate the stylistic distribution, and Table and Graph 6.22 the chronological distribution. These surveys indicate that the majority of writing associated glyptic administrative items were provenanced from Susa, and that such items were more common in the earlier periods of this study. The dominance of Susa may be indicative both of the extensive occupation and excavation (and thus publication) of this large site, but also its position as a capital or major ‘Elamite’ city for much of this occupation.

Writing associated administrative site distribution Susa 322 Godin Tepe 5 Chogha Mish 9 Chogha Gavaneh 2 Haft Tepe 26 Tal-i Malyan 3 Table 6.20. Site distribution of the writing associated administrative items.

Writing associated administrative style distribution STS 77 AGD 3 UTRS 14 EME 40 JNRS 9 STF 11 PEU 1 AS 2 CPE 168 SF 1 OBRS 5 No Image 21 GS 6 LSF 1 PEO 1 Unclassifiable 7 Table 6.21. Style distribution of the writing associated administrative items.

ECS Corpus, Volume I, Part III 667 360 350 340 330 320 310 300 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0 Sealed tablet Sealed envelope Administration seal Graph 6.19. Type distribution of the writing associated administrative items.

ECS Corpus, Volume I, Part III 668 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 Susa Chogha Mish Haft Tepe Godin Tepe Chogha Tal-i Malyan Gavaneh Graph 6.20. Site distribution of the writing associated administrative items.

170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

S F F E e PE GS GD S EU EO AS ble ST C STF LS P P EM JNRS A UTRS OBRS Imag ifia o ss N la nc U Graph 6.21. Style distribution of the writing associated administrative items.

ECS Corpus, Volume I, Part III 669 Writing associated administrative chronological distribution Susa II 77 Ur III & Shimashki 15 Susa III 183 Sukkalmah 6 Susa III/IV 11 Early Middle Elamite 40 Susa IV 2 Late Middle Elamite 2 Table 6.22. Chronological distribution of the writing associated administrative items.

190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

II II V i te te I IV II/ sa I almah lami Susa Susa Su imashk Elami E sa I e Su Sukk I & Sh iddl II M Ur arly ate Middle E L Graph 6.22. Chronological distribution of the writing associated administrative items.

Tables and Graphs 6.21 and 6.22 do not imply that tablets (sealed or otherwise) were more commonly used in the earlier (pre- or proto-literate) periods generally, but that sealed tablets were more common in this period. This early major usage is possibly functional. It is here proposed that certain elements of information that were transferable in the later fully developed written language of the true cuneiform texts, were not included in the earlier proto-cuneiform and numerical tablets (elements such as origin, destination, location, source, authorising or validating individuals or groups amongst others). In these early protoliterate tablets, this information is instead transferred by the seal impression on the tablet (a similar function may also be attributed to the non-writing associated bullae sealing types). Thus in later periods, after writing was sufficiently developed so as to convey such pieces of information, the necessity of sealing was reduced, and thus the number of sealed tablets was greatly reduced. In these later period texts, the action of sealing a

ECS Corpus, Volume I, Part III 670 tablet (or envelope) presumably took on a different function, that may be described as essentially that of identification and authorisation of the participants and witnesses of a transaction. Why some tablets were sealed, while others apparently of the same type were unsealed is unclear, and currently unknown (Collon 2005: 113 – 119). The apparent increase of sealed tablets and other writing associated administrative sealings in the early Middle Elamite period (as demonstrated by the EME items), may be functional (that is, for some unknown, undiscerned reason, the sealing of a tablet was required more often in this period), or may merely be an accident of discovery (that is, pure more items of this date have simply been uncovered, possibly linked to the excavation and contribution of Haft Tepe).

6.1.3 Summation of Administration Function The first, and defining, function of ‘Elamite’ glyptic material is the administrative function. This function is described as defining as this is the action of sealing, that lends the name seal to these devices. The catchall ‘administration’ (which is used to imply control and recording mechanisms) contains several separate functional classes and uses. In a general sense, two broad groups can be identified. The first contains sealings not associated with writing, and the second those items that are so associated. The non-writing associated administrative devices include jar, bale, door lock, wall, wall lock and clay tag sealings and bottle and jar stopper sealings, as well as a group labelled in French bulle (or balls or clumps of clay), and the specific group bullae. Writing associated administrative devices include inscribed sealed tablets and inscribed sealed clay envelopes. Table and Graph 6.23 illustrate the proportions of the administrative function items within the Elamite Corpus.

Summary of the glyptic types of the administrative function items Sealing(s) 953 Sealed Tablets 357 Bulla(e) 184 Sealed Envelopes 9 Bulle(s) 77 Administrative seal 1 1581 Table 6.23. Survey of the glyptic types of the administrative function items.

ECS Corpus, Volume I, Part III 671 1000 980 960 940 920 900 880 860 840 820 800 780 760 740 720 700 680 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

) (s) (e ets eal a ulle opes ing B el ve S d Tabl Seal e rati ed Env st Seal Seal mini Ad

Graph 6.23. Survey of the glyptic types of the administrative function items.

ECS Corpus, Volume I, Part III 672 In terms of identification, all sealings were classified as items that were ‘sealed’, as a means of control or for recording purposes, and so were classified as administrative devices. The type of sealing further allowed for these items to be classified as non-writing or writing associated. As will be further detailed below, in the current study glyptic function was defined by type and context. Thus, a sealing was identified as an administrative device on the basis of its glyptic type. For seals, function allocation is more uncertain, and can only be definitely reconstructed in certain instances of specific contexts. These instances will be further detailed below (but basically are defined as votive, funerary, ‘palatial’ and [non-votive] hoard items), but for now it should be noted that a single seal only, 2713, from Chogha Gavaneh was allocated an (associated with writing) administrative function, due to its provenance within a strongly administrative context (an ‘archive’ at that site). The vast majority of cylinder seals in the Corpus were not allocated a specific function however, as their context was generally unknown, or when known in detail, the general reconstruction of the context, and the association of the items within these contexts were too fragmentary to allow for such an identification. It may be assumed, and is indeed probable, that a significant number of the items classified as ‘unknown function’ here had a basic administrative function to a lesser or greater degree (indeed, it should also be recalled, and will be further addressed below, that at least some of the cylinders allocated another function below, were also employed as administrative devices prior to their deposition in certain contexts [such as a tomb] that led to their ultimate functional allocation). As these items could not be sufficiently so identified, and as their general fragmentary provenance and contextual information would add little of substance to the current discussion besides, these seals have not been classified as administrative here. Table and Graph 6.24 illustrate the site distribution of the ‘Elamite’ administrative devices. Graph 6.24 demonstrates the relative proportions of this distribution by the writing (coloured blue) and non-writing (coloured red) associated administrative devices. These figures indicate both that the sealing site corpora conform to similar patterns and trends for the overall Corpus (thus Susa is the dominant site, with significant sized corpora from Chogha Mish, Haft Tepe and, to a slightly lesser degree, Tal-i Malyan, with small, limited corpora from Tepe Sharafabad, Kamtarlan, Godin Tepe and Chogha Gavaneh), and that in most cases non-writing associated administrative devices out-number writing associated devices.

ECS Corpus, Volume I, Part III 673 Thus, only at two small corpora, from sites that underwent limited, small scale excavations, Godin Tepe and Chogha Gavaneh, do no non-writing associated items appear. This would seem to indicate that general, non-writing sealing to mark, identify and secure both mobile (vessels, baskets, bales) and immobile (rooms and building areas) storage devices was the most common sealing function in ‘Elam’, and not, as may have previously been assumed (or at least implied by the literature focus) the practice of sealing tablets. Thus it may be said that across ‘Elam’ (as defined here), sealings were the dominate form of administrative devices. Sealings occurred at all sites that did not have a specific and general function, be it ‘cultic’ (Choga Zanbil and Surkh Dum-i-Luri), funerary (Bani Surmah, Kalleh Nisar, Tepe Djamshidi) or non-standard (Chigha Sabz).

Summary site distribution of administrative function glyptic items Non-writing Writing Total Non-writing Writing Total Susa 859 322 1181 Kamtarlan 1 – 1 Mish 129 9 138 Godin 3 5 8 Haft 177 26 203 Gavaneh – 2 2 Sharafabad 3 – 3 Tal-i Malyan 42 3 45 Table 6.24. Survey of the site distribution of the administrative function glyptic items.

ECS Corpus, Volume I, Part III 674 1200 1180 1160 1140 1120 1100 1080 1060 1040 1020 1000 980 960 940 920 900 880 860 840 820 800 780 760 740 720 700 680 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 Susa Chogha Haft Tepe Tepe Kamtarlan Godin Chogha Tal-i Mish Sharafabad Tepe Gavaneh Malyan Graph 6.24. Survey of the site distribution of the administrative function glyptic items, red indicates non-writing associated items, blue those items associated with writing.

ECS Corpus, Volume I, Part III 675 Table and Graph 6.25 illustrate the proportional style distribution of the administrative sealings (according to those classified as ‘non-writing’ and those associated with writing), and Table and Graph 6.26 the chronological distribution. As is evidenced by Table and Graph 6.26, non-writing associated sealings are more common than writing associated sealings (tablets) across the entire chronological span of this study. Even in the earlier periods (Susa II and III), when, as outlined above, sealings appeared to be more frequently impressed upon tablets (as a means of information transference not yet facilitated by writing), non-tablet associated sealings are more common. Table and Graph 6.26 demonstrate that non-writing associated sealings dominate writing associated sealings in all stylistic divisions, except for the CPE style. The dominance of sealed tablets in this style may be linked to the over representation of ‘Proto-Elamite’ tablets (upon which a large number of the CPE glyptic items were impressed) in published materials, and thus their over representation here. Though some functional cause (that is, tablet sealing was simply more regularly undertaken than non-writing sealing in this period) cannot be dismissed however.

Summary style distribution of administrative function glyptic items Non-writing Writing Total Non-writing Writing Total STS 437 77 514 OBRS 21 5 26 JNRS 40 9 49 PEO 3 1 4 CPE 100 168 268 EME 180 40 220 GS 74 6 80 KRS 13 – 13 AGD 60 3 63 LME 19 – 9 STF 138 11 149 AS 8 2 10 SF 22 1 23 LPS 2 – 2 LSF 2 1 3 LGD 2 – 2 ARS 13 – 13 Miscellaneous 2 – 2 PEA 1 – 1 No Image – 21 21 UTRS 40 14 54 Unclassifiable 19 7 26 PEU 12 1 13 Not Illustrated 6 – 6 Table 6.25. Survey of the style distribution of the administrative function glyptic items.

ECS Corpus, Volume I, Part III 676 520

500

480

460

440

420

400

380

360

340

320

300

280

260

240

220

200

180

160

140

120

100

80

60

40

20

0

e S F O S S ed RS G S EA E ME RS ME A P GD ag ble t STS CPE STF LSF ARS P PEU P E K L L L m ia ra JN AGD UTRS OBRS neous st o I la ssif u N el la ll c t I o Misc Un N Graph 6.25. Survey of the style distribution of the administrative function glyptic items, red indicates non-writing associated items, blue those items associated with writing.

ECS Corpus, Volume I, Part III 677 These surveys indicate that non-writing associated administrative sealing is the most common sealing action across the area of Elam (where any administrative sealing took place, thus excluding Choga Zanbil, Surkh Dum-i-Luri, Bani Surmah, Kalleh Nisar, Tepe Djamshidi and Chigha Sabz), and across the entire period of study here. It may be stated that this function is therefore the standard, or regular, glyptic function. Unfortunately, due to the nature of the material and this study, the majority of the items classified as non-writing associated administrative sealing devices cannot be adequately allocated a classificatory type. This means that further reconstructions regarding sealing practices, such as which type of non-writing associated sealing was more common (for example, jar sealing, door lock sealing and so on) cannot be undertaken here, and must await a thorough and complete physical examination of the material so as to resolve this classification dilemma.

Summary chronological distribution of administrative function glyptic items Non-writing Writing Total Susa II 437 77 514 Susa III 214 183 397 Susa III/IV 138 11 149 Susa IV 24 2 26 Akkadian & Awan 14 – 14 Ur III & Shimashki 52 15 67 Sukkalmah 24 6 30 Early Middle Elamite 180 40 220 Late Middle Elamite 42 2 44 Table 6.26. Survey of the chronological distribution of administrative function glyptic items.

Following the general proliferation of sealing types (both associated with writing and not) in the early chronological periods (Susa II and III), a general pattern of decline and reduced use is demonstrated by Tables and Graphs 6.24 and 6.25. An explanation for this trend may be hypothesised, though it must be only a preliminary suggestion at the moment. As already hypothesised, in the period before the codification and development of a script that can be described as ‘true’ writing (that is, a writing system that can express abstract ideas and complex sentences, and that is not purely a mnemonic or recording device, a writing system that in short, can express language), seal impressions carried some of the information not expressible by the numerical or other notations (information such as direction or location that the sealed item may be ‘dispatched’, individuals or groups associated with the transaction, or time or date of the operation). This sealing information transference can be discerned

ECS Corpus, Volume I, Part III 678 on the sealed bullae, and the sealed (numerical, proto-cuneiform and ‘Proto-Elamite’) tablets, and possibly in a similar, though different means on the majority non-writing associated sealings. Following the development of a fully articulated and functioning written language, that could transfer ideas and detailed, abstract pieces of information, the use of seal impressions for this purpose was greatly diminished and eventually no longer required. Thus the practice of sealing items not associated with writing, and those that were both changed. For the tablets, a seal impression, rather than providing information and details, probably indicated the participation in a given transaction, the authorisation of the contact or the validity of the details contained on the tablet, or the witnessing of the creation of the tablet by certain individual(s) or representative(s) of a group (or association). The reduction in non-writing associated sealing may also be linked to the development of a true writing system, as tablets (sealed or otherwise) could be used to record information (in a place or record removed from the actual objects themselves) regarding objects, thus not requiring the seal impression to carry such information. In the truly literate society, the practice of sealing items other than tablets (and envelopes) continued, but only in certain circumstances, perhaps when it was necessary to secure or further authorise goods. Thus, in the literate era, sealed tablets were only sealed some of the time, and it may be assumed that mobile and immobile storage units were left unsealed (as the information was recorded on tablets; thus the reduction in sealings in this era should not be taken as evidence for reduction in commercial or economic transactions, as is testified in written texts), in both cases, the motivation and reasons for sealing and non-sealing generally evade knowledge. Thus, the reduction of sealing items, both associated with texts and otherwise, that appears coincident with the development of a true writing system may be due to the reduced necessity of seals as items of information transference. Thus, the seal was superseded, and replaced as a device for carrying and transferring information by the writing system. In the survey presented by Tables and Graphs 6.25 and 6.26, the general pattern of sealing reduction following the early proliferation is broken by two periods in which sealing use increased from the previous period (though it should be noted that in neither event do sealings reach the height of the early period sealing numbers). The first such period is in the ‘Ur III and Shimashki’ period, and is associated with the PEU, but more totally with the UTRS style (that is, the ‘Mesopotamian Related’ style of this period). This rise in the use of UTRS Ur III period seals as sealing

ECS Corpus, Volume I, Part III 679 devices may be linked to the imposition of a more rigid and structured administrative system by the highly bureaucratised Ur III state (already detailed above).

520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

II h III IV ite ite usa am m S usa III/IV usa Awan l S S & imashki usa h ukkalma le Ela S ian S S d idd II & Middle E M kka r I A U arly ate E L Graph 6.26. Survey of the chronological distribution of the administrative function glyptic items, red indicates non-writing associated items, blue those items associated with writing.

ECS Corpus, Volume I, Part III 680 The second major rise in sealing function, in the early Middle Elamite period, as represented by the EME style, may be linked to the significant contribution of Haft Tepe to this style/period, and the area of this site from whence this material was sourced. It has been hypothesised that the ‘Terrace Complex’ area of Haft Tepe (see Figure 2.7 and Chapter 2 above for details) generally, and the area in which the sealings specifically were uncovered had a craft production function (in association with a large kiln). If this function for this context is accurate, then the proliferation of sealings, and particularly those of non-writing associated devices (that is, sealings and not sealed tablets) in this area at Haft Tepe may be linked to the delivery or receipt of raw materials in the area, shipped from other parts of Haft Tepe or other sites, to this area, where the seals were broken and the goods used. Unfortunately, the fragmentary information regarding Haft Tepe, and the fact that on the current evidence the specific sealings type of the Haft Tepe material is unknown (thus if these items were jar sealings, a point currently unknown, it may indicate that goods were delivered in this item, if wall or door lock sealings, this may imply that items were stored in the area), this suggestion cannot be further demonstrated or proven, and must await further investigation of this material. Thus, the general, and definitional, function of Elamite seals can be described as sealing devices as a form of administration or control. This function can be divided into two areas, writing and not writing associated types. The non-writing type is the most common, and includes (as identified above) the seal impression of clay devices attached to jars and vessels (jar sealings, basket sealings, bale sealings, jar stopper sealings, bottle stopper sealings, bulles), to walls or doors (door lock, door and wall lock sealings), or not attached to devices but not associated with writing (bullae). Those items associated with writing are impressed upon inscribed objects, be they tablets (including numerical, proto-cuneiform and ‘Proto-Elamite’ text tablets as well as standard cuneiform tablets) and sealed envelopes. In the early, pre-literate periods, seal impressions on a device (including sealed tablets and bullae) may have transferred important pieces of information, not currently transferable by the still un()developed writing system, such as identity of goods or individuals involved in a transaction or action regarding the sealed items, or location, origin, or destination of this material. In the literate era, seal impressions on non-writing associated sealings may have similarly identified an individual involved, or possibly authorised the integrity of the goods so ‘sealed’, while on a sealed tablet, may have indicated the

ECS Corpus, Volume I, Part III 681 identity of the individuals involved in or witnesses the action or transaction detailed in the written text of the tablet. In the literate era, both the sealing not associated with writing and that which is (that is, on tablets) was reduced from the proliferation in the early periods (or at least thus is the picture presented for Elam in the currently available evidence in the Corpus), possibly due to the fact that information formerly transferred by a seal impression could now be contained in the text, and thus sealing in some cases was replaced by written text. The exact motivation for why some tablets were still impressed, and others not, and some goods were presumably no longer sealed at all (that is the vessels and rooms formerly sealed and producing the non-writing associated devices) while other were still so marked, remains unclear.

6.2 Symbolic While the administrative function of seals has been described above as the standard and definitional seal function, and is indeed, what makes a seal a seal and not some other device, another glyptic function is here described as ‘symbolic’. This term implies the non-administrative, or non-sealing, function of seals. As already discussed, all seals, probably to varying degrees, had an amuletic or symbolic function. It is also held that the seals classified here with a specific, contextual symbolic function, previously had a standard administrative function prior to their deposition that indeed enabled and led to their symbolic deposition. Thus the dichotomy as presented between administrative and symbolic glyptic function is a result of the identification of these items, and should not be taken as indicative of a sharp functional division between the two groups. Administrative sealing devices were identified by type (that is, sealings were allocated an administrative function). The symbolic seals were so identified by the context in which they were found. Thus, items found in a specific archaeological context were described as symbolic in function. This means that such a functional designation is reliant both on detailed information regarding the provenance of an item, and on an accurate reconstruction of this context. Thus, as is evidenced in the preceding survey of the archaeological and glyptic evidence from Elam, in the vast majority of cases (both glyptic items and sites), such information is lacking, and thus such functional allocation on the basis of context unattainable. This means that the following reconstruction is fragmentary and incomplete (and contributed to the large ‘unknown function’ group discussed below).

ECS Corpus, Volume I, Part III 682 Thus, only items with a definite and certain provenance, from an area with an established functional classification have been so described here. Two particular symbolic functional groups can be discerned on the basis of context, ‘votive’ and ‘funerary’. These two functions are related in that they are non- administrative, and symbolic in nature (and indeed, are not in fact ‘sealing’ related), but are in fact distinct, and probably enabled by separate, though related, facets of a seal. Both are however discerned on the basis of archaeological context, a temple or cultic installation in the first instance, and a grave or tomb context in the second.

6.2.1 Votive Votive defines an object that is donated or dedicated to a deity in accordance with a vow or promise. In glyptic terms, any seal that has been deposited in a temple or other cultic structure or premise is described as a ‘votive seal’ (Collon 2005: 131; the specific meaning of a seal deposited in a temple and therefore contained within a temple inventory is meant, seals that were dedicated or devoted to a god through an inscription, but continued to be used by individuals, alone or as a group representative, is not intended here, thus in a somewhat different definition to that adopted by Collon). The motivation and precise mechanism of this donation is currently unclear, though several reconstructive hypotheses may be proposed. The above defined description of a seal as a symbolic, or metaphysical, representation of the owner (Hallo 1987) (while the representation aspect of a seal holds true for seals that belonged to or were employed by both groups and individuals, presumably in this instance only seals belonging to individuals are relevant as temple donations) may have enabled the seal to act as the representative of the owner before the god in the temple, in a manner analogous to statues of worshipers also deposited in a temple (as in the Early Dynastic Diyala temples) (Collon 1990: 21; 1997: 19; Black & Green 1992: 300; Roaf 1990: 90; Hallo 1987). Thus by depositing a seal in a temple, the owner may have essentially deposited a representation of himself in the temple, and thereby may have been considered to be always before the god in supplication. As well as containing an element of the individual’s ‘self’, a seal was a valuable item to the owner, and thus may have been deemed worthy of donation to a deity (as it holds that an item of value and worth was deemed an appropriate and fitting gift or offering). This value is due both to the material from which the seal was carved, and the value of the item to the owner (detailed directly below). Thus, for

ECS Corpus, Volume I, Part III 683 example, a lapis lazuli seal was ‘valuable’ regardless of its function or inscription, due to its worth as a semi-precious stone, and for its apparently imbued ‘magical’ powers (Collon 2005: 100). Indeed, there is also evidence that seals were deposited as votive objects on the basis of the material value alone, aside from the glyptic or ‘magical’ qualities (thus the deposition of the so-called ‘Thebes Hoard’ in Boeotia, Greece, where these items were deposited in a temple because they were valuable as lapis lazuli items, regardless of their glyptic function [Porada 1981]). It is possible that some, or all, of the Surkh Dum-i-Luri seals were deposited as such, regardless of their glyptic function, as will be detailed below. A seal was valued by an individual as it presumably was required for, and enabled participation in, the administrative and economic life of the society. This importance is highlighted by textual sources that detail the implications of the loss of an individual’s seal (Collon 2005: 113 – 122). The possession of a seal may also have been a status symbol (Gorelick & Gwinnett 1990, Collon 1997: 17 – 19; 2005: 113 – 122). This importance of a seal to an individual paradoxically makes it an object of value that may have been deemed an appropriate or worthy gift to a deity, but also raises the question of why someone would so deposit, and thus relinquish, such an important object. This will be returned to below, but a solution to this dilemma may have been found at Choga Zanbil, in the form of replica or ‘cheap’ seals, made for the expressed purpose of votive deposition. Whatever the exact motivation for the deposition of a seal in a temple or religious structure, and indeed, if such an action can be accurately characterised as ‘votive’ (that is, in accordance with a vow or promise), or more simply as an offering, devotion or donation to a deity, the fact that seals were deposited in temples and cult structures in (a part of) Elam is of note. As will be demonstrated, this function may also be described as specific to a region of Elam, in a specific period, or at least may appear so on the basis of current evidence. This evidence does indeed indicate (though probably through the extraordinary, and perhaps not sample evidence from Choga Zanbil and Surkh Dum-i-Luri) a uniquely, or primarily ‘Elamite’ phenomenon. Certainly if not a solely Elamite function, then the ‘votive’ seal appears to be more common here than elsewhere1.

1There are relatively few cases of votive seals outside of Elam known by the current author. The few known examples include a hoard of seals in the Nabu temple at Assyrian Nimrud (Parker 1962), an enigmatic reference by Matthews (1990: 11) to a large amount of votive seals “in the west”

ECS Corpus, Volume I, Part III 684 Table and Graph 6.27 indicate the site distribution of ‘votive’ seals. As has already been described, only three sites within the Corpus contributed ‘votive’ seals, Susa, Choga Zanbil and Surkh Dum-i-Luri. Thus it may be stated that the ‘votive’ function of seals was restricted to a specific part of ‘Elam’ (as currently discernable on the available evidence), two neighbouring sites in Khuzistan (and more specifically the Susiana plain, Susa and Choga Zanbil), and an isolated site in Luristan (Surkh Dum-i-Luri). The thirty-six ‘votive’ seals from Susa were all provenanced from a single temple, the ‘Temple of Inshushinak’ on the Susian Acropole (see Chapter 2 and 5 for details). The ‘votive’ seals from Choga Zanbil were sourced from several temples, chapels and other cultic installations (see Chapter 2 and 5), most noticeably in two of the four ‘Chapels’ (Chapels III and IV) that lined the southwestern façade of the Ziggurat of Choga Zanbil. In these ‘chapels’ other votive gifts (or general donations or offerings) were also found (see Chapter 2 for details), thus assuring the depositional nature of these items. It should be remembered that other seal types were also found, or discerned, at Choga Zanbil however, including items allocated an ‘unknown function’ classification (as the nature of their context is uncertain), a single funerary item and the ‘palatial’ seals discussed below.

‘Votive seals’ Site Distribution Susa 36 Choga Zanbil 134 Surkh Dum-i-Luri 181 351 Table 6.27. Site distribution of ‘votive seals’.

All the glyptic material from Surkh Dum-i-Luri is classified here as ‘votive’, as all items were found within the ‘sanctuary’ walls (see Chapter 2, and Figure 2.12 for details). It should also be noted that this was the only area of Surkh Dum that was excavated, and thus any reconstruction on the overall glyptic function at the site cannot be made on the basis of this fragmentary evidence. If further investigation at Surkh Dum were to uncover an occupational settlement, complete with administrative artefacts (including seals and sealings) one may hypothesise that the sanctuary area, and the glyptic deposition in this installation, functioned in an analogous manner to Susa. If such installations were not forthcoming, a specific, and generally cultic function may be proposed for Surkh Dum, like at Choga Zanbil (for despite the fact

(presumably from context referring to Cypriot glyptic material, though unfortunately lacking full bibliographical details), and the already mentioned ‘Thebes Hoard’ from Greece (Porada 1981).

ECS Corpus, Volume I, Part III 685 that non-votive seals occurred, in limited numbers at Zanbil, no administrative devices were uncovered, and the site retains an essentially cultic function, not merely in terms of glyptic function).

200 180 160 140 120 100 80 60 40 20 0 Susa Choga Zanbil Surkh Dum-i-Luri Graph 6.27. Site distribution of ‘votive seals’.

Similarly, the identity of the individuals who deposited their seals at Surkh Dum is currently unknown, though the identity of these persons may impact on our understanding of the ‘votive’ glyptic function at this site. It may be assumed that ‘worshipers’, native or visiting Susa were responsible for the donation of the ‘votive’ seals in the Inshushinak Temple. It has been proposed that visiting ‘pilgrims’, apparently attending Choga Zanbil for some religious devotional purpose, deposited the seals in the chapels and temples at that site (Ghirshman 1968: 47; Porada 1970: 3). It may be, given the more recent discovery of ‘occupation’ areas at Choga Zanbil (see Chapter 2), that at least some of these items were in fact deposited by residents of the site however. As no information regarding the ‘residents’ or ‘worshipers’ of Surkh Dum-i-Luri is known (that is, whether the ‘sanctuary’ was a religious construction part of a larger site that also contained residential and other sectors, or whether it was an isolated sanctuary, visited, frequented and maintained by visiting ‘worshipers’, possibly seminomadic or transhumant people groups), no certain proposal regarding the mechanisms of Surkh Dum ‘votive’ offerings may be made. If the sanctuary was part of a wider occupied site, the ‘votive’ seals were presumably deposited in a manner analogous to Susa therefore. If the site was an isolated sanctuary or outpost, possibly related to a (semi)nomadic society, than the previous functional life of the deposited seals as administrative devices may be questioned (though the possibility that non-settled peoples also participated in a seal regulated control system, and thus

ECS Corpus, Volume I, Part III 686 possessed seals cannot be entirely dismissed, but seems less probable than glyptic use in a settled society), or rather, the association of the seals as a personal object of the ‘worshiper’ and thus the interpretative motivation for their deposition may be questioned. Thus, some of the Surkh Dum votive seals may have been deposited in the sanctuary due to their value as ‘heirloom’ or ancient objects, as curios, that depicted images, depictions and inscriptions that named gods, and due to the intrinsic value of the material from which they were cut, rather than as representations of the individual owner of the seal. The ‘heirloom’ (that is, seals that were not in a style deemed contemporaneous with the flourishing of the ‘sanctuary’) and non-local style items (including the Kassite and Middle and Neo-Assyrian contemporary seals not included in the Corpus) may belong particularly to this group of non-individually associated, and not so deposited, votive seals at Surkh Dum. The presence of a local, distinctive glyptic style (the LPS), that formed the majority of items from Surkh Dum (see Chapters 4 and 5, as demonstrated by Table 5.22 and Graph 5.18) may imply a local glyptic administration. However, the fact that this style was rarely, if ever, used in an actual administration associated sealing function (thus only two LPS sealings are known in the Corpus, an extremely limited, and indeed almost negligible sample, see Graphs 4.90, 4.91) may indicate that the LPS style was in fact not an administrative seal (and therefore truly a ‘seal’) style. Thus the LPS items at Surkh Dum may have been created for the intended purpose of temple deposition, or may be described more accurately as amulets or devotional objects rather than ‘seals’. Thus the discussion must again turn to the question of when an item, that apparently was not used to seal a clay device, ceases being adequately labelled a ‘seal’ and when an alternate description is required. As already mentioned, for the sake of completeness and due to the policy that it was better to err on the side of caution, all cylindrical shaped devices, with a central bore along the longitudinal axis, and a carved or incised image on the face was classified as a ‘cylinder seal’. The fact that some of these items labelled ‘seals’ were not such should be noted however. It is here proposed that, at least some, if not all the votive cylindrical items that were deposited in the Surkh Dum sanctuary, may not be truly considered seals, as they may never have been used, nor intended to be used as administrative sealing devices. This reinterpretation thus also changes at least part of the deposition motivation for these items, for it was the administrative function of the seal as an identifier and representation of an individual that led to the proposal that these devices

ECS Corpus, Volume I, Part III 687 were deposited in a manner analogous to the Early Dynastic temple statues, as a constant reminder and presence before a deity. The proposal that these items were a valuable and worthy gift of devotion as amulets or talismans holds however. The seals at Choga Zanbil that may have been specifically created for purposes of votive deposition may also belong in this category (see below for details of this phenomenon), though in this instance these items may have been intended as replica seals, or replacements in lieu of a real item, for it is known that the glyptic styles represented amongst the Choga Zanbil devoted corpus (most notably the LME items) were used as true seals, as sealings of these items are known (albeit in limited numbers, see Chapter 4, Graph 4.64). Table and Graph 6.28 illustrate the stylistic distribution of the Elamite ‘votive’ seals, and Table and Graph 6.29 the chronological distribution of these items. A cursory examination of these surveys may indicate that the votive function occurred throughout the chronological span of this study, and across all the stylistic divisions, except the earliest STS, and thus the Susa II period. This lack of Susa II period ‘votive’ seals may indeed simultaneously validate the suggestions both for the motivation for temple donation, and the above described different administrative glyptic function in the early period of seal use. If early seals were devices of information transference, and not therefore personal identification devices, the association of seal with self was not yet made, and thus the deposition of a seal in a temple as a personal representative before a deity not conceivable. The relatively small number of Susa III period seals may also indicate this, in this period too the seal was still an item of information transference primarily. The Susa III ‘votive seals’ may therefore be considered ‘heirloom’ devices, and indeed, in the instances of those from Choga Zanbil and Surkh Dum (see Chapter 5 for detail) this identification is assured. The lack of STS ‘votive’ seals also coincides with the general absence of seals of this style in the Corpus, as already addressed above (Chapter 4).

‘Votive seals’ Style Distribution JNRS 2 SF 7 PEU 8 LME 84 CPE 1 LSF 1 OBRS 16 AS 1 GS 1 ARS 3 PEO 12 LPS 120 AGD 1 PEA 8 EME 6 LGD 33 Unclassifiable 11 STF 1 UTRS 3 KRS 19 Not Illustrated 13 Table 6.28. Style distribution of ‘votive seals’.

ECS Corpus, Volume I, Part III 688 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

S S F E E S R PE G GD TF S SF RS EA EU EO M RS M A PS GD ble ted C S L A P TRS P BRS P K L L a ra JN A U O E L lust Il nclassifiot U N Graph 6.28. Style distribution of ‘votive seals’.

‘Votive seals’ Chronological Distribution Susa III 4 Ur III & Shimashki 11 Susa III/IV 1 Sukkalmah 28 Susa IV 8 Early Middle Elamite 6 Akkadian & Awan 11 Late Middle Elamite 224 Table 6.29. Chronological distribution of ‘votive seals’.

ECS Corpus, Volume I, Part III 689 230

220

210

200

190

180

170

160

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0

I h te II IV IV hki a I/ a wan s lma mi II a lamite a Sus a im E El Sus & A Sus Sh Sukka & iddle iddle adian II M M kk I A r ly te U ar a E L Graph 6.29. Chronological distribution of ‘votive seals’.

On closer examination, the reconstruction of universal chronological distribution across the entire time span of the study is not accurate. As already mentioned (Chapter 5), there is no definite stratigraphic or periodisation information

ECS Corpus, Volume I, Part III 690 for the Temple of Inshushinak at Susa (the source of thirty-six, or 10% of the ‘votive seals’), and thus the date of the material there contained, and whether these items were ‘heirlooms’ or contemporaneous with the period of occupation is also unknown. The thirty-six ‘votive’ seals from Susa are spread across the stylistic and chronological period of this study, from the JNRS (Susa III) to the LME (late Middle Elamite) styles, as demonstrated by Table 5.6. As there is no information regarding the chronology of this ‘Temple’ area, whether it was in use for the entirety of the period demonstrated by this chronological extension, or whether some were deposited as items contemporary with the period of use, and others as ‘heirloom’ items (as with Choga Zanbil and Surkh Dum-i-Luri), or indeed what period may be considered the original occupation, is unknown. If it is assumed that this occupation belongs to a single period, than it holds that this period would have to be the late Middle Elamite period, as the latest chronological item is an LME seal, and therefore late Middle Elamite in date. The other stylistic/chronological items may be considered ‘heirloom’ items therefore. If this reconstruction is accepted (and the extreme hypothetical nature of this reconstruction should be noted) then the Susa ‘votive’ deposits would be contemporaneous with the other ‘Elamite’ votive deposits (from Choga Zanbil and Surkh Dum-i-Luri). This reconstruction applies only if the Susa Inshushinak Temple was a single period use installation, if not, then it must be concluded that some of these items were deposited across other periods (and thus votive seal deposition was not a purely late Middle Elamite phenomenon, but merely apparently one on the basis of current evidence). As already detailed (see Chapter 2), both Choga Zanbil and Surkh Dum-i-Luri were founded in the (post-ME I) Middle Elamite period. Thus both sites can be considered late Middle Elamite occupations. The main flourishing of Choga Zanbil was limited primarily to the Middle Elamite period (and even more particularly the period around the reign of Untash-Napirisha), while the main period of the Surkh Dum-i-Luri patronisation was in the later, first millennium period following the late Middle Elamite foundation. The majority of the LPS items in the following survey come from Surkh Dum-i-Luri, thus conforming to this stylistic distribution. The LME and KRS items are primarily the Choga Zanbil devoted items, and thus conform to the chronological designation of this site. All the other items, from Zanbil and Surkh Dum may be considered heirloom items therefore. Table and Graph 6.30 demonstrate the ‘heirloom’ seals in opposition to contemporary seals (that is, those items that are

ECS Corpus, Volume I, Part III 691 stylistically contemporaneous with the use of the sites, and thus it is hypothesised were deposited during the main period of the use of the style), from Choga Zanbil and Surkh Dum-i-Luri. As is illustrated, contemporary styles are in the majority, though the presence of the ‘heirloom’ items in the temple deposits should be noted. As already mentioned above, those items deposited as heirlooms were presumably done so due to the value and worth of the stone upon which they were carved, and their value as antiquities, and not because they were representatives of the seal owner as defined above (though given the fact that seal ownership could apparently be transferred, as evidenced by seal re-use, some self association may be inferred for some ‘heirloom’ seals also). It should be noted that, as evidenced by the foundation date of Choga Zanbil and Surkh Dum, the ‘heirloom’ seals were already ‘ancient’ at the time of their deposition (that is, they did not ‘age’ within the temple, as part of an ancient temple inventory, as the Susa votive items may have been if this area had a long occupation). Thus the act of deposition occurred contemporaneously with the deposition of the main body of seals, and therefore this action can be demonstrated to be essentially confined to the late Middle Elamite period (at Choga Zanbil, and Surkh Dum-i-Luri [notwithstanding its continuation into the first millennium BC] at least). The precise mechanisms whereby these ‘heirloom’ seals were eventually in the possession of individuals who deposited them in ‘temples’ is unknown, though it may be assumed that at least some, if not all of these seals were originally true administrative seals.

‘Votive seals’ Heirloom Classification Heirloom 35 Contemporary 316 Table 6.30. ‘Heirloom’ classification of ‘votive seals’.

Heirloom

Contemporary

Graph 6.30. ‘Heirloom’ classification proportion of ‘votive seals’.

ECS Corpus, Volume I, Part III 692 As already discussed, the precise meaning and motivation for the deposition of a cylinder seal in a temple is still unknown, though several explanations may be proposed, including the value of the seal as a ‘magical’ or powerful stone and the worth of this material, its value as an important personal artefact, and its associated facet of individual identification and representation. The last two motivations, those that appear specific to cylinder seals, are somewhat paradoxical however. If a seal was of such worth and value to have been an appropriate votive gift or object of devotion due to its importance as an administrative and identifying device, the act of donation in a temple would be, in effect, an act of relinquishing the seal, which would have presumably, according to its value and worth, been a significant loss. Several explanations may be hypothesised, though neither is necessarily satisfactory in all cases, and thus it may be presumed that the ‘sacrifice’ of the seal, and the inherent inconvenience of this, was the crux of the action. Firstly, as already discussed, in the case of the Surkh Dum-i-Luri votive deposit the majority of these items (the LPS classified material especially) may never have been intended as true administrative seals, and thus the act of donating these items in a temple depository may not have taken an economic or administrative participation toll on the donator. Another possibility is that the donated seals were old seals in all cases, not just in the ‘heirloom’ seal instances. Thus, in the event that a seal that was no longer used by an individual or group, and needed to be (or was desired to be) replaced the old seal, no longer in use, may have been retired, and thus deposited in a temple. The circumstances for the replacement, whether due to the death or ‘retirement’ of the original owner, or for some other, unknown function, such as the assumption of a new (political) office or role, or the start of a new year or era (king’s reign), are unknown, but it is known that certain individual’s had two or more seals across their lifetime, indeed as evidenced by some of the ‘dated seals’ in the Corpus (for example the two seals of Epirmupi, 2248 and 2461), meaning that the replacement of an old seal with a new item was not unheard of. This would mean that the act of relinquishing the seal to the temple (or deity) would not have in fact harmed or cost anything to the donator. If this proposal is accepted, the reason why this practice was, apparently, limited generally to the late Middle Elamite period in ‘Elam’ must be questioned. This may be explained as an accident and coincidence of discovery creating this false impression, that is, the excavation of Choga Zanbil and Surkh Dum, two extraordinary sites in this regard, both coincidently late Middle

ECS Corpus, Volume I, Part III 693 Elamite in date may imply this impression, that with further excavations across the area (and indeed, a greater understanding of the contextual information at Susa) this picture may be altered and demonstrate a more universal distribution. The existence of the, admittedly extremely fragmentary and little understood Susa temple deposit and the fact that the Surkh Dum-i-Luri deposit continued into the first millennium BC (and thus into another period, beyond the late Middle Elamite period) both support the coincidental accident of discovery hypothesis. It may also be argued however that the temple ‘retirement’ process was indeed generally a late Middle Elamite phenomenon, and thus the information presented from Choga Zanbil and Surkh Dum may paint an accurate picture. A final explanation of the paradox of the relinquishment of a seal in a temple may be the proposal that the ‘votive seals’, especially those from Choga Zanbil, may have been in fact replica seals, created for the explicit purpose of temple donation. This may be implied at Choga Zanbil by the general uniformity of the LME, banquet scene, faience material items, a group that forms the majority of the deposited items (nearly 60% of the Zanbil votive seals). It may be hypothesised that a seal industry of faience replica ‘seal’ production existed at Choga Zanbil (possibly with the faience production area in the annex of the Temple of Ishni-qarab, see Chapter 2 above), and thus these ‘replicas’ may have been bought and so deposited by the visiting ‘pilgrims’ already discussed. This may have been a particular local solution, for those individuals that did not wish to actually forsake their seal, and it is probable that true, former administrative seals were also deposited at Choga Zanbil (and indeed, that a similar ‘replica’ solution may have been employed elsewhere). Thus ‘seals’, or more correctly devices or replica seals, manufactured for the purpose of deposition, and never intended as true seals may have been a solution to the deposition paradox at Choga Zanbil. The fact that these items belong to a true seal style (that is, LME items were used as administrative [sealing] seals, and thus can be considered a seal style) may be explained as the replica items were made in accordance to the prevailing seal style of the time. Thus, ‘votive’ seals appear to be a distinctly ‘Elamite’ phenomenon, restricted to three sites within western ‘Elam’. It also appears on the basis of current evidence that this phenomenon was restricted to a particular period of ‘Elamite’ history, the late Middle Elamite period, though given the fragmentary remains from Susa, and the fact that the Surkh Dum votive deposit continued into the first millennium, this may be

ECS Corpus, Volume I, Part III 694 more a reflection of available archaeological evidence than of an ancient reality. Votive seals may have been deposited due to their general value, as stones, as antiquities or as representatives and reflections of their owners. The Surkh Dum items may not have been truly seals, but amuletic devices, of a similar cylinder seal form, perhaps made to be so devoted (and intentionally designed to appear like cylinder seals), though further investigation in the Luristan region, and at Surkh Dum-i-Luri itself may provide evidence that seals were employed as administrative devices in the area, and thus alter this perception. At Zanbil, some of the objects placed in the Chapels and temples may have been replica or replacement seals, designed to save the devotee the problem of seal relinquishment, and thus may also never have been used as true administrative devices. Some of the ‘votive’ seals may be assumed to have been true (administrative) ‘seals’ before their deposition however. In these circumstances, the relinquishment of a seal in a temple may have been a form of seal retirement or disposal.

6.2.2 Funerary The other ‘symbolic’ seal function discerned in the Elamite Corpus is labelled ‘funerary’. This function is less problematic than the first, and, is defined by any item found in a tomb or grave context. Presumably these objects were deposited in such a context as a grave good, as an important personal artefact of the individual there buried. Therefore it may be assumed that these items at least, were not deposited in a temple upon the death of the owner, as described above, but rather buried with them, and thus taken out of circulation. The funerary function of seals is known from Mesopotamian and other contexts, most noticeably in the Early Dynastic ‘Royal Graves of Ur’. The precise function and motivation for the deposition of a seal in a funerary context (apart from simply the inclusion amongst other personal implements), such as may have pertained to beliefs regarding use of such items in an afterlife, or the burial of an item that ‘embodied’ an element of the person’s self so as to achieve total burial may be suggested, but cannot be described or identified in regards to our fragmentary information regarding ‘Elamite’ religion and beliefs. Table and Graph 6.31 illustrate the site distribution of the ‘Elamite’ funerary seals. As is demonstrated, funerary seals occur across ‘Elam’, in the three historical provinces. Two different types of sites can be discerned in this survey. One group contains pure funerary or graveyard sites, and include Bani Surmah, Kalleh Nisar and

ECS Corpus, Volume I, Part III 695 Tepe Djamshidi (see Chapter 2). All three of these sites are located in Luristan, and may possibly be identified as graveyards, possibly of (semi)nomadic or transhumant Luristan populations, and possibly not true, settled ‘Elamites’.

‘Funerary Seals’ Site Distribution Susa 14 Kalleh Nisar 7 Choga Zanbil 1 Tepe Djamshidi 2 Tepe Giyan 3 Bani Surmah 9 Tal-i Malyan 1 37 Table 6.31. Site distribution of ‘funerary seals’.

14 12 10 8 6 4 2 0 Susa Choga Bani Surmah Kalleh Nisar Tepe Tepe Giyan Tal-i Malyan Zanbil Djamshidi Graph 6.31. Site distribution of ‘funerary seals’.

The other site group that contributes ‘funerary’ seals include those in which the burials and tombs were amongst other areas and types of occupations. These sites include Susa, Choga Zanbil, Tepe Giyan and Tal-i Malyan, where the tombs and graves were included amongst other occupation and archaeological evidence. These graves may therefore be small representations among the total possible graves and tomb areas within the sites (particularly the large site of Susa). The relatively small number of graves that contained seals may either imply that many of the graves and tombs at such sites (particularly Susa), were missed or not recorded (or the seal provenance from these contexts not recorded), or that the practice of seal deposition in a grave or tomb was relatively rare. Table and Graph 6.32 illustrate the style distribution of the ‘funerary seals’, and Table and Graph 6.33 the chronological distribution of these items. As is demonstrated ‘funerary’ seals occur across the styles and the chronological periods under discussion here. No particular style or type can claim dominance in this reasonably limited ‘funerary’ corpus.

ECS Corpus, Volume I, Part III 696 ‘Funerary Seals’ Style Distribution JNRS 2 ARS 4 LME 2 CPE 1 UTRS 3 LPS 1 AGD 3 OBRS 5 LGD 1 STF 1 PEO 3 Miscellaneous 1 SF 7 KRS 1 Unclassifiable 2 Table 6.32. Style distribution of ‘funerary seals’.

7 6 5 4 3 2 1 0

S F S S us le R GD S RS PS o b CPE STF ARS BR PEO K LME L LGD JN A UTR O ifia lass scellane Mi Unc Graph 6.32. Style distribution of ‘funerary seals’.

‘Funerary Seals’ Chronological Distribution Susa III 3 Akkadian & Awan 4 Susa III/IV 1 Ur III & Shimashki 3 Sukkalmah 8 Susa IV 7 Late Middle Elamite 4 Table 6.33. Chronological distribution of ‘funerary seals’.

8 7 6 5 4 3 2 1 0 Susa III Susa III/IV Susa IV Akkadian & Ur III & Sukkalmah Late Middle Awan Shimaski Elamite Graph 6.33. Chronological distribution of ‘funerary seals’.

6.2.3 Summation of Symbolic Function As already mentioned, the relationship between the two symbolic glyptic functions is one of classification rather than function as it is in the ‘administrative’ functional divisions of writing and non-writing associated administrative items. There is therefore no real purpose to presenting combination summary figures here, as there is

ECS Corpus, Volume I, Part III 697 no real relationship between the two. Both functions are linked in that they are not administrative in final context, but probably, in their true form, were in fact administrative devices in their earlier, pre-depositional use.

6.3 Other Functions Two miscellaneous glyptic contexts, that occur at single sites (Choga Zanbil in the first instance, Chigha Sabz in the second) are classified under the rubric ‘other functions’. Both these functions, the ‘Palatial’ and the non-votive hoard, are specific to each site, and no sense of relationship between the two functions due to there classification under the one rubric here should be inferred. As already mentioned, glyptic function of cylinder seals is allocated primarily on the basis of context (thus the ‘votive’ seals were identified according to provenance in a temple or cultic installation, and ‘funerary’ due to tomb or grave context), and thus these two divisions have been defined due to the specific, and known context of these items. Both groups are relatively small however, and of little general importance to the overall glyptic functional discussion of the Elamite Corpus.

6.3.1 ‘Palace’ Seals As already mentioned, sixteen seals from Choga Zanbil may be grouped together as ‘palace’ or palatial seals. This allocation is made due to their provenance in the ‘palace’ area of the eastern sector of Choga Zanbil (see Chapter 2, Figure 2.8 and Chapter 5 for details). The precise function and definition of this area as ‘palatial’ is unclear, though some monumentality may be demonstrated (see Chapter 2), and some suggestion of occupation also (in the appearance of food preparation areas; Ghirshman 1968). Thus the exact function of the ‘palatial’ seals is also unknown. This group is only included here for the sake of completeness, as this corpus is essentially fragmentary and functionally elusive. Table and Graph 6.34 illustrate the style distribution of the sixteen ‘palace’ seals from Choga Zanbil. As is illustrated, and as further illustrates the general enigmatic, and essentially of little use, nature of these items, the majority classification is the ‘Not Illustrated’ group (meaning that these items were not illustrated with a published image, and so were not classified here). The fact that both ‘Unclassifiable Designs’ and LGD items contribute four items to this group, adds further to the enigmatic nature of the ‘palace’ seals. Thus only two seals, both of LME

ECS Corpus, Volume I, Part III 698 style, provide any chronological (late Middle Elamite) information. Both conform to the general chronological scheme for Choga Zanbil.

‘Palace seals’ style distribution LME 2 Unclassifiable 4 LGD 4 Not Illustrated 6 16 Table 6.34. Style distribution of ‘palace seals’.

6 5 4 3 2 1 0 LME LGD Unclass. Not Illustrated Graph 6.34. Style distribution of ‘palace seals’.

6.3.2 Non-Votive Hoard As already discussed (Chapters 2 and 5), all the Chogha Sabz glyptic items in the Corpus (and several others not included here as they are classified to non- Elamite/Iranian glyptic styles, see Chapter 2) were sourced from ‘Iron Age’ hoards dug into the earlier occupation levels of the site. The definition of these hoards as non-votive, as opposed to the votive hoards from Surkh Dum-i-Luri should be noted. The Chigha Sabz hoards were not associated with religious or cultic installations, nor did their deposition appear to be motivated by any devotion or gifting. Rather the Sabz hoards appear to be buried for simple storage, or more probably, safeguarding purposes. Table 5.28 and Graph 5.24 of Chapter 5 have already presented the stylistic distribution of the Chigha Sabz corpus, and as all nine Chigha Sabz seals are here classified in the non-votive hoard group, and thus the information presented would be the same, it would be redundant to repeat these figures here. Similarly, the general functional, stylistic and chronological conclusions already articulated (in Chapter 5) do not require repeating (see also Collon 2005: 135 – 137). Like the ‘palace’ seals of Choga Zanbil, this small group is included for the sake of completeness alone, and adds little of interest to the overall scheme.

ECS Corpus, Volume I, Part III 699 6.4 Unknown Functions As already discussed, due to the general fragmentary and unarticulated nature of much of the archaeological information included in this study (that is, the lack of accurate provenance information for much of the Susa material in particular, the general limited exposures at many sites, some of which were also poorly, or inexpertly excavated, according the era of their study [Tepe Sharafabad, Kamtarlan, Chigha Sabz, Godin Tepe, Tepe Giyan], or the limited publication regarding important periods of occupation [thus the Kaftari period occupation at Tal-i Malyan has received only limited, fragmentary treatment]), the majority of the seals of the Corpus, have no, or extremely limited, provenance and contextual information available (see Chapter 2 for details). As function of seals can only be accurately defined according to context, this means that for the majority of items in the Corpus, no functional allocation may be proposed. While it may be assumed, and probably generally accurately, that most of these seals had a standard, administrative, sealing function, in the event of no contextual information, this reconstruction is purely speculative and must remain unproven. While this material may have been classified as ‘administrative’ here, with a notation as to the hypothetical nature of this reconstruction due to limited information, it was deemed more prudent, to emphasise the lack of functional information for this material, to classify this material separately, with the notation that ‘administrative’ is the most probable true function. The possibility that other ‘votive’ and ‘funerary’ seals, and indeed, other functional divisions not already discerned in the Corpus, are amongst this corpus of ‘unknown function’ seals should also be noted. Unfortunately, as the major problem and cause of this lack of classification was essentially made at the time of excavation and initial recording, no amount of reanalysis or new study will solve this problem (as may for the ‘undifferentiated sealings’ discussed above). Reassessment and reanalysis reconstructions of site function based on excavated material in association with more detailed, controlled new excavations (as undertaken more recently at Susa [Carter 1980; Le Brun 1978; 1985]), may aid in some regard this discussion, though the extant to which such problems may be resolved is debatable. Table and Graph 6.35 illustrate the site distribution of the ‘unknown function’ seals, and further, Graph 6.36 demonstrates the site distribution of the ‘unknown function’ seals excluding Susa, to facilitate more accurate representation. As is shown, all sites of general multiple, or unknown function, that produced glyptic

ECS Corpus, Volume I, Part III 700 material, contributed ‘unknown function’ seals. Thus, only Surkh Dum-i-Luri (an exclusively ‘votive’ function site), Chigha Sabz (that only contributed non-votive hoard seals), Kalleh Nisar, Bani Surmah and Tepe Djamshidi (all solely ‘funerary’ sites) did not contribute items to this group. The general proportions correspond with those of the overall site corpora size, though the even more pronounced dominance of Susa, especially over the other large sites Choga Mish, Haft Tepe (both of which size is limited due to the fact that these sites contributed more sealings than seals, probably as in each case excavations primarily unearthed ‘archive’ or particularly sealing heavy [craft production] areas) and Choga Zanbil (the cultic context and function of which is generally well known, and thus the glyptic function types, as discerned from context may be accurately functionally defined) should be noted, and emphasises the general lack of contextual information for Susa.

‘Unknown Function’ Site Distribution Susa 1524 Deh-i Now 1 Chogha Mish 11 Kamtarlan 7 Haft Tepe 7 Godin Tepe 2 Choga Zanbil 25 Tepe Giyan 7 Tepe Sharafabad 2 Tal-i Malyan 17 1603 Table 6.35. Site distribution of the ‘unknown function’ seals.

Table 6.36 and Graph 6.37 illustrate the stylistic distribution of the ‘unknown function’ seals, and Table 6.37 and Graph 6.38 the chronological distribution. As one would expect of this general classification that forms the largest type of the Corpus, these patterns generally conform to the overall stylistic and chronological patterns of the Corpus. As this group is essentially negatively defined, and includes items that are, strictly speaking, not classified, no proportional information or pattern to this distribution may be commented upon.

‘Unknown Function’ Style Distribution STS 2 STF 135 PEU 143 LPS 50 JNRS 219 SF 86 OBRS 54 LGD 33 CPE 57 LSF 18 PEO 32 No Image 1 GS 184 ARS 128 EME 19 Miscellaneous 9 PEA 72 KRS 2 Unclassifiable 41 AGD 231 UTRS 58 LME 27 Not Illustrated 2 Table 6.36. Style distribution of the ‘unknown function’ seals.

ECS Corpus, Volume I, Part III 701 1550 1500 1450 1400 1350 1300 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0

e h p n an usa is epe nbil bad rlan y S T a fa a a M Z h aft a amt H eh-i Now K odin Te epe Giya hog hog D G T al-i Mal C C T epe Shara T Graph 6.35. Site distribution of the ‘unknown function’ seals.

ECS Corpus, Volume I, Part III 702 25

20

15

10

5

0

e h p n an is epe nbil bad rlan y T a fa a a M Z h aft a amt H eh-i Now K odin Te epe Giya hog hog D G T al-i Mal C C T epe Shara T Graph 6.36. Site distribution of the ‘unknown function’ seals, excluding Susa.

‘Unknown Function’ Chronological Distribution Susa II 2 Susa III 460 Susa III/IV 135 Susa IV 104 Akkadian & Awan 200 Ur III & Shimashki 201 Sukkalmah 86 Early Middle Elamite 19 Late Middle Elamite 79 Table 6.37. Chronological distribution of the ‘unknown function’ seals.

ECS Corpus, Volume I, Part III 703 240

230

220

210

200

190

180

170

160

150

140

130

120

110

100

90

80

70

60

50

40

30

20

10

0

S E S F F S U E le d G S EO PS te ST CP STF LS ARS PEA TR PE P EM KRSLME L LGD ab JNRS AGD U OBRS stra ssifi No Image Illu cellaneouscla t s o Mi Un N

Graph 6.37. Style distribution of the ‘unknown function’ seals.

ECS Corpus, Volume I, Part III 704 460 450 440 430 420 410 400 390 380 370 360 350 340 330 320 310 300 290 280 270 260 250 240 230 220 210 200 190 180 170 160 150 140 130 120 110 100 90 80 70 60 50 40 30 20 10 0

i II III n II/IV usa I Susa S Susa IV imashk Susa Sukkalmah

y Middle Elamite Akkadian & AwaUr III & Sh Earl Late Middle Elamite Graph 6.38. Chronological distribution of the ‘unknown function’ seals.

ECS Corpus, Volume I, Part III 705 6.5 Conclusion and Summation Four broad functional groups can be discerned in the Elamite Cylinder Seal Corpus; the ‘administrative’ function, the ‘symbolic’ function, the ‘other (or miscellaneous) function’ group, and the largest, and less satisfactory ‘unknown function’ category. The first category, the administrative function group, is defined by glyptic type, thus all sealings in the Corpus are deemed to have had an administrative function. The administrative function is thus the true, or standard glyptic function, as this is the act of ‘sealing’ by which these objects are defined. This function is described as ‘administrative’ as it regards the control or regulation of a society (and particularly the economic or trade aspect). This group can be further sub-divided into administrative items that are associated with writing (particularly sealed tablets and sealed clay envelopes) and those items that are not. The non-writing associated administrative devices may be further sub-divided into items that sealed mobile storage units (vessels, including jars, baskets, bales, and bullae) and immobile storage units (storerooms and areas, and thus wall and door sealings). It is here hypothesised that across the two main administrative divisions (writing and non-writing associated) the glyptic material functioned in a generally similar manner, with the major functional difference being chronological, rather than type. In the early (Susa II and Susa III periods) pre- and proto-literate periods, before written language was fully developed so as to convey abstract ideas and pieces of information (that is, when the system was essentially mnemonic, or basic in structure, and only detailed, for instance number and type, and not direction or action, thus verbs were not yet expressible) a seal impression on a jar, door, bullae or (numerical or ‘Proto-Elamite’) tablet conveyed some of the missing, or unarticulated information. Thus the seal impression may have identified the location, origin or destination of the item sealed, or the identity of the individual involved in the sealing (or sending) of the unit, or the identity (or location) of the intended recipient. Following the full development of a true writing system (in the Susa IV period and beyond), this sealing information transference was superseded by the writing system. Thus the action of sealing changed, to one of authorising, authenticating and identifying individuals involved in a transaction or contract (in the case of sealed tablets) or in the ‘sealing’ or deposition of objects in the case of non-writing associated administrative objects. The act of sealing, both tablets and non-writing devices, was reduced in the literate period, and the precise reasoning why some tablets, vessels and areas were sealed, and

ECS Corpus, Volume I, Part III 706 others not, remains elusive. This reconstruction is hypothetical, and is proposed as an explanation for the apparent proliferation of sealings in the earlier chronological periods, followed by a general decline in use that is coincident with the introduction/development of true writing. Further investigation, of the specific image types and glyptic styles, in combination with the philological evidence of the texts, of the sealing types (that is a physical study of the, mostly, non-writing associated sealings to ascertain sealing classification) and further, controlled and stratified (and thus more chronologically reliable) excavations are required for this hypothesis to be tested. Unlike the ‘administrative’ function, the two categories of the ‘symbolic’ group are not functionally linked on anything but the most superficial level. Both are defined by context (as are all the non-administrative functions), the ‘votive’ by provenance in a temple or religious/cultic installation, the ‘funerary’ by inclusion amongst the grave goods in a tomb or grave context. ‘Votive’ seals may not have been truly ‘votive’ in function (that is, donated in accordance with a vow or promise), but can be seen as generally gifts, offerings or donations made by a ‘worshiper’ and placed in a temple or similar structure, as other, non-glyptic items were (items such as statues, metal objects such as pins, jewellery and weaponry and beads). The act of donation of a seal may be due to the basic value of the material or antiquity of the seal, or may be linked in a metaphysical sense to the association of the seal with the individual or person of the owner (and thus a donated seal may have been a representative of the owner before the deity, as statues may have been). Some ‘seals’ deposited in temples, especially in this study, may have been deliberately made for this purpose, and never had a true administrative seal function. It may be said, on the basis of the general lack of administrative evidence from Surkh Dum and for the LPS style particularly, that those items (of LPS type) deposited in the Surkh Dum ‘sanctuary’ were not true seals, but talismans or amulets, designed to look like seals, but never intended for this purpose (further investigation at Surkh Dum specifically and Luristan generally, may provide evidence for an administrative glyptic system, in which case this proposal would require reassessment). Similarly, some of the ‘votive’ seals from Choga Zanbil (particularly the faience, LME banquet seals) may have been replica items, made specifically for visiting ‘pilgrims’ to be deposited in chapels, and thus also never intended as administrative devices. The fact that some ‘heirloom’ and other seals may have been true administrative devices prior to votive deposition is

ECS Corpus, Volume I, Part III 707 assured both by the wear and type of these items, and the fact that similar style devices (of the LME and other ‘heirloom’ styles represented at Choga Zanbil and Surkh Dum) were used to create sealings elsewhere (particularly at Susa). In this instance, the deposition of a seal in a temple may have been a method of retiring or disposing of a no longer needed or appropriate administrative seal, aside from the possible religious motivations. The other ‘symbolic’ function is that of ‘funerary’ deposition. This is a reasonably limited glyptic function in the Corpus, as discerned from the current evidence, that had a large chronological and geographical spread however. It can be assumed that all funerary seals were originally administrative items, possessed by the deceased, that were buried with them amongst the other personal effects come grave goods. Thus, it may be assumed that none of the funerary seals were created for this purpose as proposed for the ‘votive’ items, but were all administrative devices originally, and thus, functioning seals. The presence of ‘funerary’ seals at sites with no discernable occupation and no administrative system (thus Bani Surmah, Kalleh Nisar and Tepe Djamshidi) is interesting therefore. The Djamshidi evidence may be generally dismissed in this regard, as the extremely limited excavations at that site, and their general poor quality, prevent any assumption or conclusion as to the real function of the site, indeed whether it was truly a graveyard site only, or contained occupational levels too is unknown (see Chapter 2). Bani Surmah and Kalleh Nisar can be demonstrated to be essentially funerary in overall function (the limited occupation at Kalleh Nisar notwithstanding, as this evidence appears unrelated to the grave sites, and indeed, bore no administrative evidence) however, and thus the presence of administrative devices in some of the tombs is of note. It has been proposed that these sites were the burial areas of (semi)nomadic or transhumant groups. As already mentioned, the participation in a glyptic administrative system by such peoples seems incongruous, but not impossible. If these sites are indeed the burial sites of non-settled peoples, then the presence of some seals, presumably originally of administrative function, would indicate that some of these individuals were indeed involved in such a system of regulation (indeed it is not inconceivable that these people used the seals to seal or mark goods traded with the settled populations). Of course, the buriers at Bani Surmah and Kalleh Nisar may have been settled peoples from other areas or regions, who chose to bury their dead at these particular sites for reasons known unto themselves, and thus their use of sealing

ECS Corpus, Volume I, Part III 708 devices should not astound. A final explanation may be that these seals were not the actual possessions of the buried individuals originally, but were kept or ‘treasured’ as heirlooms or curios, and similarly so deposited amongst the grave goods. Another two seal function types, defined according to context, are the ‘palace’ and non-votive hoard groups. Both these functions were limited to a single site each (Choga Zanbil and Chigha Sabz respectively) and offer little insight or interesting information to the overall glyptic function reconstruction of the society. The final, and unfortunately for purposes of classification, the largest, functional category (as evidenced by Table 6.38 and Graph 6.39) is the ‘unknown function’ group. It may be assumed that most of these items were administrative in general function, though lacking accurate contextual information, such a reconstruction is untestable and unproven. Reanalysis and reconstructive studies aimed at discerning the function and restoring missing contextual and (site) functional information may result in an improvement on this situation, though a general resolution to better detail and record seal provenance in future excavations may provide the greatest source of controlled information in this regard.

Distribution of Functional Types Administrative 1581 Funerary 37 Non-writing 1214 Palace 16 Writing 367 Non-votive hoard 9 Votive 351 Unknown Function 1603 3597 Table 6.38. Survey of functional classification types of the Elamite Cylinder Seal Corpus.

Thus it may said, especially if one excepts that most of the ‘unknown function’ seals had a general administrative function, that general sealing as a method of control and regulation was the most common, and standard Elamite glyptic function. This indeed, confirms the very definition of these objects as ‘seals’. As is illustrated by Graph 6.40, non-writing administrative seals are more dominate than writing associated devices, thus demonstrating that in Elam, during the period under investigation here at least, that such glyptic function was primarily not associated with tablets and the written record. This may conform to the early impression of Elam, as characterised by Porada, as an essentially oral based economic society, and not tablet and written record based to the degree that neighbouring Mesopotamia was (Porada 1965: 46). Graphs 6.39 and 6.40, and Table 6.38 also illustrate that ‘votive’ was the most common of the secondary or symbolic functional classes, though the evidence in

ECS Corpus, Volume I, Part III 709 this regard may be skewed by the extraordinary contributions of Choga Zanbil and Surkh Dum-i-Luri. This secondary function is of note however, and may be, as presented by the current evidence, a particularly ‘Elamite’ functional class.

1650 1600 1550 1500 1450 1400 1350 1300 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0 Administrative Votive Funerary Other Unknown Function

Graph 6.39. Survey of functional classification types of the Elamite Cylinder Seal Corpus.

ECS Corpus, Volume I, Part III 710 1650 1600 1550 1500 1450 1400 1350 1300 1250 1200 1150 1100 1050 1000 950 900 850 800 750 700 650 600 550 500 450 400 350 300 250 200 150 100 50 0 Non- Writing Votive Funerary Palace Non-votive Unknown writing Hoard Function

Graph 6.40. Survey of functional classification types of the Elamite Cylinder Seal Corpus, with division of the functional classes.

ECS Corpus, Volume I, Part III 711 Chapter 7 – Elamite and Mesopotamian Contact and Interaction The preceding chapters have detailed the stylistic and chronological development of the glyptic items of Elam, as well as the functional classification of this material. As has been made abundantly clear, and as presented in detail, throughout much of the period under discussion here ‘Elam’ (as here defined) was engaged in varying types of interactive patterns with neighbouring Mesopotamia. Indeed, both the general chronological and archaeological scheme of Elamite archaeology as presented here, and many of the styles here articulated, were defined in reference to, and association with, Mesopotamian evidence. This is a symptom of the above outlined dominance of Mesopotamia in the available literature, in relation to the dearth of available information from Elam. However, there can also be no denial of the clear and present chronological, political, social and stylistic synchronisms between Elam and Mesopotamia. Thus, while the reliance on Mesopotamian-orientated and focused evidence may over represent or exaggerate these synchronisms, there can be no doubt as to their existence. There is no need to reiterate or repeat the political and historical Elamite- Mesopotamian synchronisms already detailed above. The reader is directed to the relevant section of Chapter 2, and especially Tables 2.1 – 2.3, 2.7 and 2.34 for details. This Chapter will demonstrate several facets or manifestations of this interaction as pertains to cylinder seals. It will thus be shown that as well as in political and historical spheres (such as kingly interactions and wars) and other areas of material culture (such as ceramics, language [particularly as demonstrated by written scripts], statuary and other forms), there was a glyptic interaction between Elam and Mesopotamia. The first section of this chapter will return to the ‘dated seals’ and examine the evidence there contained for this interaction. The following section, through the use of the ‘dated seals’ evidence and the style classification paradigm defined above (Chapter 4), will assess the chronological, geographical and directional facets of Elamite and Mesopotamian glyptic interaction and association. The final section of this chapter will then conclude and assess this material, with particular reference to several paradigms that have been previously articulated to explain the constitution of Elam, with reference to the patterns of Elamite-Mesopotamian interaction (particularly Amiet’s ‘ethnic duality’ thesis [Amiet 1979a; 1979b], de Miroschedji’s cyclic pattern of expansion and collapse [Miroschedji 2003] and the various ‘Uruk Expansion’ models [Algaze 1993]).

ECS Corpus, Volume I, Part III 712 As already mentioned, the area of Mesopotamia that chiefly concerns this study is more specifically the southern region, variously known as , and over the span of this study. The catchall ‘Mesopotamia’ is used here as a shorthand version to represent this area for ease of reference. Elamite-Mesopotamian (historical, political, and cultural, including in some glyptic areas as already reported in Chapter 4) interactions occurred with northern Mesopotamia (, and indeed, at times Syrian Mesopotamia) as well, albeit to lesser degrees as with the south. These interactions will not chiefly concern us here however, both as the chief contributor to the pattern of interaction was the south, and as a limit or parameter was placed here for the sake of brevity. The Elamite-Assyrian (northern Mesopotamian) interaction patterns must await future study therefore.

7.1 ‘Dated Seals’ The function and definition of ‘dated seals’ has already been addressed in full (see Chapter 3.4), and these items have already been used as chronological indicators in the creation and articulation of some of the Elamite styles. Of the forty-six Elamite Corpus ‘dated seals’ (including the ‘potential’ examples, see Chapter 3), twenty-nine of these items bear the name of an ‘Elamite’ king or official that potentially has a Mesopotamian synchronistic king, as demonstrated by Table 7.1 (and already noted by Table 3.8). Also, three seals from Susa, already discussed above, bear the names of Mesopotamian kings themselves, and thus cannot be ignored in this discussion of Mesopotamian-Elamite glyptic influence (2701 names Rim-Sin, 2723 and 3465 Shu- Sin). Furthermore, the unprovenanced seal in the Gulbenkian Museum, Durham that names Ebarti I (Lambert 1979: seal 42), must be considered here as this ‘Elamite’ has known synchronisms with three Ur III rulers (, Amar-Sin and Shu-Sin).

Item Elamite King/ Mesopotamian Item Elamite Mesopotamian Official Synchronism King/ Synchronism Official 2151 Eshpum * (1) 2442 Idaddu I Ishbi-Erra* (13; 2) 2213 Eshpum Manishtushu* (1) Idaddu II – 2248 Epirmupi Naram-Sin* (14) 2446 Idaddu I Ishbi-Erra* (13; 2) 2346 Tan-Ruhurater Shu-ilishu* (6) Idaddu II – 2353 Attahushu * (3) 2455 Idaddu I Ishbi-Erra* (13; 2) Sumuabum* (1) 2359 Puzur- Ur-Nammu (2; 1) Idaddu II – Inshushinak

ECS Corpus, Volume I, Part III 713 Item Elamite King/ Mesopotamian Item Elamite Mesopotamian Official Synchronism King/ Synchronism Official 2371 Queen Shu-ilishu* (5; 1) 2456 Imazu/ Ibbi-Sin* (23; 15) Mekubi/ Kindattu Ishbi-Erra* (13; 2) Tan-Ruhurater 2372 Ebarti I Shulgi* (41; 22) 2461 Epirmupi Naram-Sin (18) Amar-Sin* (8; 4) Shu-Sin* (26; 16) Ebarti II – 2912 Tepti-ahar Kadashman-Enlil I* (2)/Kadashman-Harbe I 2374 Attahushu Gungunum* (3) 2974 Kuk-Nashur Ammisaduqa* (15; 1) II 2389 Idaddu I Ishbi-Erra* (13; 2) 2977 Attahushu Gungunum* (3) Idaddu II – 2980 Kuk-Nashur Gungunum* (3) I 2390 Idaddu I Ishbi-Erra* (13; 2) Kuk-Nashur Ammisaduqa* (15) II Idaddu II – Kuk-Nashur – III 2418 Tan-Ruhurater Shu-ilishu* (5; 1) 2981 Kuk-Nashur Gungunum* (3) I 2419 Attahushu Gungunum* (3) Kuk-Nashur Ammisaduqa* (15; 1) II 2421 Ebarti I Shulgi* (41; 22) Kuk-Nashur – Amar-Sin* (8; 4) III Shu-Sin* (26; 16) Ebarti II – 2982 Tepti-ahar Kadashman-Enlil I* (2)/Kadashman-Harbe I 2437 Idaddu I Ishbi-Erra* (13; 2) 2984 Tepti-ahar Kadashman-Enlil I* (2)/Kadashman-Harbe I Idaddu II – 2985 Tepti-ahar Kadashman-Enlil I* (2)/Kadashman-Harbe I 3454 Idaddu I Ishbi-Erra* (13; 2) Idaddu II – Table 7.1. Elamite ‘dated seals’ naming Elamite-Mesopotamian synchronous kings, with reference to the identity of the Mesopotamian synchronous king. The Mesopotamian kings marked with an asterix(*) themselves possess ‘dated seals’. The number in parentheses refers to the known number of such seals, the second number, where applicable, the actual illustrated examples.

The Mesopotamian-Elamite synchronisms and the ‘dated seals’ have already been addressed as to their worth as stylistic dating mechanism. What is however of greater significance here is the fact that many of the Mesopotamian kings synchronised with Elamites (who possess ‘dated seals’), themselves also possessed ‘dated seals’. Those Mesopotamian kings with Elamite synchronisms and ‘dated seals’ of their own are marked in Table 7.1 by an asterix, with the number of known

ECS Corpus, Volume I, Part III 714 ‘dated seals’ in parentheses1. Thus, all the Mesopotamian kings synchronised with ‘dated seal’ Elamite kings themselves possessed at least one, if not more ‘dated seals’ (the only exception being Kadashman-Harbe I, who has no known ‘dated seal’, but who has only a possible synchronism with Tepti-ahar, see Chapter 2 for details). Indeed, as one would expect given the great number of Mesopotamian ‘dated seals’, other Mesopotamian kings with Elamite synchronisms also possessed ‘dated seals’ (but have not been included in this survey as their ‘Elamite’ counterpart has no known ‘dated seal’). These figures include two seals that name Sargon of Agade (associated with the ‘Elamite’/Iranian officials Sanam-shumut, Hishep-ratep and Luh-ishan; see Chapter 2 for details of this and all such synchronisms), eighteen seals that name of (associated with Siwe-palar-huppak and Kudu-zulush I), and the (or II; the first associated with Pahir-ishshan, Igi-halki and Hurbatila) and Burna-Buriash II (associated with Untash-Napirisha; some of these items may be associated with Burna-Buriash I however) produced fourteen and five ‘dated seals’ respectively2. These items are of interest, as one may infer that some of these seals, and therefore their styles, would have been known just as the officials who used them may have been known in ‘Elam’. However, as no directly contemporaneous Elamite seals can be found (that is, ‘dated seals’ belonging to the

1It should be recognised that as Mesopotamian glyptic styles and materials were not a focus of this particular study, the recognition and discovery of Mesopotamian ‘dated seals’ was not a primary concern or focus, as time was devoted to the study and detection of Elamite items. A preliminary search of available Mesopotamian glyptic sources, for the purposes of comparison yielded over four hundred ‘dated seals’ that named Mesopotamian kings alone (that is, not including other known officials, and, even more strikingly, not including northern Mesopotamia/Assyria, in any concerted effort). This is thus significantly more than the less than fifty ‘Elamite’ examples. This may be explained by the more extensive excavation and study of Mesopotamia, by the wider geographical area characterised as such, by the more extensive dynasties of this region, often with simultaneous kings (such as in the - era of the Old Babylonian period), but also to the ancient reality that Mesopotamia produced more glyptic material than Elam. The Mesopotamian ‘dated seals’ may be used similarly as the Elamite examples were here, as anchors or testing points of glyptic styles, and evidence for absolute dates. While this has been undertaken for specific kings and styles (most notably by Matthews for the Kassite styles [Matthews 1990; 1992]), no systematic nor thorough study devoted to these items has been undertaken. This is an interesting field for future study. Only the Mesopotamian ‘dated seals’ that are related to Elamite counterparts will be addressed here however; the remaining items must await further study and analysis in the future. 2Shamshi-Adad of Assyria, who is synchronised with the ‘Elamite’ Shiruk-tuh has at least six known ‘dated seals’ (including CCO I: T.188, T.216, Ward 1920: 33; 1.7.2002, 1.7.2003; RIME 4: 1.7.2005), and Zimri-Lim of Mari (associated with Kudu-zulush I) has three (RIME 4: 1.7.2003, 1.7.2006, 1.9.2001, Legrain 1925: 268), though it should be noted that the glyptic material from these areas were not studied extensively and thus more such items may be available. Similarly, as these areas were not a primary focus, and do not strictly pertain to the discussion here, that is, this study is primarily discussing the interaction between southern Mesopotamia (variously Sumer, Akkad and Babylonia) and Elam, these items and kingly associations will not be included in this discussion, but should be noted.

ECS Corpus, Volume I, Part III 715 Elamite kings or officials with whom theses Mesopotamian figures are synchronised), these items may not be used as direct evidence for a comparative Elamite- Mesopotamian glyptic study. These seals are presented in Appendix A for reference, and the evidence they provide for the chronological markers of style should be noted (thus the evidence for Akkadian contest scenes on Sargon’s ‘dated seal’ M155 relates to the contest scenes of ARS and demonstrates this influence; similarly Hammurabi’s seals M156 – M158 illustrate the Old Babylonian standing ‘presentation’ scene, linked with the same depictions on OBRS items; and the First Kassite elongated standing figure on Burnaburiash’s M174 evidences the Kassite elongated style and representation known from the LME and KRS styles). The evidence of these items confirms the chronology of the styles, especially the ‘Mesopotamian Related’ styles, articulated in Chapter 4, and provides comparative data for the Mesopotamian- Elamite interaction. Table 7.2 presents the Elamite-Mesopotamian synchronised seals (that is, those seals that are possessed by kings with known synchronistic reigns or historical interaction events). Many of the Mesopotamian ‘dated seals’ were sourced from philological and textual sources, with no reproduced seal image (in accordance with common practice for such texts as already outlined above), these not illustrated Mesopotamian ‘dated seals’ have not been included in the synchronous Table 7.2 but are detailed in Table 7.3 for reference. A rectifying study and illustration of these items would aid in the current reconstruction. Among the not illustrated seals are all three known seals of Gungunum, thus this king, despite synchronisms with Attahushu and Kuk-Nashur I (both of whom possess, possible, ‘dated seals’), is not listed in Table 7.2 despite the listing in Table 7.1. Where applicable, the second number in parentheses in Table 7.1 indicates the actual number of illustrated (and therefore included in Table 7.2) ‘dated seals’, as opposed to the actual known or detailed amount represented by the first number. Where a second number does not occur, one can assume that all known ‘dated seals’ are illustrated (and thus included).

ECS Corpus, Volume I, Part III 716 Mesopotamia Elam Manishtushu Eshpum M1 2151 ARS (4)

2213 ARS (10)

Naram-Sin Epirmupi M2 2248 ARS (10)

M3 2461 PEU (1)

M4

M5

M6

ECS Corpus, Volume I, Part III 717 Mesopotamia Elam Naram-Sin (cont.) Epirmupi (cont.) M7

M8

M9

M10

M11

M12

M13

ECS Corpus, Volume I, Part III 718 Mesopotamia Elam Naram-Sin (cont.) Epirmupi (cont.) M14

M15

Ur-Nammu Puzur-Inshushinak M16 2359 UTRS (3)

Shulgi Ebarti I M18 2372 UTRS (3)

M19 2421 UTRS (4)

ECS Corpus, Volume I, Part III 719 Mesopotamia Elam Shulgi (cont.) Ebarti I (cont.) M20 lam: 42 PEU (7)

M21

M22

M23

M24

M25

M26

ECS Corpus, Volume I, Part III 720 Mesopotamia Elam Shulgi (cont.) Ebarti I (cont.) M27

M28

M29

M30

M31

M32

ECS Corpus, Volume I, Part III 721 Mesopotamia Elam Shulgi (cont.) Ebarti I (cont.) M33

M34

M35

M36

M37

M38

M39

ECS Corpus, Volume I, Part III 722 Mesopotamia Elam Shulgi (cont.) Ebarti I (cont.) M40

M41

M42

M43

M44

M45

Amar-Sin Ebarti I M58 see the Ebarti seals above

ECS Corpus, Volume I, Part III 723 Mesopotamia Elam Amar-Sin (cont.) Ebarti I (cont.) M59

M60

M61

Shu-Sin Ebarti I M66 see the Ebarti seals above

M67

ECS Corpus, Volume I, Part III 724 Mesopotamia Elam Shu-Sin (cont.) Ebarti I (cont.) M68

M69

M70

M71

M72

M73

M74

ECS Corpus, Volume I, Part III 725 Mesopotamia Elam Shu-Sin (cont.) Ebarti I (cont.) M75

M76

M77

M78

M79

M80

M81

ECS Corpus, Volume I, Part III 726 Mesopotamia Elam Ibbi-Sin Kindattu M92 2456 UTRS (5)

M93

M94

M95

M96

M97

M98

ECS Corpus, Volume I, Part III 727 Mesopotamia Elam Ibbi-Sin (cont.) Kindattu (cont.) M99

M100

M101

M102

M103

M104

M105

M106

ECS Corpus, Volume I, Part III 728 Mesopotamia Elam Ishbi-Erra Kindattu M115 see the Kindattu seal above

M116

Ishbi-Erra Idaddu I see the Ishbi-Erra seals above 2389 UTRS (4)

2390 UTRS (4)

2437 UTRS (4)

2442 UTRS (4)

2446 UTRS (4)

ECS Corpus, Volume I, Part III 729 Mesopotamia Elam Ishbi-Erra (cont.) Idaddu I (cont.) 2455 UTRS (4)

3454 No Image

Shu-ilishu Tan-Ruhurater I M128 2346 UTRS (2)

2371 UTRS (2)

2418 UTRS (4)

Sumuabum Attahushu M136 2353

2374 UTRS (3)

ECS Corpus, Volume I, Part III 730 Mesopotamia Elam Sumuabum (cont.) Attahushu (cont.) 2419 UTRS (4)

2977 EME (8)

Ammisaduqa Kuk-Nashur II M137 2974 EME (8)

2980 EME (8)

2981 EME (8)

Kadashman-Enlil I Tepti-ahar M152 2912 EME (6)

M153 2982 EME (8)

ECS Corpus, Volume I, Part III 731 Mesopotamia Elam Kadashman-Enlil I (cont.) Tepti-ahar (cont.) 2984 EME (8)

2985 EME (8)

Table 7.2. Elamite-Mesopotamian synchronised ‘dated seals’, see Appendix A for details of the Mesopotamian items.

Not Illustrated Mesopotam ian ‘Dated Seals’ Ur-Namm u M55 M84 M109 M21 M132 M142 M17 M56 M85 M110 M22 Gungunum M143 M46 M57 M86 M111 M23 M133 M144 M47 Amar-Sin M87 M112 M24 M134 M145 M48 M62 M88 M113 M25 M135 M146 M49 M63 M89 M114 M26 Ammisaduqa M147 M50 M64 M90 Ishbi- M27 M138 M148 Erra M51 M65 M91 M17 Shu-ilishu M139 M149 M52 Shu-Sin Ibbi-Sin M18 M129 M140 M150 M53 M82 M107 M19 M130 M141 M151 M54 M83 M108 M20 M131 Table 7.3. List of ‘Not Illustrated’ Mesopotamian ‘dated seals’, see Appendix A for details of the Mesopotamian items.

As with those items in Appendix A, the Mesopotamian ‘dated seals’ presented in Table 7.2 demonstrate the general chronological and stylistic developments outlined in Chapter 4, and provide comparative evidence for the Elamite ‘Mesopotamian Related’ styles therefore. Thus the ‘contest’, ‘presentation before a deity’ and ‘mythological’ scenes demonstrated in the ‘dated seals’ of the Akkadian kings (in this instance, those of Manishtushu and Naram-Sin) confirm the Mesopotamian influenced nature of the similar ARS scenes. Similarly, the ‘presentation scenes’ (including the ‘king and cup’ motif) of the Ur III period kings (Ur-Nammu, Shulgi, Amar-Sin, Shu-Sin and Ibbi-Sin) ‘dated seals’ demonstrate the

ECS Corpus, Volume I, Part III 732 comparisons and chronological placement for these similar scene types in the UTRS style particularly. The standing ‘presentation’ scenes of the Old Babylonian ‘dated seals’ (those of Sumuabum and Ammisaduqa) also demonstrate the contemporaneous nature and interaction between this Mesopotamian scene-type and the similar OBRS scenes. In some instances, the Elamite and Mesopotamian synchronised ‘dated seals’ are extremely similar, thus indicating a seemingly high degree of shared glyptic interaction and style. Thus, the ARS ‘contest’ scene of Epirmupi’s ‘dated seal’ 2248 is highly reminiscent of a number of Akkadian ‘contest’ scenes of his contemporary Naram-Sin (M2, M5 – M7, M12, M14, M15). In form, Puzur-Inshushinak’s UTRS 2359 is similar to the illustrated Ur-Nammu ‘dated seal’ (M16), as both demonstrate ‘presentation’ scenes, though the identification of the presentee in Ur-Nammu’s instance as a king (note the ‘kingly cap’), while on 2359 the figure wears the horned headdress of divinity indicates the differences that may occur in such comparenda. All three possible contemporaries of Ebarti (I; Shulgi, Amar-Sin and Shu-Sin) also demonstrate a similarity in form to Ebarti’s UTRS (3) ‘dated seal’ 2372 (Shulgi’s M18, M19, M25 – M37, M40 – M45; Amar-Sin’s M58, M61; Shu-Sin’s M66 – M81; the high probability that M78 is a modern forgery should be noted [Fischer 1999]), though the more cursory and lower relief (as may be ascertained from a sealing, see the Catalogue entry for a photograph) of Ebarti’s seal indicates a certain Elamite element to the design, in comparison to the rounded, high relief execution of the Mesopotamian items. Similarly, the non-standard ‘presentation’ scenes of Ishbi-Erra’s ‘dated seals’ (M115, M116), as is typical of Isin-Larsa Old Babylonian designs (though the fragmentary nature of these scenes should be noted), can also be closely associated with the seals of his contemporary Idaddu (2389, 2390, 2437, 2442, 2446, 2455). Finally, the seated single attendant ‘presentation’ scenes, with a long inscription, depicted on the Kassite seals of Kadashman-Enlil (M152, M153) are markedly similar to the first ‘dated seal’ of Tepti-ahar of EME (6) style (2912), and related in form, if not true subject matter, to the ‘serpent throne’ seals of this king (2982, 2984, 2985). Thus, the general comparative and chronological information provided by the Mesopotamian-Elamite synchronised ‘dated seals’ confirms and evidences the paradigms constructed here, through general chronological similarities in designs and themes, and through specific examples of strikingly similar designs. Other aspects of

ECS Corpus, Volume I, Part III 733 the Elamite stylistic paradigm as articulated here, and its interaction with neighbouring Mesopotamian styles can also be demonstrated by the synchronised ‘dated seals’, though not in their similarities, but in their differences. Thus the offering element in the seated ‘presentation’ scene of Ebarti’s seal 2421 does not occur in any of the possible contemporaneous Mesopotamian seals (of Shulgi, Amar- Sin and Shu-Sin), thus indicating the above noted Elamite nature of this scene type. The ‘mountain man’ motif on the unprovenanced Ebarti seal (Lambert 1979: 42) of PEU (7) style also demonstrates a distinct motif not replicated in any of the Mesopotamian contemporaneous examples. Similarly, the standing ‘presentation’ scene with an axe or standard being gifted on Kindattu’s 2456, also described as an Elamite scene, is not replicated in any of the contemporary seals of Ibbi-Sin or Ishbi- Erra. Finally, the ‘serpent throne’ motif, also described as Elamite (and indeed, specifically royal), as depicted on the ‘dated seals’ of Attahushu (2977), Kuk-Nashur (2974, 2980, 2981) and Tepti-ahar (2982, 2984, 2985; as well as others not synchronised with a Mesopotamian king, see Chapters 3 and 4) is not replicated in any of the contemporaneous Mesopotamian ‘dated seals’ (those of Sumuabum, Ammisaduqa and Kadashman-Enlil, though the structural similarity between the Kadashman-Enlil designs and this scene type already mentioned should be recalled), also confirming its Elamite nature. The fact that these designs are not represented in the other Mesopotamian ‘dated seals’ of Appendix A, those that belonged to Mesopotamian kings synchronised with Elamite kings without their own ‘dated seals’ (and indeed, in other known Mesopotamian items generally, except perhaps for some presence of the ‘offering scene’ [Collon 2005: seals 1, 170]), also implies the Elamite nature of these design elements. From this survey it may be concluded therefore that some design elements were particularly Elamite or Iranian, and were not copied or influenced by Mesopotamian glyptic. This may imply a uni-directional influence, that is, that Mesopotamian glyptic conventions and design types (such as the ‘king and cup’ motif, the ‘contest’ scene and the manner of their depiction) influenced Elamite styles and artisans, while Elamite elements and designs did not influence the contemporary Mesopotamian styles in any significant manner. This possibility will be addressed again shortly, following the examination of several other pieces of information. Unfortunately, the relatively limited extent of Elamite-Mesopotamian synchronous ‘dated seals’ prevents further use of this glyptic classification as a test to

ECS Corpus, Volume I, Part III 734 the extent and duration of Elamite-Mesopotamian (glyptic) interaction. In order for this material to be of greater use in this regard a more direct, continuous chronological development is required (indeed, in this event as is a more accurate and clear understanding of the order of succussion of the Elamite dynasties). That is, more direct evidence for the era of introduction (or indeed, the retirement) of a form or design element, and a greater understanding of the chronology of this may only be ascertained if a general sequence of uninterrupted (or little interrupted) succession of Elamite-Mesopotamian ‘dated seals’ were available. Indeed, a more thorough representation of Elamite ‘dated seals’ would aid in this discussion. Another limitation to this survey is the complication of synonymously named kings, where there is a general difficulty in attributing a seal to one of two (or more) like named kings, especially, as is the case in this study, as only one of these identities possesses a Mesopotamian synchronism. Thus Ebarti I may be synchronised with Shulgi, Amar-Sin and Shu-Sin, though any number of the Ebarti ‘dated seals’ may in fact not refer to this individual, but the second Ebarti, and thus the Mesopotamian- Elamite synchronism with these seals is non-existent. Similarly the ‘dated seals’ that name Idaddu may belong to either of the so named kings, though only those belonging to the first Idaddu are relevant as synchronous ‘dated seals’ with those of Ishbi-Erra. Also, the fact that three kings named Kuk-Nashur reigned in ‘Elam’, but only one of which (Kuk-Nashur II) is associated with Ammisaduqa, though some of these seals may refer to other Kuk-Nashurs (due to the filiation supplied on Kuk- Nashur ‘dated seal’ 2974, Kuk-Nashur II may be assumed; in a recent article Mofidi Nasrabadi [2008: 40 – 41] has associated both 2974 and 2980 with Kuk-Nashur III, according to a different order of sukkalmah succession, not fully published yet, however, as already stated, and as admitted by Mofidi, this order of succession is still unclear, thus in accordance to the system adopted here [see Chapter 2], and the filiation of the different Kuk-Nashurs, 2974 is here associated most probably with Kuk-Nashur II, and the others with any of these individuals), limits the evidential application of these items. This thus limits the chronological or directional information of the Elamite-Mesopotamian glyptic interaction as one cannot obviously ascertain which seals should be so used in this reconstruction. For this reason, all the possible synchronous ‘dated seals’ are presented here, though the possibility that any number of these are not relevant should be noted.

ECS Corpus, Volume I, Part III 735 Thus, the general presence of Elamite-Mesopotamian synchronous ‘dated seals’ evidences the pattern of Elamite-Mesopotamian interaction already seen in the political/historical realms, and in the more general stylistic development of many Elamite styles (particularly the ‘Mesopotamian Related’ styles), by providing a basic chronological indication of form and theme contemporaneity (notwithstanding the limitations placed on this evidence by synonymously named kings, uncertain order of dynastic succession, and lacunae of kings unrepresented by ‘dated seals’ generally, and those with Mesopotamian synchronisms but no such seal types specifically). This evidence may be used, in association with others, in the discussion of Elamite- Mesopotamian interactions, as will be detailed below.

7.2 Location and Chronology of Elamite-Mesopotamian Seal Style Interaction The synchronous Elamite-Mesopotamian ‘dated seals’ demonstrate that there was a degree of similarity in type and form (general style and subject matter) between the glyptic styles of the two regions. It should be recalled that all but the four Tepti-ahar ‘dated seals’ (2912, 2982, 2984, 2985; all from Haft Tepe) with a Mesopotamian synchronism were provenanced from Susa. This adds to the general impression, already mentioned, that Susa especially had a particular relationship with Mesopotamia, that may not have been generally replicated elsewhere. This thus raises the question of the geographical extent of the Mesopotamian-Elamite interaction. The ‘dated seals’ also provide particular evidence for chronological moments of Mesopotamian-Elamite glyptic interaction. The degree to which this evidence can be extended beyond the ‘dated seals’, and if other periods of Elamite-Mesopotamian glyptic interaction, not necessarily represented by preserved or known ‘dated seals’ occurred, also requires some examination. As has already been outlined, the glyptic styles have been divided into three general classifications, the ‘Mesopotamian Related’, ‘Elamite’ and ‘Shared Mesopotamian-Elamite’ styles. The ‘Mesopotamian Related’ styles demonstrate a degree of Mesopotamian influence and interaction in the Elamite glyptic styles, as these styles are indeed defined by their Mesopotamian characteristics. Even more so, the ‘Shared Elamite-Mesopotamian’ styles demonstrate this phenomenon to a greater degree, as they are defined by the fact that these items appear, in essentially the same manner, in both regions. Thus these styles, like the synchronous ‘dated seals’ (that themselves do indeed belong to this classification, generally, as ‘Mesopotamian

ECS Corpus, Volume I, Part III 736 Related’ styles) evidence the pattern of Mesopotamian-Elamite glyptic interaction. Thus the STS, JNRS, SF, ARS, UTRS and OBRS are classified as ‘Mesopotamian Related’ styles, and the GS and KRS as ‘Shared Elamite-Mesopotamian’ styles; each of which demonstrate, in slightly differing patterns, Elamite-Mesopotamian glyptic interaction. On cursory examination, a method of resolving the question of location of Elamite-Mesopotamian interaction may be an examination of the site and provincial distribution of these style classifications (the ‘Related’, ‘Shared’ and ‘Elamite’ classifications). However, the accuracy and efficacy of such a study is tempered by the quality of the data set. As already detailed above, there is great variability in the presentation and quality of the information across the sites, and of the relative proportion and presentation of the glyptic data. The nature, function and extent of ancient occupation also differs greatly across the Corpus. Thus Susa was a large city centre (if not a true capital) that was occupied for the entire chronological extent of this study (and indeed, to significant degrees before and after) that has undergone almost continual, large scale exposure and excavation for over a century. Whereas Choga Mish, on the basis of the current available evidence, may be characterised as a reasonably sized settlement with occupation generally limited to earlier pre- and proto-literate periods (only the last, Susa II equivalent phase of which is relevant here), with limited evidence for later ‘Old Elamite’ and ‘Achaemenid’ period occupations, though only a relatively limited, and fragmented proportion of this site has been excavated in any detail. Haft Tepe appears to be a single period site (or thus the available excavated material would currently suggest only a late Sukkalmah to early Middle Elamite occupation at the site, which may be considered a general unified whole), the excavated material of which (especially the glyptic material) has been published in a less than satisfactory manner. Both Haft Tepe and Chogha Mish have been excavated for a number of seasons, though of duration no where near that of Susa. The limited, controlled, and specific excavations of Tepe Sharafabad, Godin Tepe, and perhaps, though less satisfactorily achieved or executed, Surkh Dum-i-Luri, Kamtarlan, Chigha Sabz, Tepe Giyan and Chogha Gavaneh (some of which have only undergone short excavation seasons of week durations, not months or years, and in earlier, less scientific schools of archaeological investigation), all meant that relatively small parts of the wider archaeological whole were uncovered, and thus different, more limited information and evidence produced. The specific site function of Choga

ECS Corpus, Volume I, Part III 737 Zanbil, Surkh Dum-i-Luri (perhaps only as is evidenced by the currently available material), Bani Surmah, Kalleh Nisar, and, perhaps (though the limited excavation at this site may also limit the accurate reconstruction of this site function) Tepe Djamshidi (the first two sites display a specifically ‘cultic’ function, a funerary function may be discerned for the others) means that one may expect the archaeological material (including the glyptic assemblage) to differ from general occupation sites, thus making comparison somewhat flawed in general content. Finally, like Susa, Tal-i Malyan may have been a regional centre, at times a capital, that was occupied over successive periods (though this is not to imply that the extent nor size of occupation at Malyan mirrored that of Susa in any or all periods), though unlike Susa, Malyan has undergone generally directed, limited excavations, and not broad exposure, the results of which (including most particularly the important Kaftari period occupation), are not all currently available. In terms of glyptic material, the large scale excavation, and publication, at Susa has produced a massive corpus of available material (though some of the size of this corpus must also be attributed to the ancient reality of the size of this site), while it is known that a significant number of the Malyan glyptic corpus (which one would expect to be smaller than Susa regardless, given the more directed, limited exposure of this site, even without taking into account the difference in duration and size of occupation at the two sites) is not published, nor available for inclusion here. Similarly, the extensive excavation at Choga Zanbil produced a reasonably large corpus that may be generally considered to be representative of the whole (though the discovery of glyptic material in the more recent Zanbil excavations, in an area not previously excavated, may imply that further glyptic evidence awaits discovery at this site), while the looted and damaged nature of Bani Surmah and Kalleh Nisar limits the degree to which the glyptic material from these sites may be considered to be representative of the ancient whole, and thus fit for comparison. The limited excavations, and relatively small exposures, at Surkh Dum, Kamtarlan, Chigha Sabz, Godin Tepe, Tepe Djamshidi, Tepe Giyan and Choga Gavaneh means that the available corpora of these sites cannot (or should not) be considered necessarily representative of the glyptic material from these sites, and thus limits the use of these items as a sample. Finally, the accident(s) of discovery, that may have resulted in specifically, and non-representative, glyptic laden areas undergoing excavation at

ECS Corpus, Volume I, Part III 738 Choga Mish (an apparent archive) and Haft Tepe (a craft production area) also may have produced non-representative samples from these sites. Thus, the general archaeological evidence and specific glyptic remains across the entire Corpus differs greatly, not merely in terms of type and chronology, which may indicate actual ancient realities, but also in modern extent, quality and availability for study. This means that in no, or perhaps relatively isolated and less significant cases (thus Bani Surmah and Kalleh Nisar), event may we compare like with like. Thus, as has already become evident above, it is unclear whether apparent differences between and across sites may be considered a function of different cultural constitution or chronological manifestations of the given societies, and thus an ancient reality, or a function of the different level, degree, quality and availability of the modern archaeological evidence. With this limitation and difficulty in mind, the examin ation of the extent, location, timing and direction of Elamite-Mesopotamian interaction becomes difficult. As already mentioned, an analysis of the relative proportions of ‘Mesopotamian Related’, ‘Elamite’ and ‘Shared’ glyptic styles across the geographical and chronological extent of this study, may aid in a reconstruction of the location and chronology, and perhaps the direction, of Elamite-Mesopotamian glyptic interaction and influence. However, the uneven, non-sample data available for this purpose, as outlined above, must limit the efficacy, correctness and trust worthy nature of this reconstruction.

7.2.1 Direction of Elamite-Mesopotamian Seal Style Interaction The ‘Mesopotamian Related’ styles (STS, JNRS, SF, ARS, UTRS and OBRS) all demonstrate a degree of Mesopotamian influence exerted on Elamite glyptic styles. They demonstrate that subject matter themes, execution methods, materials favoured and design elements from Mesopotamia were introduced, copied, and in some instances, slightly adapted or altered in the Elamite translation (that is, through the addition of design elements of an Elamite nature), in a general manner of replication. Furthermore, the chronological coincidence of the true Mesopotamian styles and their Elamite versions (that is, the UTRS style is directly coincident with the Ur III Mesopotamian style and so on, as evidenced by ‘dated seals’ and other pieces of chronological evidence, discussed above) and the degree of similarity between these manifestations (thus the almost indistinguishable Uruk Mesopotamian and STS styles

ECS Corpus, Volume I, Part III 739 for instance) indicates that this was not a general, vague influence exerted by a neighbouring, similar visual arts system or lexicon, but was a direct, clear association. One may assume that the seal carvers working in the UTRS style (for instance) were aware of, and had seen (if not had actual items from which to copy), Ur III Mesopotamian seals. The, albeit limited, evidence of three ‘dated seals’ of certain Mesopotamian origin at Susa (the three ‘dated seals’ that bear the names of Mesopotamian kings [2701, 2723, 3465], unfortunately the last of which currently, though presumably not in antiquity, has no image preserved) indicate that at least some Mesopotamian seals were present at certain points in history at Susa as a testament to the Mesopotamian styles. Indeed, some of the items classified here to ‘Mesopotamian Related’ styles may themselves have been Mesopotamian originals (that is, carved and created at a Mesopotamian settlement) and then transported/traded to Susa/Elam (see Chapter 4, as hypothesised by others also Amiet 1972: 209 – 212), also indicating a method of Mesopotamian style glyptic transference. Thus the presence of ‘Mesopotamian Related’ styles in Elam indicates Elamite-Mesopotamian interaction (on a glyptic level). The ‘Shared Elamite-Mesopotamian’ styles (GS and KRS) also evidence Elamite-Mesopotamian glyptic interaction, quite obviously. The degree and type of interaction involved in these classifications is different to that of the ‘Mesopotamian Related’ styles however. The ‘Related’ styles are Elamite styles that are created in the image of a functioning, established, contemporaneous Mesopotamian style, they are, in essence replicas or copies of Mesopotamian originals. The ‘Shared’ styles are, as already described and as the name would suggest, shared between the two regions, with neither displaying certain primacy of origin. In the GS and KRS styles original location (Mesopotamia or Elam) cannot be discerned, as the two styles appear generally contemporaneously and basically identically in the two regions. One cannot describe either the Mesopotamian Glazed Steatite items, or the Elamite versions of this style, as copies of originals from the other region, neither appear to be replicas, neither have primacy. If the phenomenon of the ‘Related’ styles may be described as one of an unidirectional influence (Mesopotamia influences Elam; indeed, the ‘dated seals’ evidence also implies an unidirectional Mesopotamia to Elam influence pattern), then the ‘Shared’ styles detail a joint bidirectional influence, with both Mesopotamia and Elam influencing one another in the creation of the ‘Shared’ styles. The precise mechanisms of this interaction are currently unknown however.

ECS Corpus, Volume I, Part III 740 In this paradigm, the ‘Elamite’ styles provide evidence for the indigenous glyptic school and development, and, in the current discussion, a relative control against which the Mesopotamian influence may be tested. It should be recalled, that the division of this classification system implies a greater degree of separation, between the ‘Related’, ‘Shared’ and ‘Elamite’ styles, than is in fact present. While it is correct to say that the ‘Related’ styles are heavily influenced by Mesopotamian forms, types and styles, and the ‘Elamite’ styles are essentially not so influenced, but local manifestations of specific, and distinct, Elamite/Iranian themes, the ‘Elamite’ styles are not completely without influence or imitations of Mesopotamian forms. In some instances this influence may be attributed to that from the Elamite made ‘Mesopotamian Related’ forms, in other cases it may be direct from Mesopotamia. Regardless, the general characterisation of the ‘Related’ styles as Mesopotamian influenced, and the ‘Elamite’ styles as essentially indigenously Iranian/Elamite, remains.

7.2.2 Location of Elamite-Mesopotamian Seal Style Interaction and Influence An investigation as to the chronological and geographical extent and relative (to the ‘Elamite’ styles) distribution of the ‘Mesopotamian Related’ and ‘Shared Mesopotamian-Elamite’ styles may aid in a reconstruction of the timing and location of Elamite-Mesopotamian glyptic interaction. The above cited limitations on the comparison of this (and any) material across the varied and uneven sites of the Corpus should be taken as somewhat of a caveat on the results of such a survey however. Table 7.4 and Graph 7.1 illustrate the distribution of the ‘Related’, ‘Shared’ and ‘Elamite’ styles across the sites of the Corpus, with the extraordinarily large Susa corpus excluded from the graph for ease of reference. Table 7.5 and Graph 7.2 further demonstrate the classification division according to the historical province distribution. In an attempt to minimise, and acknowledge, the uneven distribution of the comparenda, and to present the data removed from Graph 7.1 due to its extraordinary size, the Susian corpus has been removed from the Khuzistan whole and included as a separate entity in Graph 7.2. It should also be recalled in this regard that Tal-i Malyan is the only glyptic producing Fars site in the Corpus, and thus in reality, the Fars province designation is a representation of Malyan. Thus the two ‘capital’ or centre sites are depicted in Graph 7.2 as separate entities, beside the combination provinces of smaller, single period, single function or limited exposure sites that are

ECS Corpus, Volume I, Part III 741 grouped together under the geographical rubrics ‘Khuzistan’ and ‘Luristan’. The difference between Susa and Tal-i Malyan, in terms of occupation, excavation and publication should be recalled however. As with the functional discussion above, only those pieces of evidence that belong within the functioning, classified paradigm have been included in this, and all the following surveys. Thus, the two geometric design classifications (AGD and LGD) and the four miscellaneous groups (‘No Image’, ‘Not Illustrated’, ‘Unclassifiable’ and ‘Miscellaneous’ classifications), that cannot be analysed or classified according to the paradigm have been excluded. The interaction and influence evidence provided by the ‘Miscellaneous’ classified items, already disc ussed above (Chap ter 4), should be noted here however.

Site Distribution of the S tyle Classifications ‘Related’ ‘Elamite’ ‘Shared’ Susa 1085 999 268 Chogha Mish 144 2 – Haft Tepe 2 190 2 Choga Zanbil 1 106 18 Tepe Sharafabad 3 2 – Deh-i Now – 1 – Surkh Dum-i-Luri 17 133 3 Kamtarlan 3 5 – Chigha Sabz – 6 1 Bani Surmah 7 – – Kalleh Nisar 5 2 – Godin Tepe 3 6 1 Tepe Djamshidi 1 – – Tepe Giyan 2 1 1 Chogha Gavaneh 2 – – Tal-i Malyan 3 47 7 Tab le 7.4. Site distribution of the style classifica tions.

Province Distribution of the Style Classifica tions ‘Mesopotamian Related’ ‘Elamite’ ‘Shared’ Khuzistan 150 301 20 Susa 1085 999 268 Luristan 40 153 6 Fars/Malyan 3 47 7 Table 7.5. Province (including Susa) distribution of the style classifications.

ECS Corpus, Volume I, Part III 742 195 190 185 180 175 170 165 160 155 150 145 140 135 130 125 120 115 110 105 100 95 90 85 80 75 70 65 60 55 50 45 40 35 30 25 20 15 10 5 0

h h r h n bil ad uri a ab L abz epe yan ne Mis T Gi Zan a S urm h Nisa in ha a araf e Gava Haft Tepe eh-i Nowum-i- ni S od Djamshidiepe Sh D Kamtarlanhigh T hog hog D C Kall G e ha Tal-i Malya C C pe urkh Tep hog Te S C Mesopotamian Related Styles Elamite Styles Shared Styles Graph 7.1. Site distribution of the style classifications, excluding Susa.

ECS Corpus, Volume I, Part III 743 1100 1080 1060 1040 1020 1000 980 960 940 920 900 880 860 840 820 800 780 760 740 720 700 680 660 640 620 600 580 560 540 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0 Khuzistan Susa Luristan Fars/Malyan Mesopotamian Related Styles Elamite Styles Shared Styles

Graph 7.2. Province (including Susa) distribution of the style classifications.

ECS Corpus, Volume I, Part III 744 Several interesting phenomena become apparent in this survey, some of which may be explained as resulting from the uneven distribution of data discussed above, while others may evidence the chronology, location and areas of interaction. The almost even proportion of ‘Related’ and ‘Elamite’ styles at Susa, with a secondary, though significant contribution of ‘Shared’ styles (as demonstrated by Graph 7.2), is descriptive of the constitution of Susa already discussed, as either the most Elamite Mesopotamian centre, or a Mesopotamian-oriented Elamite centre. In this reconstruction, or at least, in glyptic terms, it appears essentially divided between these two elements, the Mesopotamian (with over 46% of the material) and the ‘highland’ Elamite (over 43%), with a ‘Shared’ Elamite-Mesopotamian glyptic element (contributing 11% of the total) that reflects this interaction pattern. Susa’s position, as a place of interaction between ‘Mesopotamian’ and ‘Elamite’ elements is well known, and may act as a comparative control in this survey (that is, to test whether Khuzistan [excluding Susa], Luristan and Fars [Malyan] operated in a similar or different interaction pattern). Graph 7.2 would appear to indicate that the other regions under discussion, Khuzistan (excluding Susa), Luristan and Fars (Malyan) were all dominated by ‘Elamite’ styles, and thus perhaps one may argue, more Elamite in nature. The difference in relative ‘Related’ and ‘Elamite’ styles in these regions would also appear to be indicative of these areas. Thus Khuzistan contains a more significant ‘Related’ presence than Luristan (over 31% of the non-Susian Khuzistan corpus is ‘Related’, with over 20% of the Luristan corpus so classified), though the dominance of ‘Elamite’ styles in both regions is clear, and the ‘Shared’ contribution generally constant (in Khuzistan, ‘Elamite’ styles comprise over 62% of the corpus, and ‘Shared’ styles over 4%; in Luristan, ‘Elamite’ styles [including the LPS ‘Iranian’ style] number over 76% of the whole, with a 3% contribution of ‘Shared’ styles). This pattern may be explained by the fact of the geographical position of Khuzistan, including Susa, as essentially an extension of the alluvial plain of Mesopotamia, and thus a major place of interaction. Much of the ‘Elamite’ classified material from Surkh Dum is not truly of ‘Elam’, but rather may better be described as ‘Iranian’, as this material is generally comprised by the LPS style items, in accordance with the characterisation of this style detailed above. While it may not be accurate to describe this material as ‘Elamite’, it is correct however to divide it from the ‘Mesopotamian Related’ and ‘Shared’ styles,

ECS Corpus, Volume I, Part III 745 and as the subject for discussion here is the extent, location and duration of Mesopotamian interaction and influence, this is an appropriate division. On the basis of Graph 7.2 therefore it may be demonstrated that Khuzistan, particularly Susa but also the entire region more generally, was the primary location for this pattern of interaction and the position from whence Mesopotamian glyptic material exerted influence over the Elamite. It has previously been proposed that direct Mesopotamian-Elamite/Iranian interaction occurred in the Luristan region (between and along the piedmont area specifically) (Carter & Stolper 1984: 107 – 110, 188 – 189), that is, that Luristan was the gateway to Elam through which Mesopotamian forms and ideas were funnelled. Indeed, it has been suggested here that some of the glyptic material from Luristan in the Corpus may be considered true Mesopotamian glyptic material, not just ‘Related’ (particularly suggested in regards to Chogha Gavaneh, Tepe Djamshidi, Bani Surmah, and perhaps the ‘Related’ material from Kalleh Nisar), and therefore, this too would imply that Mesopotamian influence and interaction was exerted directly into Luristan, across the Diyala and Piedmont region, by-passing entirely the southern, lowland Khuzistan influence. Graph 7.1 would appear to confirm this pattern, as it demonstrates the entirely ‘Related’ (or possibly purely Mesopotamian) nature of the Bani Surmah, Tepe Djamshidi and Chogha Gavaneh corpora, and the significant sized proportions of this material in the Kalleh Nisar, Tepe Giyan, and indeed the Surkh Dum-i-Luri corpus. The significant, though not dominant, ‘Mesopotamian Related’ corpus of Godin Tepe may be attributed to the Susa II period ‘Uruk’ intrusion at this site (though the clear ‘Elamite’ dominance accords with the proposal articulated here that at least some of the Godin ‘Oval Enclosure’ phenomenon should be dated to the more ‘Elamite’ dominant Susa III period). At Surkh Dum-i-Luri and Chigha Sabz (and also to a lesser degree at Kamtarlan), the ‘Elamite’ dominance may be linked to the LPS style contribution, and thus may be more aptly described as ‘Iranian’. Regardless of the ‘Elamite’ or ‘Iranian’ nature of the Luristan dominant form, it is clearly not Mesopotamian however (though the dispropionate contribution of the large Surkh Dum corpus, itself primarily LPS in nature, may imply a stronger degree of non-Mesopotamian glyptic items than is accurate). However, in terms of comparative data, it must be recalled that the Khuzistan corpus includes the equally biased and ‘Elamite’ dominant Haft Tepe and Choga Zanbil sites (see above for details, illustrated by Graph 7.1), thus implying that the stronger ‘Mesopotamian’ aspect of

ECS Corpus, Volume I, Part III 746 the Khuzistan corpus compared to that of Luristan (regardless of whether this material may be characterised as ‘Elamite’ or ‘Iranian’) may be an ancient reality. The strong(er) ‘Mesopotamian Related’ character of Khuzistan may be attributed to the contribution of Chogha Mish (see Graph 7.1, this site’s strong ‘Related’ character is itself indicative of the Susa II [STS style], and therefore the strongly ‘Mesopotamian’ aspect of this site). However, the accuracy of this reconstruction may also be accepted, despite the Chogha Mish bias, by the equal contribution and bias of the Choga Zanbil and Haft Tepe corpus, that were from periods and styles equally ‘Elamite’ (Middle Elamite, EME and LME) as Susa II was ‘Mesopotamian’. Thus the picture of a Khuzistan location for the Elamite-Mesopotamian interaction remains. The fact that the large LPS corpus did not contribute significantly to Elamite- Mesopotamian interaction indicates that while some Mesopotamian influence occurred in Luristan, directly from the Piedmont or Diyala, it may be hypothesised that this was not directed into Elam proper therefore, and did not translate into influence on the Elamite styles. The location of Elamite-Mesopotamian influence is thus placed in Khuzistan, primarily at Susa, but also in the general region, and while some Mesopotamian influence entered directly into Luristan, this did not move from this region into Elam proper. An interesting discussion is also presented regarding the ‘Shared’ style distribution across Khuzistan and Luristan. In both cases, the GS and KRS, it has been suggested that the area of interaction and influence was in Luristan, along the Piedmont region and Early Dynastic Diyala in the Glazed Steatite example (hence the fact that ‘Piedmont’ has been an alternate name for this style), and specifically through the Diyala/Hamrin in the Kassite Related Style example (given the distribution of both ‘Pseudo-Kassite’ and ‘Elaborate Elamite’ items in northern Babylonia [], and Zubeidi in the Hamrin). However, in both events Khuzistan and Susa produced more ‘Shared’ items than Luristan. In the GS case, Susa has by far the largest corpus of this style within this study (see Chapter 4), with only virtually negligible contributions from the Luristan sites (Surkh Dum, Chigha Sabz, Godin Tepe and Tepe Giyan each produced one GS item; indeed, Tal-i Malyan in fact produced more than these, with seven, confirming the ‘Elamite’ element of this ‘Shared’ style [see immediately below for discussion of Malyan’s constitution]). In the KRS ‘Shared’ style, significant contributions from Choga Zanbil and Susa overshadows the limited number from Surkh Dum, that is indeed equal to that of Haft

ECS Corpus, Volume I, Part III 747 Tepe. In both instances therefore, the location of Mesopotamian-Elamite interaction that resulted in the creation of a ‘Shared’ glyptic style, as may be judged by the distribution of the results of these creations, may be placed in Khuzistan, not Luristan, and therefore across the lowland, alluvial route, rather than the piedmont or ‘highland’ system route. The accuracy of this reconstruction in regards to the KRS style certainly may be assured as the contemporaneous and similar function Choga Zanbil (Khuzistan) and Surkh Dum-i-Luri (Luristan) may act as appropriate comparative data sets. Finally, Graphs 7.1 and 7.2 both illustrate the style classification proportion of Tal-i Malyan (Fars), in reference to the other sites in the first instance, and the other provinces in the second. As is illustrated, and as one would expect of this highland site ‘Elamite’ styles dominate, in a greater proportion than at any of the other sites or regions (with over 82% of the total Malyan corpus, compared to 63% of the Khuzistan corpus that may be labelled ‘Elamite’, 42% of the Susa corpus and 76% of that from Luristan). The distance and removal of this site from Mesopotamia explains the significantly reduced ‘Mesopotamian Related’ contribution at Tal-i Malyan. As demonstrated in Chapter 5, the ‘Mesopotamian Related’ styles at Malyan are limited to the Susa II/III period STS and JNRS styles, and thus may be linked to the ‘expansion’ into, and beyond, Malyan in these periods, and may not be typical therefore. Similarly the ‘Shared’ element of the Malyan corpus is attributable to the GS style, thus further confirming the chronological timing of the Mesopotamian influence at Malyan (that is, restricted to the Banesh or Susa II and III periods), and the Elamite element of the ‘Shared’ styles. Thus, the significant ‘Elamite’ glyptic presence at Malyan may describe the essential nature of this site as an Elamite (Iranian) centre, that was little influenced generally by Mesopotamian forms, except in the period of major Mesopotamian expansion. In conclusion, according to the general geographic distribution of the ‘Related’, ‘Shared’ and ‘Elamite’ classifications, Susa appears to be the primary location of Elamite-Mesopotamian glyptic interaction. The fact that this interaction does not limit Susian glyptic to purely Mesopotamian imposed styles and types is evidenced both by the significant presence of ‘Elamite’ styles at Susa, and by the presence in the Corpus (and at Susa) of two ‘Shared’ styles. A secondary location, or area where Elamite-Mesopotamian interaction continued, may be located in Khuzistan province (that is, the wider region in which Susa was situated).

ECS Corpus, Volume I, Part III 748 Luristan produced styles and types within the ‘Elamite’ glyptic classification (and indeed other elements of material culture from Luristan demonstrate its ‘Elamite’, that is southern Iranian influence; see the Chapter 5 site style summaries for details of the ‘Elamite’, non-LPS styles at the Luristan sites), that has allowed for its inclusion here broadly within ‘Elam’ (see Chapters 1 and 2). However, in terms of Elamite-Mesopotamian interaction generally, Luristan appears to have functioned in a separate system to that of Khuzistan (and Susa). Some sites in Luristan, in certain periods, appear to have been directly influenced, if not supplied, by Mesopotamian glyptic style systems. In the absence of (true non-LPS) ‘Elamite’ styles generally at these sites, it is hypothesised that these items did not go on to influence the ‘Elamite’ styles or produce the ‘Related’ items, but rather the direction of influence halted. This is evidenced by the fact that in both instances, the location of Elamite-Mesopotamian glyptic interaction that produced the ‘Shared’ styles was not located in Luristan, but Khuzistan. Similarly, the LPS style, which is more correctly labelled ‘Iranian’ rather than ‘Elamite’, and its proliferation in Luristan, does not evidence a major ‘Elamite’ glyptic school, but a generally local (perhaps not even truly cylinder seal related) school that did not participate significantly within the ‘Elamite’ system. The fact that the LPS style occurs, albeit rarely, in Khuzistan (including Susa), and indeed even at Fars (see Chapters 4 and 5), and displays some general similarities with Elamite and Mesopotamian contemporary styles, does indicate however a certain association and link between these groups, as one may anticipate of styles that belong in the same, wider, visual art lexicon of the broader region. Tal-i Malyan appears essentially as an ‘Elamite’ glyptic centre, in keeping with its general perceived Elamite (highland) character. While it may be hypothesised from this that it was from Malyan that the Elamite glyptic styles originated, and then directed or transferred to Susa and Khuzistan, where they interacted with the Mesopotamian styles coming from the opposite direction (to produce the ‘Related’ and ‘Shared’ styles), the current evidence does not in fact suggest this, even in terms of the relative disproportionate comparative data between Susa and Malyan. As already mentioned, and as demonstrated in the annunciation of the styles, most of the styles characterised here as ‘Elamite’ were not without some degree of Mesopotamian type and element, if not in cutting style and rendering technique (such as the elongated EME and LME figures that are reminiscent of the Kassite styles) then in general subject matter theme (thus the ‘banquet’ scenes of the PEU and PEO styles).

ECS Corpus, Volume I, Part III 749 This would appear to imply some Mesopotamian influence even on the ‘Elamite’ styles therefore. One must look for the location of their creation in an area where such styles existed to act as the influencer therefore (either in their true Mesopotamian form, or via an indirect route through the ‘Related’ styles). In the current system as reconstructed here, this would appear to be Susa specifically and Khuzistan generally. Thus, it is hypothesised that the ‘Related’ and ‘Elamite’ styles both originated in Khuzistan (Susa), from whence they spread to Luristan and Fars. The generally non- Susian/Khuzistan styles, LPS (that is not absent from Khuzistan, but less represented) and AS (an almost totally Fars based style) are the only styles not believed to have originated in Khuzistan/Susa, but rather in Luristan and Fars respectively. Whether it was on Iranian or Mesopotamian soil that the actual original creation of the ‘Shared’ styles occurred is unknown (by their very definition), and not ultimately important to this paradigm, however, it may be hypothesised that the ‘Elamite’ elements of these styles were contributed from, if not created in, Khuzistan and Susa.

7.2.3 Chronology of Elamite-Mesopotamian Seal Style Interaction and Influence Having established a general pattern of location for the Elamite-Mesopotamian interaction, a discussion of the chronology of this pattern is required. Table 7.6 and Graph 7.3 illustrate the chronological distribution of the items from the Corpus, according to style type classification. As with the other figures, the non- chronologically defined AGD, LGD and four ‘miscellaneous’ classifications have not been included in this survey generally, though the addition of the ‘Mittanian’ style items will be explained below.

Chronological distribution of the Elamite Corpus, Style Type Classification Mesopotamian Related Elamite Shared Elamite-Mesopotamian Susa II 516 – – Susa III 272 470 266 Susa IV 123 165 – Akkadian & Awan 148 81 – Ur III & Shimashki 118 164 – Sukkalmah 101 51 – Early Middle Elamite – 245 – Late Middle Elamite – 324 35 Table 7.6. Chronological distribution of the Elamite Corpus items according to style type classification.

ECS Corpus, Volume I, Part III 750 520 500 480 460 440 420 400 380 360 340 320 300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

II h III IV ite ite usa am m S usa usa Awan l S S & imashki h ukkalma le Ela ian S S d idd II & Middle E M kka r I A U arly ate E L Mesopotamian Related Style Elamite Style Shared Elamite-Mesopotamian Style Mesopotamian Styles Graph 7.3. Chronological distribution of the Elamite Corpus items according to style type classification.

As is demonstrated the Susa II period is represented by the STS style only, a ‘Mesopotamian Related’ style. The Susa III period includes the ‘Related’ JNRS style, the ‘Shared’ GS style, and the ‘Elamite’ CPE style. In accordance with the process adopted above, and to facilitate this discussion, it is assumed that half of the STF style dates to the Susa III, half to the Susa IV (the accuracy of this division is only hypothetical, and employed for illustrative purposes only), thus in the Susa III period

ECS Corpus, Volume I, Part III 751 this half is added to the ‘Elamite’ style proportion of the CPE. The Susa IV period included the SF ‘Related’ style, and the half STF and total LSF styles under the ‘Elamite’ rubric. The ‘Akkadian and Awan’ period is represented by the ‘Elamite’ PEA and ‘Related’ ARS styles, the ‘Ur III and Shimashki’ by the ‘Elamite’ PEU and ‘Related’ UTRS and the Sukkalmah period by the ‘Elamite’ PEO and ‘Related’ OBRS styles. The separate early Middle Elamite (ME I) period is represented only by the ‘Elamite’ EME style, though the extension of this style into the late Sukkalmah, and thus contemporaneity with part of the Sukkalmah period styles should be recalled. The late Middle Elamite ‘Elamite’ classification includes the LME, AS and LPS styles, with the ‘Shared’ style comprising the KRS. The general ‘Iranian’ nature of the large LPS style, rather than truly ‘Elamite’, should be recalled, though in the current paradigm it is correct not to classify this style as ‘Mesopotamian Related’. The extension of part of the LPS into the first millennium BC should also be noted. The fact that the interaction is located at Susa primarily is also confirmed by the fact that chronological and distribution patterns at this site generally mirrors that of the wider Corpus (though this is itself also a function of the large size of the Susa corpus). The absolute dominance of ‘Related’ styles in the Susa II period is linked to the Mesopotamian ‘colonisation’ of the (see below for further discussion of the ‘Uruk Expansion’). This ‘Uruk Expansion’, or period of Mesopotamian domination, also explains the ‘Mesopotamian Related’ dominance at several other sites primarily occupied in this era. Thus the almost total ‘Related’ nature of Chogha Mish may be so described (see Graph 7.1). The majority ‘Related’ contribution of the Sharafabad corpus may also be directly linked to the Susa II Mesopotamian influenced society, as the ‘Related’ items from this site all date to this period, with the ‘Elamite’ items belonging to the later occupation of Sharafabad. As already mentioned, all the ‘Related’ material from Tal-i Malyan also dates to the Susa II period, and thus illustrates the extent of this ‘expansion’. Susa II period ‘Related’ material also occurs at Godin Tepe, and demonstrates its participation in this system also, though the fact that this material is primarily Susa III, and not Susa II (as discussed above), causes the dominance of ‘Elamite’ classified items at this site (that is, CPE Susa III material). The cursory ‘Elamite’ dominance in the Susa III period is due to the large CPE corpus, and the contribution from the not insignificant STF style. This pattern conforms to the above cited, and returned to below, perceived ‘Elamite’, and

ECS Corpus, Volume I, Part III 752 strikingly not ‘Mesopotamian’, nature of the so-called ‘Proto-Elamite’ period, in opposition to the preceding Susa II Uruk Mesopotamian dominance, as described by Amiet’s alternation theory amongst others (Amiet 1979a; 1979b; this theory will be addressed further below). However, in contrast to this perceived lack, or reduction, of ‘Mesopotamian’ influence at Susa (and elsewhere) in the Susa III period, as Khuzistan joins a highland-oriented system, there is a not insignificant presence of Mesopotamian influenced glyptic material. This is chiefly represented by the JNRS style. As already discussed, the ‘Shared’ styles offer a different degree or type of Mesopotamian-Elamite interaction, however, the very presence of a style that was also repeated and employed, in an almost identical manner in Mesopotamia, at Susa (the GS style) indicates a high degree of Mesopotamian-Elamite interaction, and certainly does not imply an absence of such a system. Indeed, if one were to combine the two classifications that involve a degree of Mesopotamian interaction (the ‘Shared’ and the ‘Related’), this corpus would in fact out number the large ‘Elamite’ group in the Susa III period corpus (even though this group itself includes the possibly disproportionately studied and published CPE style, see above for details). This would imply that far from demonstrating the absence or withdrawal of Mesopotamian influence and interaction in the Susa III period, that a continued pattern of interaction functioned, albeit a different pattern than that of the Susa II period. The ‘Elamite’ style dominance of the Susa IV period equates to the general perceived ‘independence’ of Elam in this period, but the presence of a ‘Mesopotamian Related’ corpus, in not insignificant numbers in this period demonstrates the degree of influence exerted by Mesopotamia, not in a period of conquest (or colonisation in the Susa II case), but through an era of varying trade and bellicose interactions (that is, the competing period of trade and warfare that characterised the historical Elamite- Mesopotamian Susa IV – Early Dynastic interactions). Both the ‘Akkadian and Awan’ and ‘Ur III and Shimashki’ periods have a similar basic structure, with an initial conquest of Susa, and surrounds at least, by a Mesopotamian Empire (the in the first instance, the Ur III state in the second), and the continued engagement of wider Elam in patterns of historical interaction including trade, raids, warfare and, perhaps, vassalage, followed by a secondary phase marked by the control of an indigenous, Iranian/Elamite dynasty (Awan and ultimately Puzur-Inshushinak in the first case, and the Shimashkian kingdom in the second). The significant ‘Mesopotamian Related’ corpora of both

ECS Corpus, Volume I, Part III 753 periods may be directly associated with the conquest and control of Susa, as this site contributes the majority ‘Related’ material in both instances (thus over 97% of the ARS corpus [the ‘Akkadian and Awan’ period ‘Related’ style] is sourced from Susa, and over 98% of the ‘Ur III and Shimashki’ equivalent, the UTRS style, is sourced from Susa). The fact that in the second of these periods (the ‘Ur III and Shimashki’) the ‘Elamite’ classifications dominate may be attributed to the comparative structure of the two systems. It has already been proposed that the ‘Akkadian and Awan’ period was a less glyptically literate period in the history of this study. This was proposed on the basis of the reduced general corpus in this period, at Susa and elsewhere. If the ‘Mesopotamian Related’ material from this period may be attributed to the Akkadian conquest and imposition at Susa (as is indeed evidenced in the ARS ‘dated seals’), then it holds that the ‘Elamite’ material, as well as serving the non-Mesopotamian glyptic needs during the era of the Akkadian conquest would also particularly represent the ‘Awanite’ (that is, post-Akkadian) phase of this period. The reduction of this ‘Elamite’ element in this period (and indeed of glyptic in this period generally as has already been proposed) may be linked to a lesser need for glyptic associated administration. One may hypothesise a less centralised ‘Awanite’ system, as indeed the Akkadian system was not as overly bureaucratised as the Ur III state, or that an administration function of glyptic material was completed in another manner. The indigenous (and therefore not Mesopotamian, and regardless of the stylistic origin, cylinder seal based administration may be regarded as essentially a Mesopotamian original form) Iranian invention of the ‘Linear Elamite’ script in this period cannot be disregarded in this discussion. As already mentioned, the general increased glyptic material in the ‘Ur III and Shimashki’ period may be linked to the imposition of the Ur III state and it’s overbearing bureaucracy in the first instance, and an adoption of this system in the second. The fact ‘Elamite’ styles dominate ‘Related’ in this period, despite the Ur III state’s administrative imposition may be attributed both to the fact that individuals at Susa and elsewhere used ‘Elamite’ forms in the Mesopotamian administrative system, and that, following the expulsion of the Mesopotamians, that the Shimashkian state adopted the Mesopotamian form administrative system (as also this state adopted other forms of Mesopotamian administration and recording, including an Elamite version of the Mesopotamian dating system [De Graef 2008]), albeit with ‘Elamite’ styles, and thus the PEU style was in use at Susa for a longer period than the UTRS, and thus one would expect it to be larger.

ECS Corpus, Volume I, Part III 754 The larger ‘Related’ classification in the Sukkalmah period appears incongruous to the pattern established previously of Mesopotamian types dominant during periods of Mesopotamian conquest or control, especially as, apart from the brief (and uncertain) Gungunum interregnum early in this period, Sukkalmah Elam, and indeed Susa, was not conquered or controlled by Mesopotamia, but was an independent, and ultimately very powerful state. However, the dominance of the ‘Related’ style items over the ‘Elamite’ material is probably somewhat of a false impression, created by the arbitrary separation of the early Middle Elamite EME style from the Sukkalmah material in this paradigm. As detailed in full above, the EME style most probably more accurately dates to the (late) Sukkalmah and early Middle Elamite periods, across the somewhat arbitrary division between these periods. Thus if, for the sake of argument here, only a third of the EME items was considered Sukkalmah in date and thus added to the PEO items as an ‘Elamite’ Sukkalmah period style, this classification (‘Elamite’), would out number the ‘Related’ contribution in this period, thus providing a more accurate reflection. The fact that a still significant number of ‘Mesopotamian Related’ materials occur in the Sukkalmah period corpus is of interest however, as it demonstrates that even in the absence of political control or conquest, Mesopotamian glyptic styles still influenced and interacted with ‘Elamite’ types (this is also reflected in the fact that the PEO style has a similar basic form, the standing ‘presentation’ scene, to the Mesopotamian Old Babylonian material). Like in the preceding periods, the ‘Mesopotamian Related’ portion of this corpus is primarily represented at Susa, though in a less dominant manner (over 87%, thus still significantly more common at Susa, though less so in comparison with the ARS and UTRS styles). The remainder came primarily from Luristani corpora, and it is possible that some, if not all of these ‘Mesopotamian’ items were sourced from, or directly influenced by, Mesopotamia, and not via Khuzistan (certainly the Kalleh Nisar, Tepe Djamshidi and Choga Gavaneh material may be so characterised, as indeed may some of the Surkh Dum materials). Finally, in both the early and late Middle Elamite periods the dominance of ‘Elamite’ styles, and the complete absence of ‘Related’ materials depicts a different pattern of Elamite-Mesopotamian interaction then previously demonstrated. This may be partially explained by the dominance of three large corpora sites (aside form the always significant Susa) that all produced primarily ‘Elamite’ (or ‘Iranian’) material,

ECS Corpus, Volume I, Part III 755 and no, or limited ‘Related’ items (Haft Tepe, Choga Zanbil and Surkh Dum-i-Luri). In all three cases ‘Mesopotamian Related’ material is present, though characterised as ‘heirlooms’ or items not considered contemporaneous with the actual period of occupation at these sites (see Graph 7.1), and so are not represented in the chronological Middle Elamite classifications. The treatment of Surkh Dum in this study may have also slightly altered this perception, for, as detailed above, some styles deemed to have been of Mesopotamian origin (or certainly chronologically later), have not been included here (see Chapter 2, and Table 2.18). If twelve of these items, that are classified to generally contemporaneous Mesopotamian styles (the Kassite and Middle Assyrian styles generally, Holmes Expedition [Schmidt et al. 1989] seals 34, 38, 39, 71, 82, 86, 98, 139, 147, 148, 152, 154) are included in the late Middle Elamite period survey (as in Graph 7.3) then a more accurate picture of Mesopotamian ‘Related’ or true Mesopotamian styles within the Corpus may be presented, though the general ‘Elamite’ nature of this period is retained. Similarly, the ten Mittanian seals (from Susa, Haft Tepe and Tepe Giyan), that have been classified in the ‘Miscellaneous Styles’ classification demonstrate a degree of Mesopotamian influence in the early Middle Elamite period (albeit from a northern, peripheral Mesopotamian region not usually included here; more correctly the Mittanian period corresponds with the late Sukkalmah and early Middle Elamite period), and thus may be added to Graph 7.3 also. Again, this addition may provide a more accurate depiction of the Mesopotamian influence in this period, though it does not alter dramatically the general ‘Elamite’ classification dominance in the Middle Elamite period. Indeed, this dominance is also linked to the significant contribution of the two non-Susian styles, the LPS and AS in (especially the later part) of the Middle Elamite period. While one may separate part of the LPS style from this discussion as it more accurately belongs in the first millennium BC, even if one were to therefore subtract half of the LPS material, ‘Elamite’ (or non-Mesopotamian) styles would still dominate. The fact that two non-Susian styles flourished in this period may partly be an accident of discovery (owing to the chance of the Surkh Dum discovery), but may also indicate that Elam (Iran) had entered a different Mesopotamian interaction pattern in this period. Despite the large, perhaps non-representative corpora from Surkh Dum, Choga Zanbil and Haft Tepe, the general lack of a ‘Related’ Mesopotamian style in this period may be trusted. As already detailed above, it is held

ECS Corpus, Volume I, Part III 756 that given the significant size of these corpora, and the fact that non-‘Elamite’ styles do occur at each, if there was a concurrent ‘Related’ style, this would be represented, along with the ‘Elamite’ and ‘Shared’ styles also. This different system may also be evidenced by the fact that in the Middle Elamite period the role of the ‘Mesopotamian Related’ styles appears to be played by the KRS style, a ‘Shared’ classified group. As already mentioned, the presence of a ‘Shared’ style presents evidence for Mesopotamian-Elamite interaction, albeit under a different system to that represented by the ‘Related’ styles. The presence of the ‘Mittanian’, Surkh Dum Mesopotamian, and the ‘Shared’ KRS items all indicate that in Middle Elamite Elam a pattern of Mesopotamian-Elamite glyptic interaction was not absent therefore, but the absence of a true Mesopotamian influenced (or copied) ‘Related’ style is striking. It is indeed interesting that the Middle Elamite period, a period during which the Elamite-Mesopotamian interactions are perhaps the best known historically (see Chapter 2), appears to lack this high degree of Mesopotamian influence as derived from the ‘Related’ classifications. The presence of the ‘Shared’ and other Mesopotamian styles (Kassite, Middle Assyrian, Mittanian), albeit in small numbers indicates that influence was not absent however. The absence of this higher level interaction may be the result of the long period of general Elamite independence, that culminated in the Middle Elamite period. This would indicate that despite the presence of Mesopotamian-Elamite historical and political interactions, in order for a ‘Mesopotamian Related’ classification to appear (and thus a higher level of interaction to occur) Elam may have required some degree of subservience to Mesopotamia. Thus the ‘Related’ styles may be viewed as imposed, unidirectional influences, while the ‘Shared’ styles may be described as true interactions, mixtures between Elamite and Mesopotamian forms, and therefore a vehicle by which Elamite glyptic influenced Mesopotamia, and thus be described as bidirectional. The elements of Mesopotamian influence discernable in the ‘Elamite’ styles (indeed, in the EME and LME styles) may similarly imply Mesopotamian influence, though not imposed. It is correct to say that the Sukkalmah (and indeed the Shimashki period before it) was a period of Elamite independence, and thus one should expect an absence of ‘Related’ styles in this period too. The presence of ‘Related’ material may be explained while still adhering to this general principle however. Firstly, some time delay or residual Mesopotamian influence may be expected, and thus it was not until the Middle Elamite period after a significant

ECS Corpus, Volume I, Part III 757 period of independence that the Elamite glyptic schools were able to assert their dominance. Also, the fact that a significant number of EME items belong to the late Sukkalmah period indicates that the era of Elamite dominance, and near Mesopotamian absence, does in fact occur earlier (in the late Sukkalmah period) then is apparent in this survey.

7.2.4 Summation In conclusion, the location of Elamite-Mesopotamian glyptic material interaction may be specifically placed at Susa, though Khuzistan generally also participated in this interaction system. Luristan at times appeared to receive direct Mesopotamian influence, though this was not constant, nor did it continue into Elam proper or result in any mixture of styles (that is, the creation of ‘Shared’ styles; the Mesopotamian elements in the local Luristan phenomenon LPS may be attributed to this direct contact however). The ‘gateway’ of Elam, and the location of interaction, was therefore Susa/Khuzistan. Fars/Malyan, being a removed, highland centre, did not participate in this interaction system, and only presented a degree of Mesopotamian influence in the early period of its imposition across Elam (the Susa II and III periods). Susa/Khuzistan was also the primary location from whence the ‘Elamite’ styles were created (and influenced, though in lesser degrees, by Mesopotamian forms) and transmitted to the rest of Elam. The two exceptions to this construction appear to be the LPS and AS styles, the first created in Luristan, the second at Anshan (Malyan; this location in Malyan of the AS creation is linked specifically to its current distribution pattern; the fact that it does in fact display resemblance, albeit removed, and currently little understood, to the Kassite ‘granulated’ caps is of note, and implies that Malyan was not entirely free of Mesopotamian influence therefore). Both the ‘Related’ and ‘Elamite’ classifications, the ‘Related’ more so, display a unidirectional glyptic influence, from Mesopotamia to Elam. The ‘Shared’ styles evidence a mixed, or bidirectional influence, between the two regions. The fact that this classification appears only irregularly, and generally rarely, in the Corpus is significant, and implies that this mixture of influence only occurs infrequently, in certain specific circumstances. Similarly, the fact that the Mesopotamian Glazed Steatite style may be regarded as a primarily peripheral (that is not ‘heartland’ Mesopotamian) style is also of note, and may imply a separate interaction system with the ‘Piedmont’ regions, and not specifically with southern Mesopotamia. The fact that

ECS Corpus, Volume I, Part III 758 the GS interaction and creation, as may be currently discerned, appeared in Khuzistan/Susa and not Luristan places another descriptive element on the creation of the GS style, the precise mechanisms of which remain unknown. The KRS ‘Shared’ style, and the period of its flourishing is descriptive of the period of Elamite- Mesopotamian political interactions that dominate the Middle Elamite period. The ‘Related’ Mesopotamian influenced styles indicate a degree of unidirectional imposition (at least in their original composition), as is evidenced by the general chronology of these phenomena (that is, they flourished in periods of Mesopotamian political control). The elements of Mesopotamian influence in the ‘Elamite’ styles may indicate a more general, lexicon or shared visual art system influence, and while may be described as generally unidirectional, cannot be characterised as imposed. The apparent absence of unidirectional Elamite to Mesopotamia influence is striking. This may in part be a result of the method of study. That is, as the Mesopotamian glyptic material was not examined with the same degree of thoroughness or detail as the Elamite corpus, such possible influence may have been missed. In some isolated examples a certain degree of Elamite influence on Mesopotamia may be discerned, such as the appearance of the ‘animals acting as humans’ motif in the occasional Mesopotamian seal (Pittman 1993), or the LSF elongated animal seals found in isolated Mesopotamian contexts (Collon 1982a). The flag or banner, seen so commonly in the LME ‘banquet’ scenes, may also have been adopted by later Assyrian glyptic carvers (Matthews 1990), though the general banquet form as depicted in the LME cannot be seen as indigenously Elamite, and thus this element may not be characterised as introduced from Elam to Assyria. This actually provides an impression that Elamite influence on Mesopotamian glyptic was generally, or regularly absent, except in the exceptional circumstances of the ‘Shared’ styles. The reasons for this is not entirely clear, though the simple fact that the glyptic system itself was Mesopotamian originally may indicate a certain (deliberate or inherent) arrogance in this area.

7.3 Conclusion – Ethnic Duality, Elamite Cycles and Cylinder Seals Having proposed a general outline of the chronological, locational and directional paradigm of Elamite and Mesopotamian glyptic interaction, the test or judgment of some previously articulated theses regarding the nature of Elamite civilisation, and its inherent links with Mesopotamia may be examined, particularly as pertains to

ECS Corpus, Volume I, Part III 759 cylinder seals. Two main schema have been proposed to explain this pattern, Amiet’s ‘ethnic duality and alternation’ theory (Amiet 1979a; 1979b), and de Miroschedji’s more recent cyclic proposal (Miroschedji 2003). The ‘Uruk World System’ or ‘Expansion’ model, and its associated alternative constructions, is of interest in this regard also, though it only pertains to the earlier Elamite-Mesopotamian interactions. As Mesopotamian-Elamite interactions have been the primary subject matter here, other interactive phenomena, such as the ‘Trans-Elamite’ system (Amiet 1986; 1994), or links between Elam, Central Asia (Pittman 2002; Potts 1999: 178 – 181) or the Gulf region (Potts 1999: 178 – 181; Potts 2006; Kjærum 1986), will not be addressed here, though such proposals are interesting and valid fields of study. Amiet’s landmark thesis regarding the proposed ‘ethnic duality’ of Elam (Amiet 1979a; 1979b) has already been discussed above. This thesis holds that there where two distinct, though interrelated ‘ethnic’ groups present across several periods in Elam, a ‘native’ or indigenous ‘Iranian’/Elamite population at home in the highlands, and an ethnically Mesopotamian population in the lowlands (Amiet 1979a: 195 – 197), presumably immigrant at some point, in the ancient or immediate past (though the fact that this population was not generally assimilated and retained some of its Mesopotamian identity and aspect is essential to this reconstruction). As a further extension of this duality, Amiet holds that these two varying populations alternated between periods of dominance and suppression. Thus, in the initial Susian phase (not studied in detail here as the glyptic material of this period was characterised by stamp, not cylinder, seals), the Susa I period, the indigenous ‘Elamite’ (more correctly, pre-Elamite, Iranian) population held cultural dominance at Susa, as demonstrated by the Iranian-oriented material culture (Amiet 1979a: 196). According to this reconstruction, in the following, Susa II, period, the Mesopotamian ‘ethnic’ element of the Susian population came into dominance (at the expense of the suppressed ‘Elamite’ population), as evidenced by elements of material culture that was Mesopotamian in type, and various features (amongst other factors, such as bevel rimmed bowls and other general ceramics, the STS glyptic style, a ‘Mesopotamian Related’ style may be included here) (Amiet 1979a: 196). This Mesopotamian dominant era is linked to the Uruk period in Mesopotamia, and is thus contemporaneous with the expansion of this culture into northern Mesopotamia also (at Habuba Kabira and other areas) (Amiet 1979a: 196).

ECS Corpus, Volume I, Part III 760 In keeping with the “phenomenon of alternation” of this thesis, in the following Susa III (or ‘Proto-Elamite’ in the language of Amiet) period, the indigenous ‘Elamite’ population again came to be the dominant cultural entity at Susa, now at the expense of the Mesopotamian culture of the preceding Susa II period (1979a: 196 – 197). The Iranian/‘Elamite’ nature of the Susa III period is demonstrated, as described in Amiet’s theory, by certain elements of material culture that are highland Iranian in nature, and that spread, in the manner of an expansion (similar to that of the Uruk Mesopotamian culture) into Khuzistan and across the Plateau, creating material culture and trade, and possibly political, links with Tal-i Malyan and other areas of central and eastern Iran, including , Tepe Yahya and Shahr-i Sokhta (Amiet 1979a: 196 – 197). This ‘Iranian’ Susa III material is distinctly different from that of contemporary Mesopotamia (Amiet 1979a: 196 – 197). Amongst this material may be counted the ‘Proto-Elamite’ tablets, and the ‘Classic Proto-Elamite’ glyptic style. The Elamite (Iranian) dominant culture continued into the early phases of the Early Dynastic Mesopotamian period, after which Elam again entered a long period of Mesopotamian cultural dominance, including links, and in some cases conquest by the Mesopotamians, of the Early Dynastic (Susa IV), Akkadian and Ur III empires (Amiet 1979a: 197), with the formation of ‘Elamite’ “national dynasties” outside of Elam (for which one may more generally infer that Amiet was referring to Khuzistan) (the Awanite and Shimashki dynasties), that united the entire Elamite realm. Following this ‘Old Elamite’ (Susa IV, ‘Akkadian and Awan’ and ‘Ur III and Shimashki’) “integration into the Mesopotamian world”, Elam entered into a long ‘Elamite’ phase, characterised by the Sukkalmah and Middle Elamite periods (and possibly, according to Amiet interrupted by a Kassite interlude prior to what would now be classified as the ME II period, though the extent of this interlude is now doubtful) (1979a: 197). Thus, according to Amiet, may the alternating patterns of ethnic group duality, like the to and fro of the pendulum, be described. More recently, de Miroschedji has proposed a similar, though distinct, cyclic pattern to explain the major trends of Elamite history (Miroschedji 2003). De Miroschedji accepts the ‘ethnic duality’ of Elam, with a lowland Mesopotamian population and highland ‘Elamite’ group, but describes the pattern of ancient Susian and Elamite civilisation in terms of a repeated, cyclic paradigm of growth and decline, characterised by three distinct phases. The first phase is labelled the ‘integration’

ECS Corpus, Volume I, Part III 761 phase (the period of the impetus for political integration of the lowlands and highlands), the second called the ‘expansion’ phase (when the ‘Elamite’ entity controlled Susa and the other neighbouring regions, the southern-central Zagros, Fars and Luristan, in a unified system, that in some instances threatened Mesopotamia) and the third and final phase labelled the period of ‘collapse’ (when the “excessive extension of the Elamite confederation” led to its fall and political fragmentation) (Miroschedji 2003: 23). In this scheme, it is in the final period of collapse, when the confederated constituent parts of the Elamite ‘state’ were essentially isolated and independent, and Susa, as one of these parts, looked to Mesopotamia, and thus ‘alternated’, in Amiet’s language, to a Mesopotamian-oriented phase (Miroschedji 2003: 23). At least four cycles in this pattern are discerned by de Miroschedji, the final one of which concerns the first millennium BC, Neo-Elamite kingdom (Miroschedji 2003: 34 – 36), and so need not concern us here. The first cycle is labelled ‘Proto- Elamite’, with the period of integration occurring in the Susa II period, during which the Uruk period intrusion into Khuzistan (and beyond) provided the impetus for the integration of the ‘Elamite’ entity, which led to the Susa III (‘Proto-Elamite’) expansion phase of consolidation and cultural spread (to the highland regions, Fars/Anshan, and influence beyond to Yahya, Sialk and Shahr-i Sokhta), and then a period of collapse in the terminal Susa III period (that era coincident with the Early Dynastic I Mesopotamian period) (Miroschedji 2003: 24 – 25). The second cycle is labelled ‘Old Elamite’; the integration period corresponds to the Susa IV, ‘Akkadian and Awan’ and initial ‘Ur III and Shimashkian’ (particularly the Ur III, pre- Shimashkian era) periods in the current paradigm, with the impetus for integration the intervention and interaction (often bellicose in nature) of the Early Dynastic city states and the conquest of Susa by the Akkadians and Ur III kings, the ‘expansion’ is represented by the fully confederated Shimashkian and Sukkalmah kingdoms, with the ‘collapse’ occurring in the late Sukkalmah and early Middle Elamite (ME I) periods (Miroschedji 2003: 25 – 31). The final relevant cycle is labelled the ‘Middle Elamite’ cycle, with the period of integration occurring in the (rather confused) Middle Elamite I period (the exact impetus in this period for this integration is not entirely clear, nor explicitly stated, though one may identify the possible Kassite intervention of Kurigalzu I discussed above), the expansion phase reached in the heights of the Middle Elamite II and III periods (including the foundation of Choga

ECS Corpus, Volume I, Part III 762 Zanbil), and the collapse occurring after the Babylonian invasion in 1115 BC, into the ‘dark age’ of the early first millennium BC (Miroschedji 2003: 31 – 34). Another reconstructive paradigm of interest here concerns the early periods generally and the Susa II period specifically, and is generally labelled the ‘Uruk Expansion’ (Algaze 1993). According to this model, as hypothesised by Algaze and others (Algaze 1993; Collins 2000; Rothman 2001 [ed.]), the Susa I – Susa III cultural pattern, can be explained by the intrusion of Mesopotamian peoples bearing the Uruk culture, as colonists in the Susa II period, for the direct purpose of control of resources. Thus the Susa I period, the culture of which is manifestly, and indigenously Iranian, is superseded and replaced by the “wholescale regional colonization” of Khuzistan by Mesopotamian colonists, who bring with them the Uruk material culture, that manifests itself in Iran as the Susa II culture (and evidenced here by the STS style) (Algaze 1993: 15 – 17). This Uruk intrusion apparently explains the general “Sumerian” (that is, Mesopotamian) character of the Susa II artefacts, its break from the preceding Susa I culture, and the evidence for evolution of these forms in Mesopotamia (but not in Iran) (Algaze 1993). In this reconstruction, following the collapse of the Uruk World System (the mechanisms and causes of which need not concern us here), Khuzistan reverted to its original cultural constitution in an highland oriented culture, manifested by the Susa III period (Algaze 1993). This reconstruction thus differs from that of Amiet’s in that the Mesopotamian culture manifested in the Susa II period is intrusive, borne by immigrant colonists from southern Mesopotamia, whereas Amiet holds that the Mesopotamian element was always (or at least, prior to this period) present in Elam, and thus generally indigenous (though presumably this cultural group adopted the current forms and styles of neighbouring Mesopotamia), and merely gained dominance over the ‘Elamite’ element in the Susa II period. Other scholars, including Oates (1993), Steinkeller (1993), Collins (2000), Rothman ([ed] 2001) and Potts (1999: 52 – 79) have all criticised elements of Algaze’s thesis, particularly as regards the motivation and structure of this action (whether for control of resources, as structured for deliberate colonisation, or movement of peoples, or cultural/political refugees), but retain the general colonisation hypothesis (that is, an intrusion or influx of immigrant peoples, bearing the Uruk Mesopotamian culture). Several criticisms, both generally and specifically, may be described regarding these theses, and thus the effectiveness in regards to their application to Elam, and

ECS Corpus, Volume I, Part III 763 indeed the glyptic material, may be raised. The concept of Elamite ‘ethnic duality’ is an ancient one, and can be attributed to Scheil, who first proposed this constitution due to the presence of texts at Susa both in ‘native’, or indigenous ‘Elamite’ (Scheil’s ‘Anzanite’) and Semitic, Mesopotamian Akkadian language (Scheil 1900). The presence of these texts, and other elements of material culture, including cylinder seals as we have seen, simultaneously at Susa (and elsewhere in Elam) does indicate that two separate cultural entities were present at Susa, in certain periods, to greater and lesser degrees, exemplified most plainly by the varying use of Elamite and Akkadian in written texts (though whether one may equate these different elements with ethnic groups or other social classifications must be hypothesised). Whether these groups were ethnically divided, or indeed, both permeant and present since time immemorial may be debated however. The very fact that for periods of its history, as detailed here, ‘Elam’ functioned as an independent entity (during the Susa III, Susa IV and ‘Awanite’ eras, and the Shimashkian, Sukkalmah and Middle Elamite kingdoms), that was interrupted and conquered by clearly foreign groups and conquers (the Akkadian and the Ur III state certainly, but possibly also the colonists if not conquers of Uruk Mesopotamia, and the at times allies at others rivals of Early Dynastic and Kassite Mesopotamia) does not imply that two, parallel populations were competing for and alternating between patterns of dominance and suppression, but rather that the foreign imposition, of conquers or colonists created the dual cultural system. Whether there is some ancient link for all or part of the Susian/Khuzistan population in ancestral Mesopotamia is possible, but immaterial to this discussion. In terms of Amiet’s alternation, it seems incongruous that time and again an imbedded population, within a ‘foreign’ or alien cultural system would rise to power, control or dominance, and adopt and express the same cultural system as that flourishing contemporaneously in their ancestral homeland, after the suppression (to the degree that the ancestral tradition was not archaeologically discernable) without some intervention from the ‘homeland’. Rather it seems more accurate to describe the Mesopotamian elements and influence, as evidence of an intrusion or imposition from Mesopotamia, that indeed may have been welcomed or rejected by the people of Elam, depending on any familial or ancestral ties that they personally may have held, no matter how ancient the link.

ECS Corpus, Volume I, Part III 764 Indeed, as a general criticism of Amiet’s thesis as discerned in the glyptic material, an assemblage that was indeed, chief among Amiet’s evidence, the polarity of Mesopotamian and Elamite styles and forms across these periods, especially in the Susa III period (Amiet’s centrepiece argument), as detailed above, appears to be a false impression. The presence of the JNRS (‘Related’) and GS (‘Shared’) styles within the Susa III corpus (along with the ‘Elamite’ CPE and half STF style) means that the Susa III period corpus is by no means uniquely and distinctly ‘Elamite’ (though the CPE style may be so described certainly), but characterised by significant Mesopotamian influence. Also, as outlined above (Chapter 4), thematically, and in some depiction and rendering elements, the CPE is related to the STS, in a developmental sense, thus again disproving a complete break between the Susa II (apparently Mesopotamian) and Susa III (apparently Elamite) glyptic expressions. De Miroschedji’s thesis particularly, but all those discussed here generally, also have the effect of essentially distilling the Elamite civilisation into a desiccated pattern of inevitable rise and fall, irrespective of the ancient reality. This has the effect of removing any and all distinctions that may have existed across these cycles (for indeed, as has been demonstrated above, the Elamite-Mesopotamian patterns that characterised the Susa II, Susa III, Susa IV, ‘Akkadian and Awan’, ‘Ur III and Shimashki’, Sukkalmah and Middle Elamite periods are all different, and involve periods of colonisation, conquest, political interactions, trade and alliance, that manifested itself in different patterns of Mesopotamian influence over the glyptic material), and providing the false impression that the history of Elam was a cyclic whole that always concurred with such a pattern. Indeed, for instance, the nature of the Akkadian intrusion was significantly different from that of the Ur III, in scope, design and impact, while the independence of the ‘Awanite’ dynasty also differed significantly from that of the Shimashkian kingdom, and the ‘world power’ of the Sukkalmah state. Furthermore, on a purely theoretical level, the conflation of the ‘Old Elamite’ system into one cyclic pattern does not concur with the evidence presented above, nor does the ‘impetus’ era of the ME I period concur generally with the pattern established previously. The ‘Uruk Expansion’ model may or may not accurately describe the cultural phenomenon of that period, and there is not the time, space nor inclination to discuss in detail the relevant arguments as to its appropriateness here. Besides which, this model only applies to the earliest phase of this study. Whatever the accuracy of the

ECS Corpus, Volume I, Part III 765 model, and whether the Uruk Expansion may be characterised as ‘simple’ movements of people, or organised population groups, a structured desire for resource control, or the immigration of peoples fleeing an imposed social order, or merely in search of land, the intrusion of an Uruk cultural phenomenon into Elam, as borne by individuals may be accepted. This sets the general pattern for the wider chronological span under investigation here. This paradigm, as may be judged from the cylinder seal evidence, presents a picture of an Elamite civilisation (which indeed, culturally, politically, socially and chronologically can never be truly described as a monolithic entity) that was at times independent, and at times controlled or under the influence of foreign Mesopotamian peoples. This influence may have been the result of colonists (the Susa II), rivals that were intermittingly allies (the Susa IV, and perhaps more allies then rivals for much of the Middle Elamite period), and conquerors (the Akkadians and Ur III kings). Glyptically, this influence is expressed in the presence of Mesopotamian influenced (and in some cases imposed) glyptic styles, and rarely, and in specific circumstances (and never in the context of conquest or Mesopotamian control) the creation of ‘Shared’ glyptic styles, that were mixed Elamite and Mesopotamian. The variation across time of the patterns and degree of this Mesopotamian influence may be directly linked to the manifestation of these styles. Thus, the only period when true Mesopotamian colonisation may be inferred (the Susa II) is the only period where solely Mesopotamian influenced material occurs (in the glyptic realm). Periods of ‘Elamite’ independence may be discerned by ‘Elamite’ glyptic material (the Middle Elamite period specifically, but also in the Susa III, Susa IV, and ‘Awanite’, Shimashkian and Sukkalmah phases), and periods of Mesopotamian control by Mesopotamian influenced material (including the degree of this control, as evidenced by the proliferation, and likeness of Ur III styles). This creates a pattern of almost continual Mesopotamian glyptic presence or influence in the Elamite Corpus (for even in the Middle Elamite absence of a ‘Related’ styles, foreign styles [Kassite, Middle Assyrian, Mittanian], the ‘Shared’ KRS style, and the presence of Mesopotamian elements in the ‘Elamite’ styles all evidence some Mesopotamian influence). It is proposed that this is not attributed to any necessary ethnic duality in the Elamite population however, but rather to the residual and continual imposition from foreign (Mesopotamian) intervention. This does mean, nor infer, that there was never any originally Mesopotamian element in the Elamite (and especially Susian) population, for indeed there may have been, and Mesopotamian peoples may have

ECS Corpus, Volume I, Part III 766 been immigrant at varying times, and residual to varying degrees, and so a manner of ethnic duality (or indeed, in regard to the multiple constituent and confederated parts of Elam over the span of this study, a degree of poly-ethnicity) may have existed. It is merely argued here that this did not express itself, nor was the chief catalyst for the Mesopotamian influence discerned in the Elamite cylinder seal styles, but rather, that this was the result of foreign, Mesopotamian intervention, and the varying pattern of this manifestation linked to varying patterns of Mesopotamian-Elamite interaction.

ECS Corpus, Volume I, Part III 767 Chapter 8 – Conclusion 8.1 General Conclusion and Summation The aim of this study has been the articulation of a cylinder seal stylistic development or paradigm for the Elamite glyptic material from c.3500 BC (or the period of the creation/development of cylinder seals in Elam and Mesopotamia) to c.1000 BC (or the end of Middle Elamite period and beginning of a new glyptic and cultural system). Following the articulation of these styles, the secondary aims were the analysis of the function of Elamite cylinder seals, and the identification and examination of the pattern of Elamite-Mesopotamian interaction, especially as pertains to the glyptic styles. In all cases, the motivation for this study was the relative lack of relevant sources and studies devoted to these subjects regarding Elamite glyptic material (and not due to a dearth of actual information), in reference to the proliferation of Mesopotamian studies, and the desire to unite what limited Elamite sources were available, into a single, functioning descriptive system. The primary result of this study has been the production of the Elamite Cylinder Seal Corpus Catalogue, presented as a separate volume to this study. This Catalogue is fully annotated, and includes all currently available, fully published (that is with type, image, material and dimensional data) ‘Elamite’ cylinder seal glyptic materials (according to the chronological and geographical boundaries of Elam established here). The Catalogue has been structured according to chronological and stylistic order, and thus provides a functioning reproduction of the proposed chronological and stylistic development of Elamite glyptic material aside from the details and discussion of these phenomena presented in the body of this study. The styles are presented in general chronological order, though naturally contemporaneous styles have been presented in succession so as to facilitate the Catalogue. The items are also presented according to the geographical and subject matter divisions. Thus the Susa II Style (STS) items are presented first in the catalogue as this is described as the first style. The items are then divided according to the subject matter based sub-group divisions, which are themselves ordered geographically. The Susa material is always presented first, and the order of operations then follows the order established in the discussion of the ‘Elamite’ sites in this text (particularly as presented in Chapters 2 and 5). Sixteen sites were included in this Corpus as glyptic producing (contributing) sites, divided according to broad historical provinces; the Khuzistan sites include Susa, Chogha Mish, Haft Tepe, Choga Zanbil, Tepe Sharafabad and (the surface surveyed site only)

ECS Corpus, Volume I, Part III 768 Deh-i Now; the Luristan sites include Surkh Dum-i-Luri, Kamtarlan, Chigha Sabz, Bani Surmah, Kalleh Nisar, Godin Tepe, Tepe Djamshidi, Tepe Giyan and Chogha Gavaneh; only one site in Fars, Tal-i Malyan, produced relevant and included glyptic material. Eighteen styles, two geometric design classifications and four miscellaneous classifications were articulated to describe the Elamite Cylinder Seal Corpus. These styles have been classified according to whether they may be characterised as ‘Elamite’, ‘Mesopotamian Related’ or ‘Shared Elamite-Mesopotamian’ styles. The ‘Elamite’ styles are here described as essentially indigenous Elamite (or in the case of the LPS, and the pre-Elamite proper periods [that is pre-Susa IV], ‘Iranian’) creations, though in some cases a degree of Mesopotamian influence may be discerned in general subject matter, depiction methods, renderings and materials. In these styles (except in rare circumstances, and in regards to the ‘Shared’ styles), the Mesopotamian-Elamite influence/interaction appears to be unidirectional, from Mesopotamia to Elam, but not reciprocated from Elam to Mesopotamia. The ‘Mesopotamian Related’ styles are defined as Elamite copies or replicas of contemporaneous Mesopotamian styles, to which they obviously demonstrate a significant degree of Mesopotamian influence. Again this influence is unidirectional, Mesopotamia to Elam. The ‘Shared’ styles demonstrate original (that is not copied or replicated), and certain, placement within both the Elamite and Mesopotamian glyptic lexicon and thus cannot be described as created and then copied in one or the other region. These styles thus also demonstrate Elamite-Mesopotamian glyptic interaction and influence, though in this regard the influence is bidirectional, with each system engaging and influencing the other in the creation of these styles. In all cases, the ‘Elamite’ and ‘Related’, and possibly the ‘Shared’, the place of origin of these styles (the places where they were influenced by Mesopotamian glyptic sources and created) appears to be Khuzistan generally and Susa in particular, from whence they spread throughout ‘Elam’. The exception to this creation location are the two specifically non-Susian styles, LPS (essentially a Luristan based style) and the AS (an almost exclusively Malyan style) styles, which were created in Luristan and Malyan respectively, as may be evidenced in their geographical distribution. Table 8.1 presents a slightly altered from of Table 4.111, and summarises the chronological and classification allocation of the styles of the Corpus. Styles coloured red indicate ‘Elamite’ styles, blue marks the ‘Mesopotamian Related’ styles and green

ECS Corpus, Volume I, Part III 769 the ‘Shared Elamite-Mesopotamian’ styles. To demonstrate the different locational and derivational nature of the LPS and AS styles these items are presented under a different hue of red. Yellow indicates the generally non-specifically chronologically definable geometric classifications (AGD and LGD), and the yellow scored bars indicate the possible projected extension of any number of these items into the intervening ‘Old Elamite’ period. The structure of this table is also designed to demonstrate the contemporaneity of certain styles, including the partial synchronicity and overlapping use, as well as the developmental or continuum nature of some of these styles. Thus those styles depicted along a general linear horizontal placement demonstrate some developmental and continuation, while the vertical axis demonstrates (general or partial) contemporaneity. The dashed lines between certain ‘Old Elamite’ styles further indicate a certain degree of blending, overlap and contemporaneity. It should be noted that no scale has been used in this figure, and thus no deduction as to chronological span (apart from the period of use indicated) should be inferred from the relative size of each plot.

‘Archaic’ ‘Old Elamite’ Middle Elamite Susa Susa Susa Akkadian Shimashki Sukkalmah ME II III IV & Awan & Ur III I II III IV CPE PEA PEU PEO EME LME STF Popular Elamite AS LSF LPS AGD ? ? LGD STS SF ARS UTRS OBRS JNRS GS KRS Table 8.1. Summary and survey of the Elamite Cylinder Seal Corpus Styles according to chronological placement, style classification and development.

In general summary (though the reader is directed to Chapter 4 for details of type, theme, material, and design), the first, chronologically speaking, style is the Susa II Style (STS), a ‘Mesopotamian Related’ style. The Susa III styles include the Classic Proto-Elamite (CPE) style and the initial part of the Susa III/IV (STF) style, both ‘Elamite’ styles, the Jemdet Nasr Related Style (JNRS; a ‘Related’ style, that may have been initially developed in the late Susa II period) and the ‘Shared Elamite- Mesopotamian’ Glazed Steatite (GS) style (that may have continued into the earliest Susa IV period, and which is also known from Mesopotamia, particularly the ‘Piedmont’ peripheral region of this area). The remaining part of the STF style and

ECS Corpus, Volume I, Part III 770 the Late Susa IV (LSF) style represent the ‘Elamite’ styles of the following Susa IV period, and the Susa IV (SF) style the ‘Related’ group. The Popular Elamite style is a continuous ‘Elamite’ style, that is divided according to subject matter and chronological groups into the PEA (dating to the ‘Akkadian and Awan’ period), the PEU (dating to the ‘Ur III and Shimashki’ period) and the PEO (of the Sukkalmah period) sub-styles. Throughout the continual PE development several ‘Mesopotamian Related’ styles flourished in the specific periods. Thus the Akkadian Related Style (ARS) occurred in the ‘Akkadian and Awan’ period, and is contemporaneous generally with the PEA, the Ur III Related Style (UTRS) dates to the ‘Ur III and Shimashki’ period, contemporaneously with the PEU style and Old Babylonian Related Style (OBRS) occurs in the Sukkalmah period in the same general period as the PEO style. Some extension of the PEA/ARS into the ‘Ur III and Shimashki’ and the PEU/UTRS into the Sukkalmah period demonstrates the general continuum nature of these styles. The Early Middle Elamite (EME) style is an ‘Elamite’ classified style that was initially in use in the late Sukkalmah period (and indeed probably itself developed from the PEO style) as well as the early Middle Elamite (ME I) period. The late Middle Elamite (post-ME I) period produced the ‘Elamite’ classified Late Middle Elamite (LME) style, Anshanite Style (AS) and Luristan Provincial Style (LPS), the later two of which were not, uniquely amongst the styles of the Corpus, developed in Khuzistan/Susa, but in Malyan/Fars and Luristan respectively. A ‘Shared Elamite- Mesopotamian’ style also flourished in the late Middle Elamite period, labelled the Kassite Related Style (KRS), and includes the ‘Pseudo-Kassite’ and ‘Elaborate Elamite’ divisions. All the styles (except for the very small LSF style) were divided into subject matter based sub-groups. An analysis of the distribution pattern of the ‘Elamite’, ‘Related’ and ‘Shared’ classified styles of the Corpus indicates the degree of Mesopotamian influence and interaction with Elam. It is hypothesised that Mesopotamian elements and influence in Elamite glyptic material may be attributed to the interaction, intervention and imposition of Mesopotamian peoples and systems, variously as colonists (the Susa II period), conquers (the Akkadian and Ur III Empires, and perhaps the brief Old Babylonian Gungunum interregnum), rivals (the Early Dynastic – Susa IV interaction, and part of the Old Babylonian – Sukkalmah and Kassite – Middle Elamite interactions), allies (the other interactions of the Old Babylonian – Sukkalmah and Kassite – Middle Elamite periods) or through simply

ECS Corpus, Volume I, Part III 771 trade and cultural interaction systems, and not through any residual or remnant continuous Mesopotamian ‘ethnic’ group amongst the Elamite population (though this may not necessary disallow such a group or groups in some or all periods). Several different functions may be discerned in the Elamite Corpus, defined both by type and provenance. In terms of type, all cylinder sealings demonstrate an administrative function, some associated with writing (sealed tablets and sealed clay tablets particularly) and some that are not associated with writing, including immobile storage device sealings (jar sealings, bottle and jar stopper sealings, basket and bale sealings, bulles, and bullae) and immobile storage devices (door and wall lock sealings and door sealings). The administrative function of cylinder seals may be described as the primary, and definitional, function of cylinder seals. In the Elamite Corpus administrative sealings not associated with writing are dominant, and thus imply that this was the more regular Elamite glyptic function, and not the action of sealing tablets as may previously have been generally assumed. This non-writing associated administration aspect of Elamite glyptic may imply that Elam generally functioned primarily in a non-written, or oral (and glyptic) administrative system, as previously suggested (Porada 1965), though the presence of tablets, both sealed and otherwise, should be noted. In both the writing and non-writing associated administrative systems the actual functioning and action of sealing appears to be the same, with the difference occurring chronologically rather than functionally. It is hypothesised that in the earlier periods under discussion here (specifically the Susa II and Susa III periods), or the pre- and proto-literate periods, before written language was developed to the degree that it could convey abstract ideas and complex pieces of information (that is, when the system was essentially mnemonic, or basic in structure, and only detailed, for instance, number and type, and not direction or action, thus verbs were not yet expressible) a seal impression on a jar, door, bullae or (numerical or ‘Proto-Elamite’) tablet conveyed some of the missing, or unarticulated information. Thus the seal impression may have identified the location, origin or destination of the item sealed, or the identity of the individual involved in the sealing (or sending) of the unit, or the identity of the intended recipient. Following the full development of a true writing system (in the Susa IV period and beyond), this sealing information transference was superseded by true writing. Thus the action of sealing changed, to one of authorising, authenticating and identifying individuals involved in a transaction or contract (in the

ECS Corpus, Volume I, Part III 772 case of sealed tablets) or in the ‘sealing’ or deposition of objects in the case of non- writing associated administrative objects. The act of sealing, both tablets and non- writing devices was reduced in the literate period, with many tablets not sealed, and presumably many vessels and storerooms not marked with a sealing. The precise reasoning why some tablets, vessels and areas were sealed, and others not, remains elusive however. This proposed reconstruction is due to the fact that sealings, on numerical and ‘Proto-Elamite’ tablets, bullae and regular sealings are all more common in the earlier, pre- and proto-literate periods across the whole Corpus (thus disallowing for bias of unrepresentative early period occupation sites, such as Chogha Mish). It may be hypothesised that many of the items classified in the two geometric design groups, the Archaic Geometric Design (AGD) and Late Geometric Design (LGD) classifications, that have not been given further chronological classification due to the difficulty in judging and classifying geometric deign devices, may be beads rather than cylinder seals. The real definition or classification of these devices must remain elusive however, though the fact that some of the geometric design items were employed as true seals (that is, there are sealings of geometric design items) indicates that at least in some cases these items have been correctly classified here as true seals. As well as the primary administrative sealing function(s), several functions, loosely classified as ‘symbolic’ may be discerned in the Elamite Corpus. As well as the ‘funerary’ function (where a cylinder seal was deposited in a tomb or grave amongst the other grave goods of the deceased), which apparently occurred relatively rarely in the Elamite Corpus (though this may be a symptom of poor recording rather than an ancient reality), and is understood from elsewhere, the more interesting, and seemingly unique, or characteristically, Elamite symbolic function is the so-called ‘votive’ function. This function is particularly found at Choga Zanbil and Surkh Dum- i-Luri, though also more rarely at Susa. This function may be enabled by the interpretation of seals as representatives or emblems of the owner’s self or identity, and thus the deposition in a temple or cultic structure or area may equate to the placement of a representative or reminder before the deity. It is also possible that the act of devoting a cylinder in a temple may have been a method of seal retirement, or disposal of a no longer required device. In the instance where retirement of a seal was not desired, replacement or replica seals may have been created to save the owner the

ECS Corpus, Volume I, Part III 773 inconvenience of seal loss (this may be particularly apparent at Choga Zanbil in the production of the near uniform faience LME banquet scene seals). As a related phenomenon, the general absence of LPS style sealing items, and the significant number of these items in the Surkh Dum-i-Luri ‘sanctuary’ may imply that these items were not created as, nor intended to be true, functioning seals, but (amuletic) devices intended explicitly for the purpose of temple deposition (thus again explaining the problematic paradox of seal relinquishment). The bias of the Surkh Dum ‘sanctuary’ and the general cursory archaeological evidence from Luristan may present this pattern however, and that may therefore, not be entirely accurate. This is merely an hypothesis therefore, and confirmation or rejection, and greater understanding, must await further excavation and investigation at Surkh Dum-i-Luri and Luristan.

8.2 Directions of Future Study Some results of this study, and indeed some of the limitations of these results, imply several areas and directions for future research. Some of these include subjects and problems not addressed here due to time and subject area constraints. In particular these areas include the complete, thorough examination of the Mesopotamian ‘dated seals’ (including the investigation of those seals/sealings that have not previously been illustrated), to provide a tested paradigm of chronological development and possible absolute dating mechanisms for the Mesopotamian glyptic styles; and the study of the Assyrian (northern Mesopotamia) interaction with Elamite styles (including ‘dated seals’), as has been undertaken here for the southern Mesopotamian glyptic material. In regards to the current Corpus material, the efficacy of some of the reconstructions, and indeed the degree to which proposals may be suggested, has been hampered by the quality and detail of the available material. These problems include the lack of type classification for most of the sealings (that is, the large category), the general absence of petrographic/scientific seal material classifications and the poor or non-existent data regarding glyptic provenance (and indeed, general site functional or classification information aside from the glyptic material). Thus a part resolution to this problem would be the undertaking of a complete, thorough, detailed, physical (in regards to the sealings) or petrographic (in regards to the seals) study of the glyptic items. Such an undertaking

ECS Corpus, Volume I, Part III 774 would provide a more complete and trustworthy data set from which to judge proportions, and thus function, of types and uses. Microscopic and reconstructive studies of the cylinder seals, as undertaken by Sax and her colleagues at the British Museum (Sax et al. 1998; Sax & Meeks 1994; 1995; Sax et al. 2000; Sax & Middleton 1992), and Gorelick and Gwinnett (1978; 1981a; 1981b; 1990; 1992; Gwinnett & Gorelick 1979; 1987), in order to reconstruct cutting and carving techniques, would also provide better, more detailed understanding of the use, development and chronology of cutting methods, and thus aid in the stylistic definition of the Corpus material, particularly the largely unclassifiable geometric design items. Reconstructive reanalyses, and directed control excavations, in the case of thoroughly excavated sites or site areas (such as those that have already been applied at Susa [Carter 1980]), or further exposure and investigation at sites that have only undergone limited actual investigation, would both provide more accurate and detailed site (and thus possibly glyptic) function information and thus may resolve part of the problems with the limited data set. However, it must be recognised that much of the lacking provenance, and thus functional, information is beyond reconstruction and recovery, and thus the material must merely be accepted for what is currently known. In this same regard, it is likely that any number of the may be non-diagnostic items that will evade classification even with thorough physical investigation, thus also providing no new functional information than already presented here. A final area of future study is the associated analysis of the image and styles of the multiple seal sealed bullae, and the associated, though distinct, study of the sealing types in association with the textual (both of type, that is, ‘Proto-Elamite’, Akkadian, numerical tablets) and detail (the data actually recorded on the tablet, and thus in combination with philological study) information of the tablets, may provide further functional information and understanding of the Elamite Cylinder Seal Styles and their function.

ECS Corpus, Volume I, Part III 775 Bibliography Abbreviations AA Arts Asiatiques AAASH Acta Antiqua Academiae, Scientiarum Hungaricae AfO Archiv für Orientforschung AIO L. De Meyer & E. Haerinck (eds.), 1989, Archaeologia Iranica et Orientalia: Miscellanea in honorem Louis Vanden Berghe, Louvain: Peeters. AION Annali dell’Istituto Universitario Orientale di Napoli AJA American Journal of Archaeology AMI Archäologische Mitteilungen aus Iran BaM Baghdader Mitteilungen BAR British Archaeological Reports BiOr Bibliotheca Orientalis CAH Cambridge Ancient History CAJ Cambridge Archaeological Journal CANE J. M. Sasson (ed.), 1995, Civilizations of the , vols. I – IV, New York: Charles Scrbiner’s Sons CDAFI Cahiers de la Délégation archéologique Française en Iran CDR H. Gasche, M. Tanret, C. Janssen & A. Degraece (eds.), 1994, Cinquante-deux reflexions sur le Poche-Orient ancien offertes en hommages à Léon de Meyer, Gent: Peeters EnIr Encyclopædia Iranica EW East and West FHE L. De Meyer, H. Gasche & F. Vallat (eds.), 1986, Fragmenta Historiae Elamicae: Mélanges offerts à M. J. Steve, Paris: Editions Recherche sur les Civilisations IA Iranica Antiqua IRSA Sollberger & Kupper 1971 JA Journal Asiatique JAOS Journal of the American Oriental Society JCS Journal of Cuneiform Studies JEA Journal of Egyptian Archaeology JFA Journal of Field Archaeology

ECS Corpus, Volume I, Part III 776 JNES Journal of Near Eastern Studies LED Luristan Excavation Documents MDP Mémoires de la Délégation en Perse, Mémoires de la Mission Archéologique de Susiane, Mémoires de la Mission archéologiques de Perse, Mémoires de la Délégation Archéologique en Iran MHEOP I L. De Meyer & H. Gasche (eds.), 1991, Mésopotamie et Elam, Ghent: Mesopotamian History and Environment Occasional Publications I. MJP F. Vallat (ed.), 1990, Mélanges Jean Perrot, Paris: Editions Recherche sur les Civilisations NABU Nouvelles Assyriologiques Brèves et Utilitaires NATN Owen 1982 Or Orientalia OrAnt Oriens Antiquus RA Revue d’Assyriologie et d’archéologie Orientale RIA Reallexikon der Assyriologie RIME The Royal Inscriptions of Mesopotamia ZA Zeitschrift für Assyriologie

ABDI, K., 1999 ‘Archaeological Research in the Islamabad Plain, Central Western : Preliminary Results of the First Season, Summer 1998’, Iran 37: 33 – 34. 2000 ‘Islamabad, 1999’, Iran 38: 162 – 163. 2001 ‘Islamabad, 2000’, Iran 39: 299 – 300.

ABDI, K. and BECKMAN, G., 2007 ‘An Early Second Millennium Cuneiform Archive from Chogha Gavaneh, Western Iran’, JCS 59: 39 – 91.

ADAMS, R. McC. 1962 ‘Agriculture and Urban Life in Early Southwestern Iran’, Science 136: 109 – 122.

ECS Corpus, Volume I, Part III 777 AL-GAILANI WERR, L., 1988a Studies in the chronology and regional style of Old Babylonian cylinder seals, Malibu: Bibliotheca Mesopotamica 23. 1988b ‘Cylinder seals made of clay’, 50: 1 – 24.

AL-GAILANI WERR, L. (ed.), 1992 Old Babylonian Cylinder Seals from the Hamrin, London: NABU.

ALGAZE, G., 1993 The Uruk World System: the Dynamics of Expansion of Early Mesopotamian Civilisation, Chicago, London: University of Chicago Press.

ALIZADEH, A., 2008 Chogha Mish, Volume II, The Development of a Prehistoric Regional Center in Lowland Susiana, Southwestern Iran: Final Report on the Last Six Season of Excavations, 1972- 1978, Chicago: Oriental Institute Publications 130.

AMIET, P., 1957 ‘Glyptique Susienne Archaïque’, RA 51: 121 – 129. 1959 ‘Représentations antiques de construction susiennes’, RA 53: 39 – 44. 1960 ‘Le Temple Alié’, RA 54: 1 – 9. 1963 ‘La glyptique syrienne archaïque’, 40: 57 – 83. 1966 Elam, Archée Éditeur: Auvers-sur-Oise. 1971 ‘La Glyptique de l’Acropole (1969 – 1971). Tablettes Lenticulaires de Suse’, CDAFI 1: 217 – 233. 1972 Glyptique Susienne – des origines à l’époque des perses achéménides, cachets, sceaux-cylindres et empreintes antiques découverts à Suse de 1913 – 1967, Paris: MDP 43. 1973 ‘Glyptique Elamite – à propos de documents nouveaux’ AA 26: 2 – 35. 1978 ‘Phase 10 at Shahr-i Sokhta: Excavations in Square XDV and the Late 4th Millennium B.C. Assemblage of Sistan’, EW 28: 9 – 31. 1979a ‘Archaeological discontinuity and ethnic duality in Elam’, Antiquity 53: 195 – 204.

ECS Corpus, Volume I, Part III 778 1979b ‘Alternance et Dualité – essai d’interpretation de l’histoire élamite’ Akkadica 15: 2 – 22. 1980a La Glyptique mésopotamienne archaïque, revised edition, Paris: Centre National de la Recherche Scientifique. 1980b ‘La glyptique du second millénaire en provenance des chantiers A et B de la Ville Royale de Susa’, IA 15: 133 – 147. 1980c ‘Antiquités de Serpentine’, IA 15: 155 – 166. 1980b ‘La période IV de Tépe Sialk reconsidérée’, in J. –L. Huot, M. Yon and Y. Calvet (eds.), De L’Indus aux Balkans – Recueil à la mémoire de Jean Deshayes, Paris: Éditions Recherche sur les civilisations, pgs. 293 – 312. 1986 L’âge des échanges inter-iraniens – 3500 – 1700 avant J. –C., Paris: Notes et Documents des Musées Nationaux 11. 1988 Suse – 6000 ans d’histoire, Paris: Ministere de la Culture et de la Communication Editions de la Reunion des musees nationaux. 1992a ‘Sur l’Histoire Elamite’, IA 27: 75 – 94. 1992b Corpus des cylindres de Ras Shamra-Ougarit II/Sceaux-cylindres en hematite et pierres diverses, Ras Shamra-Ougarit IX, Paris: Editions Recherhe sur les Civilisations. 1994 ‘Un Sceau Trans-Élamite à Suse’, RA 88: 1 – 4. 1996 ‘Observations sur les sceaux de Haft Tépé (Kabnak)’, RA 90: 135 – 143. 1999 ‘Les Sceaux de Kabnak (Haft Tappeh)’, in The Iranian World – Essays on Iranian Art and Archaeology Presented to Ezat O. Negahban, Y. M. A. Alizadeh and S. M. Shahmirzadi (eds.), Tehran: Iran University Press, pgs. 101 – 113. 2005 ‘Les Sceaux de l’Administration Princière de Suse à l’Époque d’Agadé’, RA 99: 1 – 12.

AMIET, P. and LAMBERT, M., 1973 ‘Objects Incrits de la Collection Foroughi’, RA 67: 157 – 162.

ANDRÉ-SALVINI, B., 1992 ‘Historical, Economic and Legal Texts’, in P. O. Harper, J. Aruz and E. Tallon (eds.), The Royal City of Susa: Ancient Near Eastern Treasures in the , New York: Metropolitan Museum of Art, pgs. 261 – 265.

ECS Corpus, Volume I, Part III 779 ARUZ, J., 2002 ‘Power and Protection: A Little Proto-Elamite Silver Bull Pendant’, in E. Ehrenberg (ed.), Leaving No Stones Unturned – Essays on the Ancient Near East and Egypt in Honor of Donald P. Hansen, Winona Lake: Eisenbrauns, pgs. 1 – 14.

BALKAN, K., 1954 Kassitenstudien I. Die Sprache de Kassiten, New Haven: AOS 37.

BALL, C. J., 1889 Light from the East, or the Witness of the Monuments, London: Eyre And Spottiswoode.

BANKS, E. J., 1912 Bismya or the Lost City of – A Story of Adventure, of Exploration, and of Excavation among the Ruins of the Oldest of the Buried Cities of Babylonia, New York: G. P. Putnam’s Sons.

BANNISTER, F. A. and PLENDERLEITH, H. J., 1936 ‘Physico-chemical examination of a Scarab of Tuthmosis IV bearing the name of the god Aten’, JEA 22: 3 – 6.

BARAG, D., 1985 Catalogue of Western Asiatic Glass in the British Museum, London: British Museum Publications.

BERAN, T., 1957 – 58 ‘Die babylonische Glyptik der Kassitenzeit’, AfO 18: 255 – 278.

BERGAMINI, G., 1998 ‘In-na-ba seal of Aakalla, governor of , granted by Su-Sîn’, NABU: 129 – 130.

ECS Corpus, Volume I, Part III 780 BERNBECK, R., 1996 ‘Siegel, Mythen, Riten: und die Ideologie der Akkad-Zeit’, BaM 27: 159 – 213.

BIENKOWSKI, P. and MILLARD, A. (eds.), 2000 Dictionary of the Ancient Near East, London: British Museum Press.

BLACK, J. and GREEN, A., 1992 Gods, Demons and Symbols of Ancient Mesopotamia, Austin: University of Texas Press.

BLACKMAN, M. J., 1985 ‘Chemical Characterization of Sealing Clays from Fourth Millennium BC Sites in Iran’, Transactions of the American Nuclear Society 9: 168 – 169.

BLEIBTREU, E., 1981 Rollsiegel aus dem Vorderen Orient – Zur Steinschneidkunst Zwiscen etwa 3200 und 400 vor Christus nach Beständen in Wien und Graz, Vienna: Exhibition Catalogue.

BLOCHER, F., 1988 ‘Einige Altbabylonische Siegelabrollungen aus Kiš im Louvre’, RA 82: 33 – 46.

BOEHMER, R. M., 1965 Die Entwicklung der Glyptik während der Akkad-Zeit, Berlin: Walter de Gruyter. 1967 ‘Zur Datierung des Epirmupi’, ZA 58: 302 – 310. 1981 ‘Glyptik der späpten Kassiten-Zeit aus dem nordöstlichen Babylonien’, BaM 12: 71 – 81. 1999 Uruk – Früheste Abrollungen, Mainz an Rhein: Ausgrabungen in Uruk-Warka Endberichte 24.

ECS Corpus, Volume I, Part III 781 BOEHMER, R. M. and DÄMMER, H. –W., 1985 Tell Imlihiye, Tell Zubeidi, Tell Abbas, Mainz am Rhein: P. von Zabern.

BOEHMER, R. M. and GÜTERBOCK, H. G., 1987 Glyptik aus dem Stadtgebiet von Boğazköy, Grabungs-kampagnen 1931 – 1939, 1952 – 1978, Berlin: Boğazköy-Hattuša XIV.

BÖRKER-KLÄHN, J., 1970 Untersuchungen zur altelamischen Archäologie, Berlin.

BOTTÉRO, J., 1998 Religion in Ancient Mesopotamia, translated by T. L. Fagan, Chicago; London: The University of Chicago Press.

BOTHMER, D. von, 1961 Ancient Art from New York Private Collections Catalogue of an Exhibition Held at the Metropolitan Museum of Art December 17, 1959 – February 28, 1960, New York.

BRANDES, M. A., 1979 Siegelabrollungen aus archaischen Bauschichten in Uruk-Warka, Wiesbaden: Freiburger altorientalische Studien 3.

BRINKMAN, J., 1970 ‘Notes on Mesopotamian History in the Thirteenth Century B.C.’, BiOr 27: 301 – 314. 1976 A Catalogue of Cuneiform Sources Pertaining to Specific Monarchs of the Kassite Dynasty, Chicago: Oriental Institute of the University of Chicago. 1977 ‘Mesopotamian chronology of the historical period’, in A. L. Oppenheim (ed.), Ancient Mesopotamia: Portrait of a dead civilization, Chicago: University of Chicago Press, pgs. 335 – 348.

ECS Corpus, Volume I, Part III 782 BUCHANAN, B., 1966 Catalogue of Ancient Near Eastern Seals in the Ashmolean Museum, Volume I, Cylinder Seals, Oxford; New York: Clarendon Press; Oxford University Press. 1981 Early Near Eastern Seals in the Yale Babylonian Collection, New Haven; London: Yale University Press.

CAMERON, G. G., 1936 History of Early Iran, Chicago; London: University of Chicago.

CANAL, D., 1978 ‘La Haute Terrasse de l’Acropole de Suse’, Paléorient 4: 169 – 175.

CARNEGIE, H. (ed.), 1908 Catalogue of the Collection of Antique Gems Formed by James Ninth Earl of Southesk K. T., London.

CARTER, E., 1971 Elam in the Second Millennium B. C.: the Archaeological Evidence, PhD dissertation, University of Chicago. 1979 ‘The Susa Sequence 3000 – 2000 BC: Susa Ville Royale I’, AJA 83: 451 – 454. 1980 ‘Excavations in the Ville Royale I at Susa: The Third Millennium B. C. Occupation’, CDAFI 11: 7 – 134. 1985 ‘Notes on archaeology and the social and economic history of Susiana’, Paléorient 11: 43 – 48. 1990 ‘Elamite Exports’, MJP, pgs. 89 – 100. 1996 Excavations at Anshan (Tal-e Malyan): The Middle Elamite Period, Philadelphia: The University Museum of Archaeology and Anthropology, University of Pennsylvania. 1999 ‘Kabnak (Haft Tappeh) and Al Untash-Napirisha (): Elamite Cities of the Late ’, in A. Alizadeh, Y. Majidzadeh and S. M. Sharmirzadeh (eds.), The Iranian World: Essays on Iranian Art and

ECS Corpus, Volume I, Part III 783 Archaeology Presented to Ezat O. Negahban, Tehran: Iran University Press, pgs. 114 – 130. 2001 ‘Review: Chogha Mish’, JAOS 121(2): 311 – 314.

CARTER, E. and STOLPER, M. W., 1984 Elam – Surveys of Political History and Archaeology, Berkeley: University of California Publications; Near Eastern Studies 25, University of California Press.

CLERCQ, H. F. X. de and MENANT, J., 1888/1903 Collection de Clercq, catalogue méthodique et raisonné, antiquités, assyriennes, cylindres orientaux, cachets, briques, bronzes, bas-reliefs etc., 2 vols., Paris.

COLBOW, G., 1995 Die Spataltbabylonische Glyptik Sudbabyloniens, Munich: Müncher Vorderasiatische Studien 18, Profil Verlag.

COLE, S. W. and DE MEYER, L., 1999 ‘Tepti-ahar, King of Susa, and Kadašman-dKUR.GAL’, Akkadica 112: 44 – 45.

COLLINS, P., 2000 The Uruk Phenomenon – the role of social ideology in the expansion of the Uruk culture during the fourth millennium BC, Oxford: BAR International Series 900.

COLLON, D., 1975 The Seal Impressions from Tell Atchana/Alalakh, Kevelaer: Butzon & Bercker. 1982a Catalogue of Western Asiatic Seals in the British Museum – Cylinder Seals II, Akkadian – Post-Akkadian – Ur III Periods, London: British Museum.

ECS Corpus, Volume I, Part III 784 1982b The Alalakh Cylinder Seals – a new catalogue of actual seals excavated by Sir Leonard Woolley at Tell Atchana, and from neighbouring sites on the Syrian- Turkish border, Oxford: BAR, International Series 132. 1986 Catalogue of Western Asiatic Seals in the British Museum – Cylinder Seals III, Isin-Larsa and Old Babylonian Periods, London: British Museum. 1990 Near Eastern Seals, London: British Museum Publications. 1997 ‘Ancient Near Eastern Seals’, in D. Collon (ed.), 7000 Years of Seals, London: the British Museum Press, pgs. 11 – 30. 1998 ‘Review of Ricerche sui sigilli a cilindro Vicino-Orientali del periodo di Uruk/Jemdet Nasr by E. Rova’, JAOS 117(4): 730 – 732. 2005 First Impressions – Cylinder Seals in the Ancient Near East, rev. ed. (1987), London: British Museum Publications.

CONNAN, J., 1999 ‘Use and Trade of Bitumen in Antiquity and Prehistory: Molecular Archaeology Reveals Secrets of Past Civilizations’, Philosophical Transactions: Biological Scienes 354(1379): 33 – 50.

CONNAN, J. and DESCHESNE, O., 1996 Le Bitume à Suse – Collection du Musée du Louvre, Paris: Éditions de la Réunion des musées nationaux et Elf Aquitaine Productions. 1998 ‘Un matériau énigmatique: le mastic de bitume de Suse – La coupe à la gazelle, La coupe-mortier’, Techne 7: 13 – 16.

CONTENAU, G. and GHIRSHMAN, R., 1935 Fouilles du Tépé-Giyan près de Néhavend de style oriental du Musée du Louvre, vol. I – Fouilles et Missions, Paris: Librairie Orientaliste Paul Geuthner.

DAMEROW, P. and ENGLUND, R. K., 1989 The Proto-Elamite Texts from Tepe Yahya, Cambridge, Mass.: The American School of Prehistoric Research Bulletin 39; the Peabody Museum of Archaeology and Ethnology, Harvard University.

ECS Corpus, Volume I, Part III 785 DE GRAEF, K., 2005 Les archives d’Igibuni. Les documents Ur III du chantier B à Suse, Gand: Université de Gand, MDP 54. 2006 De la dynasite Simaški au Sukkamahat. Les documents fin PE IIB – début PE III du Chantier B à Suse, Gand: Université de Gand, MDP 55. 2008 ‘Annus Simaškensis, l’usage des noms d’année pendant la période Simaškéene (ca. 1930 – 1888 AV. Notre ère) à Suse’, IA 43: 67 – 87.

DELAPORTE, L., 1909 Catalogue du Musée Guimet – Cylindres orientaux, Paris: Annales du Musée Guimet 33. 1910 Catalogue des Cylindres Orientaux et des Cachets Assyro-Babyloniens, Perses et Syro-Cappadociens de la Bibliothèque Nationale, Paris. 1920 Catalogue des Cylindres Orientaux – cachets et pierres gravées de style oriental du Musée du Louvre, vol. I – Fouilles et Missions, Paris: Librairie Hachette. 1923 Catalogue des Cylindres Orientaux, vol. II – Acquistions, Paris: Librairie Hachette.

DELOUGAZ, P. and KANTOR, H. J., 1996 Chogha Mish. Volume I. The First Five Seasons Of Excavations 1961 – 1971, Chicago: the Oriental Institute of the University of Chicago.

DESCHESNE, O., 2003 ‘Le Mastic de Bitume: Un Matériau Remis en Question’, IA 38: 25 – 40.

DIEULAFOY, M., 1893 L’acropole de Suse d’après les fouilles executes en 1884, 1885, 1886 sous les auspices du Musée du Louvre, Paris: Librairie Hachette.

DIJK, J. van, 1978 ‘Išbi’erra, Kindattu, l’homme d’Elam, et la chute de la ville d’Ur’, JCS 30: 189 – 208.

ECS Corpus, Volume I, Part III 786 DITTMANN, R., 1986a ‘Seals, Sealings and Tablets – Thoughts on the Changing Pattern of Administrative Control from the Late-Uruk to the Proto-Elamite Period at Susa’, in U. Finkbeiner and W. Röllig (eds.), Ğamdat Nasr: Period or Regional Style?, Wiesbaden: TAVO Beiheft B 62, pgs. 332 – 366. 1986b ‘Susa in the Proto-Elamite Period and Annotations on the Painted of Proto-Elamite Khuzestan’, in U. Finkbeiner and W. Röllig (eds.), Ğamdat Nasr: Period or Regional Style?, Wiesbaden: TAVO Beiheft B 62, pgs. 171 – 198.

DUCHENE, J., 1986 ‘La Localisation de Huhnur’, in L. De Meyer, H. Gasche and F. Vallat (eds.), FHE, pgs. 65 – 74.

DUNHAM, S., 1985 ‘The Monkey in the Middle’, ZA 75: 234 – 264.

DURAND, J. – M., 1986 ‘Fragments rejoints pour une histoire Elamite’, in L. De Meyer, H. Gasche and F. Vallat (eds.), FHE, pgs. 111 – 128.

DURING CASPERS, E. C. L., 1970 – 71 ‘Some motifs as evidence for maritime contact between Sumer and the Indus Valley’, Persica, pgs. 107 – 118.

DYSON, R. H. Jr., and HARRIS, M. V., 1986 ‘The Archaeological Context of Cylinder Seals Excavated on the Iranian Plateau’, in M. Kelley-Buccellati (ed.), Insight Through Images – Studies in Honour of Edith Porada, Malibu: Bibliotheca Mesopotamica 110, pgs. 79 – 110.

ECS Corpus, Volume I, Part III 787 EDZARD, D. O., 1967 ‘The Early Dynastic Period’, in J. Bottéro, E. Cassin and J. Vercoutter (eds.), The Near East: the Early Civilizations, New York: Delacorte Press, pgs. 133 – 176.

EISEN, G. A., 1940 Ancient Oriental Cylinder and Other Seals with a Description of the Collection of Mrs. William H. Moore, Chicago: the University of Chicago Oriental Institute Publications 47.

ELLIS, R. S., 1995 ‘The Trouble with ‘Hairies’’, Iraq 57: 159 – 165.

ENGLUND, R. K. and GRÉGOIRE, J. –P., 1991 The Proto-Cuneiform Texts from Jemdet Nasr, Berlin: Gebr. Mann Verlag.

FERIOLI, P. and FIANDRA, E., 1979 ‘The Administrative Functions of Clay Sealings in Protohistorical Iran’, in G. Gnoli & A. V. Rossi (eds.), Iranica, Napoli: Istituto Universitario Orientale Seminario di Studi Asiatici, vol.10, pgs. 307 – 312. 1983 ‘Clay Sealings from Arslantepe VIA: Administration and Bureaucracy’, Origini 12: 455 – 509. 1994 ‘Archival Techniques and Methods at Arslantepe’, in P. Ferioli, E. Fiandra, G. Giacomo Fissore & M. Frangipane (eds.), Archives before Writing, Rome: Centro Internazionale di Recerche Archaeologische Anthropologische e Storiche, pgs. 149 – 161.

FIANDRA, E., 1979 ‘The Connection between Clay Seals and Tablets in Administration’, in H. Härtal (ed.), South Asian Archaeology, Berlin: Dietrich Reimer Verlag, pgs. 29 – 43.

ECS Corpus, Volume I, Part III 788 FINKBEINER, U. and RÖLLIG, W., (eds.), 1986 Ğamdat Nasr: Period or Regional Style? Wiesbaden: Beihefte zum TAVO B 62.

FISCHER, C., 1999 ‘The Perils of Perfection’, NABU: 17 – 18.

FRAME, G., 1995 The Royal Inscriptions of Mesopotamia, Babylonian Periods, Volume 2 – Rulers of Babylonia from the Second to the End of Assyrian Domination (1157 – 612 BC), Toronto: University of Toronto Press.

FRANKE, J. A., 1977 ‘Presentation of the Ur III/Isin-Larsa Period’, in McG. Gibson and R. D. Biggs (eds.), Seals and Sealings in the Ancient Near East, Undena: Bibliotheca Mesopotamica 6, pgs. 61 – 66.

FRANKFORT, H., 1939 Cylinder Seals – A Documentary Essay on the Art and Religion of the Ancient Near East, London: Macmillan and Co. 1955 Stratified Cylinder Seals from the Diyala Region, Chicago: OIP 72.

FRANKFORT, H., LLOYD, S. and JACOBSEN, T., 1940 The Gimilson Temple and the Palace of the Rulers at Tell Asmar, Chicago: The University of Chicago Press.

FRAYNE, D., 1990 The Royal Inscriptions of Mesopotamia Early Periods, Volume 4 – Old Babylonian Period (2003 – 1595 BC), Toronto: University of Toronto Press. 1997 The Royal Inscriptions of Mesopotamia Early Periods, Volume 3/2 – Ur III Period (2112 – 2004 BC), Toronto: University of Toronto Press.

ECS Corpus, Volume I, Part III 789 FRIBERG, J., 1978 The Third Millennium Roots of Babylonian Mathematics I. A Method for the Decipherment, through Mathematical and Metrological Analysis, of Proto- Sumerian and Proto-Elamite Semi-pictographic Inscriptions, Gothenburg: Chalmers University of Technology and the University of Gothenburg. 1979 The Early Roots of Babylonian Mathematics II. Metrological Relations in a Group of Semi-pictographic Tablets of the Jemdet Nasr Type, probably from Uruk-Warka, Gothenburg: Chalmers University of Technology and the University of Gothenburg.

GARRISON, M. B., 1989 ‘An Early Dynastic III Seal in the Kelsey Museum of Archaeology: the Relationship of Style and Iconography in Early Dynastic III Glyptic’, JNES 48(1): 1 – 13.

GASCHE, H., 1973 La poterie du deuxième millénaire a. C, Leiden; Paris: MDP 47.

GASCHE, H., ARMSTRONG, J. A., COLE, S. W. and GURZADYAN, V. G., 1998 Dating the Fall of Babylon – a reappraisal of second-millennium chronology (a joint Ghent-Chicago-Harvard Project), Ghent; Chicago: Mesopotamian History and Environment, Series II Memoirs IV, the University of Ghent and the Oriental Institute of the University of Chicago.

GASSAN, M., 1989 ‘Hurpatila, roi d’Elammat’, AION 49(3): 223 – 229.

GELB, I. J., 1977 ‘Typology of Mesopotamian Seal Inscriptions’, in McG. Gibson and R. D. Biggs (eds.), Seals and Sealings in the Ancient Near East, Undena: Bibliotheca Mesopotamica 6, pgs. 107 – 136.

ECS Corpus, Volume I, Part III 790 GELB, I. J. and KIENAST, B., 1990 Die Altakkadischen Königsinchriften des Dritten Jahrtausends V. Chr, Stuttgart: Freiburger altorientalische Studien 6.

GHIRSHMAN, R., 1966 Tchoga Zanbil (Dur-Untash) I. La Ziggurat, Paris: MDP 39. 1968 Tchoga Zanbil (Dur-Untash) II. Temenos, temples, palais, tombes, Paris: MDP 40.

GIBSON, Mc.G. and BIGGS, R. D. (eds.), 1977 Seals and Sealings in the Ancient Near East, Undena: Bibliotheca Mesopotamica 6.

GLASSNER, J. – J., 1991 ‘Les textes de Haft Tépé, la Susiane et l’Elam au 2ème millénaire’, MHEOP I, 109 – 126. 1994 ‘Ruhušak – mar ahatim: la transmission du pouvoir en Elam’, JA 282: 219 – 236.

GOLDBERG, J., 2004 ‘The Berlin Letter, Middle Elamite Chronology and Šutruk-Nahhunte I’s Genealogy’, IA 39: 33 – 42.

GORDON, C. H., 1939 ‘Western Asiatic Seals in the Walters Art Gallery’, Iraq 6: 3 – 34.

GORELICK, L. and GWINNETT, A. J., 1978 ‘Ancient Seals and Modern Science – Using the Scanning Electron Microscope as an Aid in the Study of Ancient Seals’, Expedition 20(2): 38 – 48. 1981a ‘Close Work Without Magnifying Lenses? – A Hypothetical Explanation for the Ability of Ancient Craftsmen to Effect Minute Detail’, Expedition 23(2): 27 – 34.

ECS Corpus, Volume I, Part III 791 1981b ‘The Origin and Development of the Ancient Near Eastern Cylinder Seal’, Expedition 23(4): 17 – 38. 1990 ‘The Ancient Near Eastern Cylinder Seal as Social Emblem and Status Symbol’, JNES 49: 45 – 64. 1992 ‘Minoan Versus Mesopotamian Seals: Comparative Methods of Manufacture’, Iraq 54: 57 – 64.

GWINNETT, A. J. and GORELICK, L., 1979 ‘Ancient Lapidary – A Study Using Scanning Electron Microscopy and Functional Analysis’, Expedition 22(1): 17 – 32. 1987 ‘The Change from Stone Drills to Copper Drills in Mesopotamia’, Expedition 29(3): 15 – 24.

HAERINCK, E., 1986 ‘The Chronology of Luristan, Pusht-i Kuh in the late fourth and first half of the third millennium BC’, in J. –L. Huot (ed.), Préhistoire de la Mésopotamie, Paris: Editions du CNRS, pgs. 55 – 72.

HAERINCK, E. and OVERLAET, B., 2006 Bani Surmah – An Early Bronze Age Graveyard in Pusht-i Kuh, Luristan, Leuven: LED VI. 2008 The Kalleh Nisar Bronze Age Graveyard in Pusht-i Kuh, Luristan, Leuven: LED VII.

HAKEMI, A., 1997 Shahdad – Archaeological Excavations of a Bronze Age Centre in Iran, Translated by S. M. S. Sajjadi, Rome: IsMEO.

HALLO, W. W., 1957 Early Mesopotamian Royal Titles, a Philological and Historical Analysis, New Haven: American Oriental Series 43. 1977 ‘Seals Lost and Found’, in McG. Gibson and R.D. Biggs (eds.), Seals and sealing in the ancient Near East, Malibu: Bibliotheca Mesopotamica 6.

ECS Corpus, Volume I, Part III 792 1987 ‘As the Seal upon thine Arm’: Glyptic Metaphors in the Biblical World’, in L. Gorelick & E. Williams-Forte (eds.), Ancient Seals and the , Malibu: Undena Publications, pgs. 7 – 17.

HALLO, W. W. and WINTER, I. J., (eds.), 2001 Seals and Seal Impressions, Proceedings of the XLVe Rencontre Assyriologique Internationale Pt. II, Bethesda: CDL Press.

HANSEN, D. P., 1971 ‘Some Early Dynastic I Sealings from Nippur’, in D. G. Miten, J. G. Pedley, and J. A. Scott (eds.), Studies Presented to George M. A. Hanfmann, Mainz: Verlag Philipp von Zabern.

HARPER, P. O., ARUZ, J. and TALLON, F. (eds.), 1992 The Royal City of Susa: Ancient Near Eastern Treasures in the Louvre, New York: Metropolitan Museum of Art, H.N. Abrams.

HEINRICH, E., 1931 Ergebnisse der Ausgrabungen der deutschen Orient-Gesellschaft in Fara und Abu Hata 1902/03, Berlin.

HENRICKSON, E. F. 1988 ‘Chalcolithic seals and sealings from Seh Gabi, Central Western Iran’, IA 23: 1 – 19.

HENRICKSON, R. C., 1985 ‘The Chronology of Central Western Iran, 2600 – 1400 B.C.’, AJA 89: 569 – 581.

HERRERO, P., 1976 ‘Tablettes administratives de Haft-Tépé’, CDAFI 6: 93 – 116.

HERZFELD, E. E., 1941 Iran in the Ancient East, New York: Oxford University Press.

ECS Corpus, Volume I, Part III 793 HINZ, W., 1973 Neue Wege im Altpersischen, Wiesbaden: Göttinger Orientforschungen III/I.

HULBUT, C. S., Jr., 1971 Dana’s Manual of Mineralogy, 18th Edition, New York: John Wiley & Sons Inc.

JACOBSEN, T., 1939 The , Chicago: AS 11.

JEAN, C. –F., 1922 ‘L’Élam sous la dynastie d’Ur: les indemnités allouée aux “chargés de mission” des rois d’Ur’, RA 19: 1 – 44.

JOHNSON, G. A., 1973 Local Exchange and Early State Development in Southwestern Iran, Ann Arbor: Anthropological Papers of the Museum of Anthropology 51.

JOHNSTON, R. G., MARTINEZ, D. D. and GARCIA, A. R. E., 2001 ‘Were ancient seals secure?’, Antiquity 75: 299 – 305.

KAWAMI, T. S., 2001 ‘The Cattle of Uruk: Stamp Seals and Animal Husbandry in the Later Uruk/’, in W. W. Hallo and I. J. Winter (eds.), Seals and Seal Impressions, Proceedings of the XLVe Rencontre Assyriologique Internationale, Bethesda, Maryland: CDL Press, pgs. 31 – 47.

KEEL-LEU, H., 2004 The Ancient Near Eastern Cylinder Seals of the Collections “Bible + Orient” of the University of Fribourg, Gottingen: Academic Press, Vandenhoeck & Ruprecht.

ECS Corpus, Volume I, Part III 794 KEISER, C. E., 1971 Neo-Sumerian Account Texts from Drehem, New Haven: Yale University, 3, Babylonian Inscriptions in the Collection of J. B. Nies.

KELLY-BUCCELLATI, M. (ed.), 1986 Insight Through Images – Studies in Honor of Edith Porada, Malibu: Bibliotheca Mesopotamica, vol. 21.

KJÆRUM, P., 1984 The Stamp and Cylinder Seals, Failaka/Dilmun: the second millennium settlements, Aarhus: Jysk arkæologisk selskab, vol. 1. 1986 ‘The Dilmun seals as evidence of long distance relations in the early second millennium BC’, in H. A. A. Khalifa & M. Rice (eds.), Bahrain through the Ages: the Archaeology, London: Distributed by Routledge & Kegan Paul, pgs. 269 – 271.

KORDEVANI, M., 1971 ‘Excavations at Chogha Gavaneh’, Bastan Chenassi va Honar-e Iran 7 – 8: 36 – 71. (in Persian).

LABAT, R., 1975a ‘Elam, c. 1600-1200 B.C.’, in CAH II/2, pgs. 482 – 506. 1975b ‘Elam and western Persia, c. 1200-1000 B.C.’, in CAH II/2, pgs. 482 – 506.

LÆSSØE, J., 1965 ‘IM 62100: A Letter from Tell Shemshara’, in H. G. Gütterbock and Th. Jacobsen (eds.), Studies in Honor of Benno Landsberger on his Seventy-fifth Birthday, April 21, 1965, Chicago: University of Chicago Press, pgs. 189 – 196.

LAHANIER, C., 1976 ‘Note sue l’emploi de l’heulandite et de la mordenite dans la fabrication de sceaux cylindres proto-élamites’, Annales du Laboratoire de Recherche des Musées de France, 65 – 66.

ECS Corpus, Volume I, Part III 795 1977 ‘Analyse d’objects en bitume provenant de Suse’, Annales du Laboratoire de Recherche des Musées de France, 48 – 68.

LAMBERT, M., 1966 ‘Cylindres de Suse des Premiers Temps des Sukalmah’, IA 6: 34 – 42. 1971 ‘Investiture de fonctionnaires en Elam’ JA 259: 217 – 221.

LAMBERT, W. G., 1966 ‘Ancient Near Eastern Seals in Birmingham Collections’, Iraq 28: 64 – 83. 1979 ‘Near Eastern Seals in the Gulbenkian Museum of Oriental Art, University of Durham’, Iraq 41: 1 – 45. 1992 ‘Further notes on a Seal of Ebarat’, RA 86: 85 – 86

LAYARD, A. H., 1846 ‘Description of the Plain of Khuzistan’, Journal of the Royal Asiatic Society 16: 78 – 86.

LE BRUN, A., 1978 ‘La Glyptique du Niveau 17B de l’Acropole (Campagne de 1972)’, CDAFI 8: 11 – 59. 1985 ‘Le niveau 18 de l’acropole de Suse. Mémoire d’argile, mémoire du temps’, Paléorient 11: 31 – 36.

LE BRUN, A. and VALLAT, F., 1978 ‘L’origine de l’écriture à Suse’, CDAFI 8: 11 – 59.

LEGRAIN, L., 1911 Catalogue des Cylindres Orientaux de la Collection Louis Cugnin, Paris: Honore Champion. 1921 Empreintes de Cachets Élamites, Paris: MDP 16. 1925 The Culture of the Babylonians from their Seals in the Collections of the Museum, Philadelphia: University Museum 1936 Archaic Seal-Impressions Ur Excavations III, Philadelphia: the British Museum and the University Museum.

ECS Corpus, Volume I, Part III 796 1951 Seal Cylinders, Ur Excavations X, Philadelphia: the British Museum and the University Museum of Pennsylvania.

MALBRAN-LABAT, F., 1995 Les inscriptions royales de Suse: Briques de l’époque paléo-élamite à l’Empire néo-élamite, Paris: Éditions de la Réunion des musées nationaux.

MARTIN, H. P., 1988 Fara: A Reconstruction of the Ancient Mesopotamian City of Shuruppak, Birmingham.

MARTIN, H. P. and MATTHEWS, R. J., 1993 ‘Seals and sealings’, in A. Green (ed.), Excavations 4. The 6G Ash-tip and its Contents: Cultic and Administrative Discard from the Temple?, London: British School of Archaeology in Iraq.

MATTHEWS, D. M., 1990 Principles of Compositions in Near Eastern Glyptic of the Later Second Millennium B.C., Fribourg & Göttingen: Orbis Biblicus et Orientalis, Series Archaeologica 8. 1992 The Kassite Glyptic of Nippur, Fribourg: Orbis Biblicus et Orientalis 116, Vandenhorck & Ruprecht. 1997 The Early Glyptic of Nippur, Fribourg & Göttingen: Orbis Biblicus et Orientalis, Series Archaeologica 15.

MATTHEWS, R. J., 1992 ‘Defining the Style of the Period: Jamdet Nasr 1926 – 1928’, Iraq 54: 1 – 34. 1993 Cities, Seals and Writing: Archaic Seal Impressions from Jemdet Nasr and Ur, Berlin: Mann.

MAZZONI, S., 1984 ‘Seal-impressions on jars from in EB I (V) A – B’, Akkadica: 18 – 45.

ECS Corpus, Volume I, Part III 797 MECQUENEM, R. DE, CONTENAU G., PFISTER, R. and BELAIEW, N., 1943 Archéologie Susienne, Paris: MDP 29.

MECQUENEM, R. DE, 1922 ‘Inventaires de Cylindres et Cachets’, RA 19: 165 – 153. 1925 ‘Inventaire de Cachets et de Cylindres (Suse, 1923 – 1924)’, RA 22: 1 – 11. 1927 ‘Inventaire de Cachets et Cylindres (Suse, 1925 – 1926)’, RA 24: 7 – 21. 1928 ‘Choix d’Intailles Susiennes’, RA 25: 169 – 174. 1937 ‘Vases Susiens a Personages’, RA 34: 149 – 153. 1943 ‘Les Fouilles de Suse 1929 – 1933’, MDP 29, pgs. 5 – 150. 1949 Épigraphie Proto-Élamite, Paris: MDP 31

MECQUENEM, R. DE and MICHALON, J., 1953 Recherches à Tchogha Zembil, Paris: MDP 33.

MÉNANT, M. J., 1883 – 1886 Les pierres gravées de la Haute-Asie, recherches sur la glyptique orientale, 2 vols, Paris.

MEYER, L. DE, 1986 ‘Les archives d’Igibuni’, FHE, pgs. 75 – 77.

MICHALOWSKI, P., 1978 ‘Foreign tribute to Sumer during the Ur III period’, ZA 68: 34 – 49. 1989 The Lamentation over the Destruction of Sumer and Ur, Winona Lake: Eisenbrauns. 2008 ‘Observations on “Elamite” and “Elam” in Ur III Times’, in P. Michalowski (ed.), On the – Studies in Honor of Marcel Sigrist, Boston: American Schools of Oriental Research, pgs. 109 – 123.

MICHALOWSKI, P. (ed.), 2008 On the Third Dynasty of Ur – Studies in Honor of Marcel Sigrist, Boston: American Schools of Oriental Research.

ECS Corpus, Volume I, Part III 798 MIROSCHEDJI, P. De, 1978 ‘Stratigraphie de la période néo-élamite à Suse’, Paléorient 4: 213 – 227. 1981 ‘Le dieu élamites au serpent et aux eaux jaillissantes’, IA 16: 1-25. 2003 ‘Susa and the Highlands – Major Trends in the History of Elamite Civilization’, in N. F. Miller and K. Abdi (eds.), Yeki bud, yeki nabud – essays on the archaeology of Iran in honor of William M. Sumner, Los Angeles: The Costen Institute, University of California, pgs. 17 – 35.

MOFIDI NASRABADI, B., 2003-04a ‘Untersuchungen zu Siedlungsstrukturen in der Peripherie von Čoga Zanbil (Dur Untaš)’, AMI 35 – 36: 241 – 265. 2003-04b ‘Archäologische Untersuchungen in Haft Tape (Iran), AMI 35 – 36: 225 – 239. 2007 Archäologische Ausgrabungen und Untersuchungen in Čoga Zanbil, Münster.

MØLLER, E., 1992 Ancient Near Eastern Seals in a Danish Collection, Copenhagen: Carsten Niebuhr Institute Publications.

MOOREY, P. R. S., 1994 Ancient Mesopotamian Materials and Industries: the archaeological evidence, Oxford; New York: Clarendon Press.

MOOREY, P. R. S. and GURNEY, O. R., 1978 ‘Ancient Near Eastern Cylinder Seals acquired by the Ashmolean Museum, Oxford 1963 – 1973’, Iraq 40: 41 – 60.

MOORTGAT, A., 1940 Vorderasiatische Rollsiegel, Berlin.

MOUSAVI, A., 1996 ‘Early Archaeological Adventures and Methodological Problems in Iranian Archaeology: the Evidence from Susa’, IA 31: 1-16.

ECS Corpus, Volume I, Part III 799 MUNN-RANKIN, J. M., 1959 ‘Ancient Near Eastern Seals in the Fitzwilliam Museum, Cambridge’, Iraq 21: 20 – 37.

MUSCARELLA, O. W., 1981 ‘Surkh Dum at the Metropolitan Museum of Art: a Mini-Report’, JFA 8(3): 327 – 359.

NEGAHBAN, E. O., 1969 ‘Haft Tepe’, Iran 7: 173 – 177. 1990 ‘The Haft Tepe Bronze Plaque – an example of Middle Elamite art’, MJP, 137 – 142. 1991 Excavations at Haft Tepe, Iran, Philadelphia: The University Museum of Archaeology and Anthropology, University of Pennsylvania. 1994 ‘The Artist’s Workshop of Haft Tepe’, in CDR, pgs. 31 – 41. 1996 Marlik, the Complete Excavation Report, Philadelphia: The University Museum, University of Pennsylvania.

NICHOLAS, I. M., 1990 The Proto-Elamite Settlement at TUV, Philadelphia: The University Museum of Archaeology and Anthropology, University of Pennsylvania.

NICHOLSON, P. T. and PELTENBURG, E., 2000 ‘Egyptian faience’, in P. Nicholson & I. Shaw (eds.), Ancient Egyptian materials and technology, Cambridge: Cambridge University Press, pgs. 177 – 199.

NIJHOWNE, J., 1999 Politics, religion and cylinder seals: a study of Mesopotamian symbolism in the second millennium B. C., Oxford: J. and E. Hedge, Archeopress.

ECS Corpus, Volume I, Part III 800 NISSEN, H. J., 1977 ‘Aspects of the Development of Early Cylinder Seals’, in McGuire Gibson and R. D. Biggs (eds.), Seals and Sealing in the Ancient Near East, Malibu: Bibliotheca Mesopotamica 6, Undena Publications, pgs. 15 – 23. 1986 ‘The development of writing and glyptic art’, in U. Finkbeiner and W. Röllig (eds.), Ğamdat Nasr – Period or Regional Style?, Wiesbaden: TAVO Beiheft B 62, pgs. 316 – 331.

NOVECK, M., 1975 The Mark of Ancient Man – Ancient Near Eastern Stamp Seals and Cylinder Seals: the Gorelick Collection, New York: The Brooklyn Museum.

OATES, J., 1993 ‘Trade and power in the fifth and fourth millennia BC: new evidence from northern Mesopotamia’, World Archaeology 24: 403 – 422.

OSTEN, H. H., von der, 1934 Ancient Oriental Seals in the Collection of Mr. Edward T. Newell, Chicago: OIP 22. 1936 Ancient Oriental Seals in the Collection of Mrs. Agnes Baldwin Brett, Chicago: OIP 37. 1957 Altorientalische Siegelsteine der Sammlung Hans Silvius von Aulock, Sweden: Uppsala.

OTTO, A., 2000 Die Entstehung und Entwicklung der Klassich-Syrischen Glytik, Berlin.

OWEN, D., 1975 The John Frederick Lewis Collection, Rome. 1982 Neo-Sumerian Archival Texts primarily from Nippur, Winona Lake: Eisenbrauns.

ÖZGÜÇ, N., 1965 The Anatolian Group of Cylinder Seals from Kültepe, Ankara.

ECS Corpus, Volume I, Part III 801 1968 Seals and Seal Impressions from Level Ib from Karum Kanish, Ankara.

PARKER, B., 1962 ‘Seals and seal impressions from the Nimrud excavations, 1955 – 58’, Iraq 24: 26 – 40.

PÉZARD, M., 1912 ‘Etude sur les intailles susiennes’, Paris: MDP 12.

PITTMAN, H., 1993 ‘Pictures of an Administration: the Late Uruk Scribe at Work’, in M. Frangipane, H. Hauptmann, M. Liverani, P. Matthiae & M. Mellink (eds.), Between the Rivers and Over the Mountain – Archaeologica Anatolica et Mesopotamica Alba Palmieri Dedicata, Rome: Dipartimento di Scienze Storiche Archeologiche e Antropologiche dell’Antichita, Universita di Roma, “La Sapienza”, pgs. 236 – 245. 1994 The Glazed Steatite Style: Structure and Function of an Image System in the Administration of Protoliterate Mesopotamia, Berlin: BBVO 16. 1997 ‘The Administrative Function of Glyptic Art in Proto-Elamite Iran: a Survey of the Evidence’, Res Orientales 10: 133 – 161. 2002 ‘The “Jeweler’s” Seal from Susa and Art of Awan’, in E. Ehrenberg (ed.), Leaving No Stones Unturned – Essays on the Ancient Near East and Egypt in Honor of Donald P. Hansen, Winona Lake: Eisenbrauns, pgs. 211 – 235. 2003 ‘The Proto-Elamite Period’, in J. Aruz (ed.), 2003, Art of the First Cities: the Third Millennium B. C. from the Mediterranean to the Indus, New York: Metropolitan Museum of Art, pgs. 42 – 55.

PITTMAN, H. and ARUZ, J., 1987 Ancient Art in Miniature: Near Eastern Seals from the Collection of Martin and Sarah Cherkasky, New York: Metropolitan Museum of Art.

POLLOCK, S., 1991 ‘Of priestesses, princes and poor relations: the dead in the Royal Cemetery of Ur’, CAJ 1: 171 – 189.

ECS Corpus, Volume I, Part III 802 PONS, N., 1994 ‘Tchoga Zanbil Apres Untaš-Napiriša’, in CDR, pgs. 43 – 51.

PORADA, E., 1948 Corpus of Ancient Near Eastern Seals in North American Collections, The Collection of the Pierpont Morgan Library, Washington: The Bollingen Series 14. 1950 ‘A Leonine Figure of the Protoliterate Period of Mesopotamia’, JAOS 70: 223 – 226. 1957 ‘Syrian Seal Impression on Tablets Dated in the Time of Hammurabi and Samsu-Iluna’, JNES 16: 192 – 197. 1965 Ancient Iran: the Art of Pre-Islamic Times, New York, Crown. 1970 Tchoga Zanbil (Dur Untash). Vol. IV: La Glyptique, Paris: MDP 42. 1974 – 77 ‘Die Siegelzylinder-Abrollung auf der Amarna-Tafel’, AfO 25: 132 – 142. 1981 ‘The Cylinder Seals Found at Thebes in Boeotia’, AfO 28: 1 – 78. 1982 ‘Problems of Method in the Archaeology and Art History of the Ancient Near East’, JAOS 102(3): 501 – 506. 1995 ‘Understanding Ancient Near Eastern Art: A Personal Account’, in CANE, pgs. 2695 – 2714.

POTTS, D. T., 1990a ‘Notes on Some Horned Buildings in Iran, Mesopotamia and Iran’, RA 84: 33 – 40. 1990b The Arabian Gulf in Antiquity, Oxford, New York: Clarendon Press. 1997 Mesopotamian Civilization: The Material Foundation, London: Athlone Press. 1999 The Archaeology of Elam – Formation and Transformation of an Ancient Iranian State, Cambridge: Cambridge University Press. 2001 Excavations at Tepe Yahya, Iran, 1967 – 1975: the third millennium, Cambridge, Mass.: Peabody Museum of Archaeology and Ethnology, Harvard University. 2006 ‘Elamite and Kassites in the ’, JNES 65(2): 111 – 119.

ECS Corpus, Volume I, Part III 803 QUINTANA, E., 1998 ‘De Nuevo sobre la dinastia elamita de Šimaški’, NABU, 4.

RAPP, G. R., 2002 Archaeomineralogy, Berlin; New York: Springer Verlag.

RAPP, G. R. and HILL, C. L., 1998 Geoarchaeology – The Earth-Science Approach to Archaeological Interpretation, New Haven: Yale University Press.

RASHAD, M., 1990 Die Entwicklung der Vor- und Frühgeschichtlichen Stempelsiegel in Iran, Berlin: Archäologische Mitteilungen aus Iran Ergänzungsband 13, Dietrich Reimer Verlag.

RATHJE, W. L., 1977 ‘New Tricks for Old Seals: A Progress Report’, in McG. Gibson and R. D. Biggs (eds.), Seals and Sealing in the Ancient Near East, Malibu: Bibliotheca Mesopotamica 6, Undena Publications, pgs. 25 – 32.

RAVN, O. E., 1960 A Catalogue of Oriental Cylinder Seals and Impressions in the Danish National Museum, Kobenhavn: Nationalmuseet.

READ, H. H., 1970 Rutley’s Elements of Mineralogy, 26th edition, London: Thomas Murby Co.

READE, J., 2003 ‘The Royal Tombs of Ur’, in J. Aruz (ed.), 2003, Art of the First Cities: the Third Millennium B. C. from the Mediterranean to the Indus, New York: Metropolitan Museum of Art, pgs. 93 – 132.

REINER, E., 1963 ‘Malamir’, RA 57: 169 – 174.

ECS Corpus, Volume I, Part III 804 1973a ‘The location of Anšan’, RA 67: 57 – 62. 1973b ‘Inscription from a Royal Elamite tomb’, AfO 24: 87 – 202.

RENGER, J., 1977 ‘Legal Aspects of Sealing in Ancient Mesopotamia’, in McG. Gibson and R. D. Biggs (eds.), Seals and Sealings in the Ancient Near East, Undena: Bibliotheca Mesopotamica 6, pgs. 73 – 88.

ROAF, M., 1990 Cultural Atlas of Mesopotamia and the Ancient Near East, New York: Facts on File.

ROBERTS, W. L., CAMPBELL, T. J. and RAPP, G. R., 1990 Encyclopedia of Minerals, Second Edition, New York: Van Nostrand Reinhold Company.

ROTHMAN, M. S., 1994a ‘Seal and Sealing Findspot, Design Audience, and Function: Monitoring Changes in Administrative Oversight and Structure at during the Fourth Millennium B.C.’, in P. Ferioli, E. Fiandra, G. Giacomo Fissore and M. Frangipane (eds.), Archives before Writing, Rome: Centro Internazionale di Recerche Archaeologische Anthropologische e Storiche, pgs. 97 – 119. 1994b ‘Sealings as a Control Mechanism in Prehistory’, in G. Stein and M. S. Rothman (eds.), Chiefdoms and Early States in the Near East, Madison, Wi.: Prehistory, pgs. 103 – 120. 2007 ‘The Archaeology of Early Administrative Systems in Mesopotamia’, in E. C. Stone (ed.), Settlement and Society: Essays Dedicated to Robert McCormick Adams, Los Angeles: Costen Institute of Archaeology, University of California, the Oriental Institute of the University of Chicago, pgs. 235 – 254.

ROTHMAN, M. S. (ed.), 2001 Uruk Mesopotamia and Its Neighbours: Cross-cultural interactions and their consequences in the era of state formation, Oxford; Santa Fe; School of American Research and James Currey.

ECS Corpus, Volume I, Part III 805 ROVA, E., 1994 Ricerche sui sigilli a cilindro Vicin-Orientali del periodo di Uruk/Jemdet Nasr, Rome: Orientis Antiqui Collectio, 20, Istituto per l’Oriente C. A. Nallino.

RUTTEN, M., 1949 Archéologie Susienne, Paris: MDP 31. 1950 ‘Glyptique Susienne – Empreintes et Cylindres’, RA 44: 161 – 178.

SAX, M., McNABB, J. and MEEKS, N. D., 1998 ‘Methods of Engraving Mesopotamian Cylinder Seals: Experimental Confirmation’, Archaeometry 40(1): 1 – 21.

SAX, M. and MEEKS, N. D., 1994 ‘The Introduction of Wheel Cutting as a Technique for Engraving Cylinder Seals: Its Distinction from Filing’, Iraq 56: 153 – 166. 1995 ‘Methods of Engraving Mesopotamian Quartz Cylinder Seals’, Archaeometry 37(1): 25 – 36.

SAX, M., MEEKS, N. D. and COLLON, D., 2000 ‘The Introduction of the Lapidary Engraving Wheel in Mesopotamia’, Antiquity 74: 380 – 387.

SAX, M. and MIDDLETON, P., 1992 ‘A System of Nomenclature for Quartz and Its Application to the Material of Cylinder Seals’, Archaeometry 34(1): 11 – 20.

SCHACHT, R. M., 1975 ‘A Preliminary Report on the Excavations at Tepe Sharafabad, 1971’, JFA 2(4): 307 – 329.

SCHAEFFER-FORRER, C. F. –A., 1983 Corpus I des cylindres-sceaux de Ras Shamra- et d’Enkomi-Alasia, Paris: Editions Recherche sur les Civilisations, ‘synthèse’ no.13.

ECS Corpus, Volume I, Part III 806 SCHEIL, V., 1990 Textes élamites-sémitiques, première série, Paris: MDP 2. 1902 Textes élamites-sémitiques, deuxième série, Paris: MDP 4. 1905 Textes Élamites-Sémitiques, troisième série, Paris, MDP 6. 1923 Textes de Comptabilité Proto-Élamites, Paris: MDP 17. 1927 ‘Raptim’, RA 23: 35 – 48. 1931 ‘Dynasties Élamites d’Awan et de Simaš’, RA 28: 1 – 8. 1935 Textes de Comptabilité Proto-Élamite – Troisième Série, Paris: MDP 26. 1939 ‘Actes Juridiques Susiens’, in Mélanges Epigraphiques, Paris: MDP 28, pgs. 37 – 161.

SCHMANDT-BESSERAT, D., 1978 ‘The Earliest Precursor of Writing’, Scientific American 238(6): 38 – 47. 1979 ‘An Archaic Recording System in the Uruk-Jemdet Nasr Period’, AJA 83: 19 – 48. 1981 ‘From Tokens to Tablets: a Re-Evaluation of the So-Called “Numerical Tablets”’, Visible Language 15(4): 321 – 344. 1986 ‘Tokens at Susa’, OrAnt 25” 93 – 125. 1994 ‘Token: A Prehistoric Archive System’, in P. Ferioli, E. Fiandra, G. G. Fissore and M. Frangipane (eds.), Archives Before Writing, Archives before Writing, Rome: Centro Internazionale di Recerche Archaeologische Anthropologische e Storiche, pgs. 13 – 33.

SCHMIDT, E. F., 1940 Flights Over Ancient Cities of Iran, Chicago: University of Chicago Press.

SCHMIDT, E. F., van LOON, M. N. and CURVERS, H. H., 1989 The Holmes Expedition to Luristan, Chicago: The Oriental Institute of the University of Chicago.

SEIDL, U., 1990 ‘Altelamische Siegel’, MJP, 129 – 135.

ECS Corpus, Volume I, Part III 807 SIGRIST, M., 1992 Drehem, Bethesda: CDL Press.

SIGRIST, M. and GOMI, T., 1991 The Comprehensive Catalogue of Published Ur III Tablets, Bethesda: CDL Press.

SOLLBERGER, E. and KUPPER, J. –R., 1971 Inscriptions Royales Sumeriennes et Akkadiennes, Paris: Littératures du Proche-Orient 3.

SOMMERFIELD, W., 1995 ‘The Kassites of Ancient Mesopotamia: Origins, Politics, and Culture’, in CANE, pgs. 917 – 930.

SPELEERS, L., 1917 Catalogue des Intailles et Empreintes Orientales des Musées Royaux du Cinquantenaire, Bruxelles: Vromant & Co. Imprimeurs. 1943 Catalogue des intailles et empreintes orientales des Musées Royaux d’Art et d’Histoire, Supplément, Bruxelles: Vromant & Co. Imprimeurs.

SPYCKET, A., 1960 ‘La Déesse Lama’, RA 54: 73 – 84. 1992 Les figurines de Suse, Paris: MDP 52.

STEIN, D., 1993 Das Arciv des šilwa-Teššub 8. The seal impressions (2 vols.), Wiesbaden.

STEINKELLER, P., 1977 ‘Seal Practice in the Ur III Period’, in McG. Gibson and R. D. Biggs (eds.), Seals and Sealings in the Ancient Near East, Undena: Bibliotheca Mesopotamica 6, pgs. 41 – 53. 1988 ‘On the identity of the toponyms LÚ.SU(.A)’, JAOS 108: 197 – 202.

ECS Corpus, Volume I, Part III 808 1993 ‘Settlement patterns and material culture of the Akkadian period: continuity and discontinuity’, in M. Liverani (ed.), Akkad, the First World Empire: Structure, Ideology, Traditions, Padua: History of the Ancient Near East, pgs. 91 – 129. 2007 ‘New Light on Šimaški and its Rulers’, ZA 97: 215 – 232.

STERN, E. M., 1995 Roman mold-blown glass: the first through sixth centuries, Rome: “L’Erma” di Bretschneider in association with the Toledo Museum of Art.

STEVE, M. –J., 1967 Tchoga Zanbil (Dur-Untash) III. Textes élamites et accadiens de Tchoga Zanbil, Paris: MDP 41. 1968 ‘Fragmenta Elamica’, Or 37: 290 – 307. 1987 Nouveaux mélanges épigraphiques, Inscriptions royales de Suse et de la Susiane, Nice: MDP 53. 1989 ‘Des Sceaux-Cylindres de Simaški?’, RA 83: 13 – 26. 1991 ‘Elam: histoire continue ou discontinue?’, MHEOP 1, 1 – 9. 1992 Syllabaire élamite: Histoire et paléographie, vol. 1, Neuchâtel-Paris: Civilisations du Proche-Orient, Ser. II, Philogie.

STEVE, M. –J. and GASCHE, H., 1971 L’Acropole de Suse, Paris and Leiden: MDP 46. 1990 ‘Le tell de l’Apadana avant les Achéménides: Contribution à la topographie de Suse’, MJP, 15 – 60.

STEVE, M. –J. and VALLAT, F., 1989 ‘La Dynastie des Igihalkides: Nouvelles Interpretations’, AIO, pgs. 223 – 238.

STEVE, M. –J., GASCHE, H. and MEYER, L. DE, 1980 ‘La Susiane au deuxième millénaire: à propos d’une interpretation des fouilles de Suse’, IA 15: 49 – 154.

ECS Corpus, Volume I, Part III 809 STOLPER, M. W., 1982 ‘On the Dynasty of Šimaški and the Early Sukkalmahs’, ZA 72: 42 – 67. 1987 – 90 ‘Malamir. B. Philologische’, RIA 7: 276 – 280. 2003 ‘Three Stray Elamite Tablets from Malyan’, in N. F. Miller and K. Abdi (eds.), Yeki bud, yeki nabud – essays on the archaeology of Iran in honor of William M. Sumner, Los Angeles: The Costen Institute, University of California, pgs. 201 – 206.

SUMNER, W. M., 1974 ‘Excavations at Tall-i Malyan, 1971 – 72’, Iran 12: 155 - 179. 1976 ‘Excavations at Tall-i Malyan (Anshan) 1974’, Iran 14: 103 – 115. 1985 ‘The Proto-Elamite city wall at Tal-i Malyan’, Iran 23: 153 – 161. 1986 ‘Proto-Elamite civilization in Fars’, in U. Finkbeiner and W. Röllig (eds.), Ğamdat Nasr – Period or Regional Style?, Wiesbaden: TAVO Beiheft B 62, pgs. 199 – 211. 1988 ‘Maljan, Tall-e (Anšan)’, RIA 7: 306 – 320. 1989 ‘Anshan in the Kaftari phase: pattern of settlement and land use’, AIO, pgs. 135 – 161. 1997 ‘Malyan’, in E. M. Meyers (ed.), The Oxford Encyclopaedia of Archaeology in the Near East, Volume 3, New York: Oxford University Press, pgs. 406 – 409. 2003 Early Urban Life in the Land of Anshan: Excavations at Tal-e Malyan in the Highlands of Iran, Philadelphia: The University of Pennsylvania Museum of Archaeology and Anthropology; Chicago: The Oriental Institute, University of Chicago.

TEISSIER, B., 1984 Ancient Near Eastern Cylinder Seals from the Marcopoli Collection, Berkeley. 1994 Sealing and Seals on Texts from Kültepe Kārum Level 2, Leiden: Nederlands Historisch-Archaeologisch Instituut te Istanbul LXX.

TOBLER, A. J., 1950 Excavations at Tepe Gawra II, Philadelphia.

ECS Corpus, Volume I, Part III 810 TOUROVETS, A., 1996 ‘La Glyptique de Bani Surmah, Pusht-i Kuh-Luristan’, IA 31: 19 – 45. 1997 ‘Observations concernant l’existence d’une ancienne voie cérémonielle au nord-est du site de Tchoga Zanbil’, IA 32: 71 – 90.

VALLAT, F., 1986 ‘The most ancient scripts of Iran: the current situation’, World Archaeology 17: 335 – 347. 1989a ‘Le scribe Ibni-Adad et les premiers sukkalmah’, NABU, 34. 1989b ‘La Dynastie des Igihalkides: Nouvelles Interpretations’, in AIO, pgs. 223 – 235. 1993 ‘Kuk-Našur et Ammisaduqa’, NABU, 136. 1994 ‘Succession royale en Elam au IIème millénaire’, in CDR, pgs.1 – 14. 1995 ‘Susa and Susiana in second-millennium Iran’, CANE, 1023 – 1033. 1996a ‘Šu-ilišu, Iddin-Dagan et Imazu, roi d’Anšan’, NABU, 77 – 78. 1996b ‘L’Élam à l’époque paléo-babylonienne et ses relations avec la Mésopotamie’, in J. – M. Durand (ed.), Mari, Ébla et les Hourrites, dix ans de travaux, première partie, Paris: Amurru 1, pgs. 297 – 319. 2000 ‘L’Elam du IIe millénaire et la chronologie courte’, Akkadica 119 – 120: 7 – 17.

VAN BUREN, E. D., 1954 ‘The Esoteric Significance of Kassite Glyptic Art’, Orientalia 23: 1-39. 1959 Catalogue of the Uga Sissa collection of stamp and cylinder seals of Mesopotamia.

VANDEN BERGHE, L., 1968 ‘La Nécropole de Bani Surmah. Aurore d’une civilisation du Bronze’, Archéolgia 24: 52 – 68. 1973 ‘Excavations in Luristan, Kalleh Nisar’, Bulletin of the Asia Institute of Pahlavi University 3: 25 – 56.

VANDEN BERGHE, L. and TOUROVETS, A., 1994 ‘La Glyptique de Kalleh Nisar, Pusht-i Kuh-Luristan’, IA 29: 9 – 45.

ECS Corpus, Volume I, Part III 811 VOIGT, M. M. and DYSON, R. H., Jr., 1992 ‘The Chronology of Iran, ca.8000 – 2000 BC’, in R. W. Ehrich (ed.), Chronologies in Old World Archaeology, Chicago: University of Chicago Press, pgs. 122 – 178.

VOLK, J. G., 1979 Habib Anavian Collection: Ancient Near Eastern Cylinder and Stamp Seals from the Early 6th Millennium B.C. to 651 A.D, New York: Habib Anavian Galleries.

VOLLLNWEIDER, M. –L., 1967 Musée d’Art et d’Histoire – Genève – Catalogue raisonné des sceaux cylindres et inailles, Vol. I, Geneva. 1983 Musée d’Art et d’Histoire – Genève – Catalogue raisonné des sceaux cylindres et inailles, Vol. III, Geneva.

WILCKE, C., 1987 ‘Inschriften 1983 – 1984 (7.-8 Kampagne)’, in B. Hrouda, Isin-Išan Bahriyat III, Munich: Bayr. Akademie der Wissenschaften, philosopohische-historische Klasse, Abh. N. F. 94, pgs. 83 – 120.

WARD, W. H., 1920 The Seal Cylinders of Western Asia, Washington, D.C.

WEISS, H. and YOUNG, T. C., Jr., 1975 ‘The Merchants of Susa – Godin V and Plateau-Lowland Relations in the Late Fourth Millennium BC’, Iran 13: 1 – 17.

WILLIAMS, M. F., 1928 ‘The Collection of Western Asiatic Seals in the Haskel Oriental Museum’, The American Journal of and Literatures 44: 232 – 253.

ECS Corpus, Volume I, Part III 812 WILLIAMS FORTE, E., 1976 Ancient Near Eastern Seals – a Selection of Stamp and Cylinder Seals from the Collection of Mrs. William H. Moore, New York: Metropolitan Museum of Art. 1981 ‘Cylinder Seals; in Surkh Dum at the Metropolitan Museum of Art: a Mini- Report’, JFA 8: 351 – 359.

WINTER, I. J., 1986 ‘The King and the Cup: Iconography of the Royal Presentation Scene on Ur III Seals’, in M. Kelly-Buccellati (ed.), Insight through Images: Studies in Honor of Edith Porada, Malibu: Undena Publications, pgs. 253 – 268.

WISEMAN, D. J., 1962 Catalogue of Western Asiatic Seals in the British Museum – Cylinder Seals I, Uruk – Early Dynastic Periods, London: British Museum.

WISEMAN, D. J. and FORMAN, B., 1959 Cylinder Seals of Western Asia, London.

WOOLLEY, C. L., 1934 The Royal Cemetery, Ur Excavations 2, New York; Philadelphia: the Joint Expedition of the British Museum and of the University Museum, University of Pennsylvania.

WRIGHT, H. T., 2007 ‘Ancient Agency: Using Models of Intentionality to Understand the Dawn of Despotism’, in E. C. Stone (ed.), Settlement and Society: Essays Dedicated to Robert McCormick Adams, Los Angeles: Costen Institute of Archaeology, University of California, the Oriental Institute of the University of Chicago, pgs. 173 – 184.

WRIGHT, H. T. and JOHNSON, G. A., 1985 ‘Regional perspectives on southwest Iranian state development’, Paléorient 11: 25 – 30.

ECS Corpus, Volume I, Part III 813 WRIGHT, H. T., MILLER, N. and REDDING, R., 1980 ‘Time and Process in an Uruk Rural Center’, in M. T. Barrelet (ed.), L’Archéologie de l’Iraq du début de l’Époque Néolithique à 333 avant notre ère – perspectives et limites de l’interprétation anthropologique des documents, Paris: Éditions du Centre National de la Recherche Scietifique, pgs. 265 – 284.

WRIGHT, H. T., REDDING, R. W. and POLLOCK, S. M., 1989 ‘Monitoring interannual variability: an example from the period of early state development in southwestern Iran’, in P. Halstead and J. Shea (eds.), Bad Year Economics: Cultural Responses to Risk and Uncertainty, Cambridge: Cambridge University Press, pgs. 106 – 113.

YOUNG, T. C., Jr., 1969a Excavations at Godin Tepe: First Progress Report, Toronto: Royal Ontario Museum. 1969b ‘The Chronology of the Late Third and Second Millennia in Central Western Iran as seen from Godin Tepe’, AJA 73: 287 – 291. 1978 ‘The Comparative Stratigraphy of Second Millennium Khuzistan’, Paléorient 4: 237-244. 1986 ‘Godin Tepe Period VI/V and Central Western Iran at the End of the Fourth Millennium’, in U. Finkbeiner & W. Röllig (eds.), Ğamdat Nasr: Period or Regional Style?, vol. 62, Wiesbaden, TAVO Beiheft B, pgs. 212 – 228.

YOUNG, T. C., Jr. and LEVINE, L. D., 1974 Excavations of the Godin Project: Second Progress Report, Toronto: Royal Ontario Museum.

YUSIFOV, Y. B., 1974 ‘The Problem of the order of succession in Elam again’, AAASH 22: 321 – 331.

ECS Corpus, Volume I, Part III 814 ZADOKS-JOSEPHUS JITTA, A. N., 1952 Catalogue sommaire des cylindres orientaux au Cabinet Royal des Médailles à la Haye, The Hague.

ZETTLER, R. L., 1987 ‘Sealings as Artifacts of Institutional Administration in Ancient Mesopotamia’, JCS 39: 197 – 240.

ECS Corpus, Volume I, Part III 815 Appendix A

Current Inscription ‘Elamite’ Image Number Relation Sargon M154 En-hedu-anna, Sanam- daughter of Sargon: shimut ….-kikudu, Hishep-ratep the scribe, Luh-ishan (is your servant)

M155 En-hedu-ana, Sanam- daughter of Sargon: shimut Ilum-pal[il] Hishep-ratep (is) her coiffeur Luh-ishan

Hammurapi M156 Ikun-pi-…., Siwe-palar- sanga priest of the huppak god Zaba[ba], son of Sin-magir, servant of Hammu- [rapi] M157 -mušallim, Siwe-palar- scribe, son of huppak Siatum, servant of Hammu-rapi

M158 …., [….] …., Siwe-palar- servant of huppak H[a]mm[a]-rapi

M159 Hammurabi ili Siwe-palar- no published image son of Iribaman …. huppak servant of Hammurabi M160 Sin-iqiša[m], Siwe-palar- no published image archivist (and) huppak gudapsum priest of the god [Nanna], son of Ilšu-ibbi[šu], servant of Hammu- ra[pi]

ECS Corpus, Appendix A 816 Current Inscription ‘Elamite’ Image Number Relation Hammurapi M161 Anum-pi-[….], son Siwe-palar- no published image of Lala[….], huppak servant of Hammu- rap[i] M162 …., son of Siwe-palar- no published image Haliqum, servant of huppak Hammu-rap[i] M163 Nabium-[nasirum], Siwe-palar- no published image son of Sin-[….], huppak servant of Hammu- rapi M164 Sin-ma[gir], son of Siwe-palar- no published image Silli-za[….], huppak servant of Hammu- ra[bi] M165 Warad-[….], [s]on Siwe-palar- no published image of Erra-nad[a], huppak servant of Hammu- r[api] M166 [….], [son of Siwe-palar- no published image …]rat, servant of huppak Hammu-rapi M167 [Ilum]-pi-[Šamaš], Siwe-palar- no published image [son] of Sin-[….], huppak [servant] of Hammu-[rapi] M168 Šep-Sin, son of Siwe-palar- no published image Šamaš-muballi[t], huppak servant of Hammu- ra[pi] M169 Šamaš-Iamassašu, Siwe-palar- no published image son of huppak Muhad[dum], servant of Hamm[u-rapi] M170 ….BI-[….], son of Siwe-palar- no published image Šamaš-na[….], huppak servant of Ham[mu-rapi] M171 [A]qba-hamm[u], Siwe-palar- no published image diviner, [s]on of huppak Himdi-Samaš, [ser]vant of Hammu-rapi M172 Šamaš-ha[zir], son Siwe-palar- no published image of I-[….], servant huppak of Ha[mmu-rapi]

ECS Corpus, Appendix A 817 Current Inscription ‘Elamite’ Image Number Relation Hammurapi M173 Šamaš-liwwir, son Siwe-palar- no published image of Šamaš-isqam- huppak kin, overseer of the kulmaš[itum] priestesses, servant of Hammu-ra[pi] Burnaburiaš M174 Iškur, exalted Lord, II – Untash- provider of rain and Napirisha abundance, whose fame is far- (reaching) womb fame … is the godly grain the pleasant gift! Uri- Šudu, son of Kaššû, servant of Burnaburiaš king of the world M175 Adad-ushabashi, II – Untash- son of Kidin- Napirisha Marduk, the ša reši official of Burnaburiash

M176 …. II – Untash- Nin-Eanna Napirisha Burnaburiaš

M177 Kidin-Marduk, II – Untash- son of Ša-ilimma- Napirisha damqa, the ša reši official of Burnaburiaš, king of the world

M178 Kidin-Marduk, II – Untash- son of Ša-ilimma- Napirisha damqa, the ša reši official of Burnaburiah, king of the world, as long as he lives may he be noble

ECS Corpus, Appendix A 818 Current Inscription ‘Elamite’ Image Number Relation Kurigalzu M179 [To …(some deity) I – Pahir- Kurigalzu ishshan the son of Humban- Burnaburi]ash numena for his life II – Hurbatila and his soul presented (this)

M180 To Adad, the lord I – Pahir- of (the temple) ishshan Enamhe, his lord, Humban- Kurigalzu numena the son of II – Hurbatila Burnaburiash offered (?) his prayers and for his life presented (this) M181 Son of Kuri-[galzu] I – Pahir- king of the world ishshan Humban- numena II – Hurbatila

M182 Shamash(?)-…., I – Pahir- guda-priest of ishshan, Nuska, Humban- (priest/official) of numena Sharrat-Nippuri, II – Hurbatila servant of Kurigalzu, king of the world

M183 Uballissu-Marduk, I – Pahir- son of Warad-Ea, ishshan ‘shatammu’ of Humban- Kurigalzu numena …. II – Hurbatila …. Viceroy of Dilmun

M184 Duri-ulmash, I – Pahir- son of -shunu, ishshan servant of Humban- Kurigalzu, numena king of the II – Hurbatila universe, governor of the city of Dur-Kurigalzu

ECS Corpus, Appendix A 819 Current Inscription ‘Elamite’ Image Number Relation Kurigalzu M185 Kurigalzu I – Pahir- ishshan Humban- numena II – Hurbatila

M186 Oh goddess I – Pahir- Ninsun, mighty ishshan lady, eldest Humban- daughter …., great numena inspector of the god II – Hurbatila Enlil’s land survey, whose wisdom makes everything perfect: may joy be on him who seeks you, may his way be happy, behind him …., Uballissu- Marduk, son of Warad-Ea, treasurer, servant of Kurigalzu, king of the universe M187 Kurigalzu I – Pahir- ishshan Humban- numena II – Hurbatila

M188 Nuremlil, I – Pahir- high priest of Enlil, ishshan son of Kurigalzu Humban- the king of the numena world II – Hurbatila

M189 Duriulmas, I – Pahir- son of Belsunu, ishshan servant of Humban- Kurigalzu, numena king of Assar, II – Hurbatila sakkanaku of the city Dur-Kurigalzu

ECS Corpus, Appendix A 820 Current Inscription ‘Elamite’ Image Number Relation Kurigalzu M190 Kurigalzu I – Pahir- ishshan Humban- numena II – Hurbatila

M191 Terimangar, I – Pahir- son of ishshan Gishkuranshidada. Humban- chief official of the numena shrine of Ishtar, II – Hurbatila servant of Kurigalzu

M192 May the name I – Pahir- revealed progress ishshan on high, so that the Humban- net of his royalty numena may reach farther; II – Hurbatila his lot is abundance, riches …. his life far renowned for its fullness; a plenitude of days heavenly bright, for the great leader of men, the chief …. Kurigalzu, king of the whole world

ECS Corpus, Appendix A 821 Concordance M45: RIME 3/2: 1.2.78 M87: Buchanan 1981: M1. Speleers 1917: 594 M46: IRSA: IIIA2x 648x M2. Boehmer 1965: 763 M47: IRSA: IIIA2y M88: RIME 3/2: M3. Ball 1989: p.153 M48: RIME 3/2: 1.4.2005 M4: CCO I: T.104 1.2.2008 M89: RIME 3/2: M5: Boehmer 1965: 603 M49: RIME 3/2: 1.2.22 1.4.2006 M6: CCO I: T.36 M50: RIME 3/2: 1.2.67 M90: RIME 3/2: M7: CCO I: T.44 M51: RIME 3/2: 1.2.70 1.4.2007 M8: CCO I: T.101 M52: RIME 3/2: 1.2.69 M91: RIME 3/2: M9: CCO I: T.103 M53: RIME 3/2: 1.2.71 1.4.2010 M10: CCO I: T.105 M54: RIME 3/2: 1.2.68 M92: Buchanan 1981: M11: Ménant 1983 – 86: M55: RIME 3/2: 1.2.74 636 S.76 M56: RIME 3/2: 1.2.75 M93: Buchanan 1981: M12: CCO I: T.35 M57: RIME 3/2: 1.2.76 649 M13: Banks 1912: p.301 M58: Buchanan 1966: M94: Buchanan 1981: M14: CCO I: T.64 627 653 M15: Porada 1957: pl.29 M59: Buchanan 1966: M95: Legrain 1925: 284 M16: Collon 1982a: 469 681 M96: Porada 1948: 292 M17: IRSA: IIIA1n M60: Collon 1982a: 87 M97: Porada 1948: 447 M18: Legrain 1925: 277 M61: Legrain 1951: 428 M98: CCO I: T.221 M19: Legrain 1925: 286 – 429 M99: Collon 2005: 533 M20: Buchanan 1966: M62: RIME 3/2: M100: Collon 1982a: 417 1.2.2003 446 M21: Porada 1948: 295 M63: RIME 3/2: 1.3.20 M101: Legrain 1951: 416 M22: Collon 1982a: 470 M64: RIME 3/2: 1.3.22 M102: Legrain 1951: 418 M23: Collon 1982a: 471 M65: RIME 3/2: M103: Legrain 1951: 421 M24: Lambert 1979: 90 1.3.2003 M104: Legrain 1951: 435 M25: Buchanan 1981: M66: Buchanan 1981: M105: Legrain 1951: 436 606 637 M106: Legrain 1951: 438 M26: Buchanan 1981: M67: Buchanan 1981: M107: IRSA: IIIA5h 639 646 M108: IRSA: IIIA5i M27: Buchanan 1981: M68: Buchanan 1981: M109: IRSA: IIIA5j 642 647 M110: IRSA: IIIB5c M28: Buchanan 1981: M69: Buchanan 1981: M111: IRSA: IIID1b 643 650 M112: NATN: 631 M29: Buchanan 1981: M70: Buchanan 1981: M113: RIME 3/2: 1.5.8 648 651 – 652 M114: RIME 3/2: M30: Buchanan 1981: M71: Buchanan 1981: 1.5.2006 660 654 M115: RIME 4: 1.1.2002 M31: Munn-Rankin M72: CCO I: T.217 M116: RIME 4: 1.1.2006 1959: 13 M73: CCO I: T.219 M117: RIME 3/2: M32: Collon 2005: 458 M74: Ward 1920: 34 1.5.2012 M33: Collon 2005: 797 M75: Ward 1920: 35 M118: IRSA: IVA1a M34: CCO I: T.186 M76: Ward 1920: 52b M119: IRSA: IVA1b M35: CCO I: T.215 M77: Legrain 1951: 432 M120: RIME 4: 1.1.2005 M36: CCO I: T.218 M78: Bergamini 1998 M121: RIME 4: 1.1.2001 M37: CCO I: T.224 M79: Buchanan 1966: M122: RIME 4: 1.1.2003 M38: de Clercq & 439 M123: RIME 4: 1.2.2004 Ménant 1888/1903: 86 M80: Collon 1982a: 445 M124: RIME 4: 1.2.2008 M39: Ward 1920: 51 M81: RIME 3/2: 1.4.31 M125: RIME 4: 1.1.2010 M40: Ward 1920: 51a M82: RIME 3/2: 1.4.31 M126: RIME 4: 1.1.2011 M41: Legrain 1951: 407 M83: CCO I: T.192 M127: RIME 4: 1.1.2012 M42: Legrain 1951: 408 M84: IRSA: IIIA4k M128: Collon 1982a: 77 M43: Legrain 1951: 422 M85: IRSA: IIIA4l M129: RIME 4: 1.2.4 M44: RIME 3/2: 1.2.77 M86: IRSA: IIIA4l M130: RIME 4: 1.2.2001

ECS Corpus, Appendix A 822 M131: RIME 4: 1.2.2002 M167: RIME 4: M132: RIME 4: 1.2.2003 3.6.2016 M133: RIME 4: 1.3.2001 M168: RIME 4: M134: RIME 4: 1.5.2001 3.6.2017 M135: RIME 4: 1.5.2002 M169: RIME 4: M136: Buchanan 1981: 3.6.2018 753 M137: Buchanan 1981: M170: RIME 4: 760a-c 3.6.2019 M138: RIME 4: 2.6.2007 M171: RIME 4: M139: RIME 4: 2.7.2004 3.6.2020 M140: RIME 4: 2.7.2003 M172: RIME 4: M141: RIME 4: 2.7.2005 3.6.2014 M142: RIME 4: M173: RIME 4: 1.5.2003 3.6.2009 M143: RIME 4: 2.7.2007 M174: Porada 1948: 577 M144: RIME 4: 2.8.2001 M175: Matthews 1990: M145: RIME 4: 2.8.2002 164 M146: RIME 4: 2.8.2003 M176: Matthews 1990: M147: RIME 4: 2.8.2004 36 M148: RIME 4: 2.8.2006 M177: Porada 1981: 26 M149: RIME 4: 2.8.2007 M178: Porada 1981: n M150: RIME 4: 2.8.2008 M179: Osten 1934: 660 M151: RIME 4: 2.8.2009 M180: Osten 1934: 665 M152: Matthews 1992: M181: Matthews 1992: 61 48 M153: Matthews 1992: M182: Matthews 1992: 18 188 M154: Boehmer 1965: M183: Matthews 1990: 9 458 M184: Matthews 1990: M155: Collon 1982a: 54 38 M156: Buchanan 1981: M185: Matthews 1990: 791 45 M157: Parrot 1959: M186: Matthews 1990: pl.49ME193 68 M158: Legrain 1925: 328 M187: Matthews 1990: M159: RIME 4: 69 3.6.2005 M188: Osten 1934: 276 M160: RIME 4: M189: Ward 1920: 40a 3.6.2006 M190: Ward 1920: 41 M161: RIME 4: M191: Ward 1920: 41a M192: Legrain 1925: 531 3.6.2007

M162: RIME 4: 3.6.2010 M163: RIME 4: 3.6.2011 M164: RIME 4: 3.6.2012 M165: RIME 4: 3.6.2013 M166: RIME 4: 3.6.2015

ECS Corpus, Appendix A 823