Acknowledgements

We wish to thank our colleagues for sharing unpublished information, off-prints of forthcoming publications, and general reflections on the Late Chalcolithic period. We especially thank Morteza Zamani Dadane, director of the Marivan Plain Survey project from the Kurdish Studies Institute at the University of Kurdistan. We also thank Khaled Abu Jayyab and Michael Lewis for their useful comments on previous iterations of this article. Productive feedback from three anonymous reviewers contributed greatly to the final version of this article.

Work on the unpublished Mahidasht survey dataset was conducted at the Royal Ontario Museum by S. Renette with the permission and support of L. Levine, director of the Mahidasht Survey Project 1975-78, and C. Reichel, Associate Curator at the ROM. New drawings of selected sherds from Pisdeli were made by S. Renette at the Penn Museum and printed here with permission of R. Zettler, Associate Curator-in-Charge of the Near Eastern Section.

1The Chalcolithic period, broadly understood to span the 5^th-4^th millennia BC in Southwest Asia, is defined by a long process that transformed dispersed, small village communities into a network of urban societies. At the beginning of this period, ca. 5000 BC, agriculture had become the dominant subsistence strategy together with village-based animal husbandry across Southwest Asia. Daily life would not reach beyond the village’s territory and its immediate neighbours. By the end, ca. 3200 BC, specialised centres, some of which displayed urban characteristics, dominated landscapes with hierarchically structured settlements. Goods, peoples, and ideas spread through a vast network that stretched from the Eastern Mediterranean to Central Asia. Yet, while this process can be generally identified in every part of Southwest Asia, trajectories along this path were highly regionalised, reflecting not only geographical preferences for communication routes and the presence of desirable resources, but also local communities’ choices to engage with or reject external cultural influences.

4In these recent studies of the Late Chalcolithic, the northern and central Zagros Mountains that form the interface between Mesopotamia and the Iranian highlands are generally ignored, with the notable exception of the site of Godin Tepe. Our knowledge of the Chalcolithic developments in this region is still largely based on a series of small soundings and surveys conducted from the 1950s to 1970s. In the 1980s, E. Henrickson, L. Levine, and T.C. Young Jr. made great efforts to synchronise these fieldwork results with the Mesopotamian and Iranian sequences through stylistic parallels and rudimentary radiocarbon dates. But while our knowledge of northern Mesopotamia and the central Iranian Plateau has changed drastically since then, the chronological framework for the Zagros region has not been updated (Voigt and Dyson 1992; Moghaddam and Javanmardzadeh 2013).

Previous studies of the Chalcolithic Zagros

9An alternative explanation for Godin Tepe emphasises the role of Susa and the emerging Proto-Elamite network for the distribution of Uruk-related material culture onto the Iranian Plateau. This approach, first formulated by H. Weiss and T.C. Young Jr., and more recently by C. Petrie, accepts the absence of similar outposts between the Tigris and the Zagros (Weiss and Young 1975; Petrie 2014). Petrie especially considers the relative and absolute chronological aspect of the formation of the Proto-Elamite network that fits better with the evidence from Godin Tepe, rather than the spread of the Uruk network somewhat earlier.

10A third approach prioritises indigenous developments in the Kangavar Plain. Both M.S. Rothman and R. Matthews preferred to take a bottom-up approach and considered local elites at Godin Tepe as a driving factor in the adoption of southern Mesopotamian material culture within part of the settlement (Matthews 2013; Rothman 2013). A recent study of local versus non-local pottery based on the chemical composition of clays at Godin Tepe provides additional support for this emphasis on the local community since most vessels were made locally, while the few imported vessels came mainly from neighbouring areas in the Iranian highlands (Gopnik et al. 2016).

11Since the height of fieldwork in the Zagros in the 1970s, there have been no new extensive excavations of Chalcolithic sites. Nevertheless, soundings and surveys conducted by Iranian teams in the past two decades have begun to fill in the gaps on the map so that we no longer need to rely exclusively on the key sequences of Godin Tepe/Seh Gabi, Tepe Siahbid/Chogha Maran, and Dalma/Pisdeli/Hasanlu. This article aims to update the chronological framework for the Zagros Chalcolithic, bringing it in line with the current chronologies of Mesopotamia and the Iranian highland. This remains largely a relative, ceramic chronology, but new ¹⁴C dates from soundings are beginning to provide much-needed anchor points.

Late Chalcolithic type sites in the Zagros

The northern Zagros: Pisdeli

Drawings by S. Renette

15A single exception is a sherd of an open jar with a triangular rim (Dyson and Young 1960: pl. 3). This vessel has a dark grey, chaff-tempered fabric with a white slip. The rim is decorated with painted lines. Just below the low neck is a metope frieze with alternating hatched panels and stylised horned animals, beneath which there are a series of horizontal lines. This sherd fits better with the closely related Seh Gabi tradition of the eastern central Zagros.

The eastern central Zagros: Godin Tepe and Seh Gabi

18The Chalcolithic sequence of Godin Tepe and Seh Gabi has been discussed extensively elsewhere (Henrickson 1985; Levine and Young 1987; Rothman and Badler 2011). This sequence (Godin IX-VI) defines our understanding of the Late Chalcolithic east of the Kuh-i Sefid (eastern Luristan and western Hamadan provinces). Due to Godin Tepe’s fame in Near Eastern archaeology, this sequence tends to be more familiar to those working outside of the Zagros region. However, due to its geographical location, the site is not representative of the entire central Zagros. Instead, this region’s cultural alignment fluctuates between northwestern Iran via the eastern flanks of the northern Zagros and the central Iranian Plateau via the Hamadan route during the Chalcolithic and Bronze Age.

20Period Godin VII forms a break with the previous periods based on the disappearance of painted wares and an orientation toward the central Iranian Plateau. The distinct plain, chaff-tempered wares consist mainly of coarse ware (smoothed and unsmoothed) basins with finger-impressed lug handles; red-slipped ware goblets, trays, and open jars with moulded zigzag decoration on the outside of the rim; and buff fine wares (sometimes white-slipped) for hemispherical bowls, shallow bowls, and sinuous-walled goblets (Levine and Young 1987: fig. 16). The best parallels for this material, as well as associated stamp seals, are found to the east, especially at Tepe Ghabristan I (Voigt and Dyson 1992: 161; Madjidzadeh 2008: fig. 8-9; see also Moghimi and Davoudi 2020 for ceramic and glyptic evidence from the nearby site at Ālou).

21Period Godin VI developed from the period VII tradition, but with a strong increase in buff (medium-) fine wares and the appearance of new vessel shapes, such as inverted rim bowls and pedestaled bowls. In addition, a newly painted ware is introduced, although uncommon, that has direct parallels within the Sialk III tradition, reflecting the intensified integration of this eastern zone of the Zagros with the central Iranian Plateau. By Godin VI:2, the first influences from the Mesopotamian world become visible in the ceramic record with the gradual introduction of BRB, as well as other selected vessel types that were widespread in Late Chalcolithic Mesopotamia (e.g., bowls with an interior bevel; jars with a straight or cannon spout; see Badler 2002: 81; Young 2004; Rothman and Badler 2011: 90). While these vessel types do have parallels within southern Mesopotamia, they are widely shared across Greater Mesopotamia by the second half of the 4^th millennium BC and do not represent the typical Uruk assemblage. Furthermore, a growing influence of southern Mesopotamian ceramic traditions has also been observed in the central Iranian Plateau where, by the end of the Sialk III tradition the local communities integrated foreign vessel types such as BRB, spouted vessels, and nose-lugged jars within their local assemblages (Fazeli et al. 2013; Helwing 2013a: 95; Helwing 2013b). In other words, the initial appearance of BRB and a few other ceramic types of Mesopotamian origins at Godin Tepe continues the site’s close connection with the central Iranian Plateau and widely shared increased communication with the Mesopotamian world.

23The absolute chronology for Godin VII-VI remains contested due to inconsistencies and a clear late bias in many of the radiocarbon dates from Godin Tepe (Voigt and Dyson 1992: 164). Those dates suggest a range of ca. 3700-3300 cal. BC for phase VII, which is almost certainly too low since it does not match the secure dates for the contemporary Sialk III:1-5 period and it leaves very little time for the long Godin VI phase (Fazeli et al. 2005, 2009, 2013; Pollard et al. 2013). A reassessment of radiocarbon dates from VI:1 produced a broad range of ca. 3500-3050 cal. BC, but no dates for the earlier phases VI:3 and VI:2 are available (Rothman and Badler 2011: 84-85). New dates from recent fieldwork at other sites with comparable material now allow us to refine this chronology.

The western central Zagros: Chogha Maran and Tepe Siahbid

24The western central Zagros, best documented in Kermanshah province, underwent a separate development that reflects contact with the Zagros Piedmont and the central Tigris region. This is not surprising considering its geographical location and natural communication routes via mountain passes that open to the Diyala/Sirwan region. Perhaps for this reason, the western central Zagros did not adopt the Dalma material culture, but instead developed its own black-on-buff and red-on-buff ware (Siahbid phase) that is closer related to the late Ubaid tradition of central Mesopotamia (Henrickson 1986: 122). For example, typical vessel shapes of the Siahbid tradition include conical bowls with simple or pinched rims and flaring-neck jars that recall typical Ubaid vessel shapes, while the painted motifs often have good parallels in central Mesopotamia (Henrickson 1985: 73, fig. 14; Levine and Young 1987: 33, fig. 12; Voigt and Dyson 1992: 158). But unlike in the lowlands, the Siahbid painted ware tradition seems to have lasted into the second half of the 5^th millennium BC. However, the absolute chronology of the Chalcolithic in the western central Zagros is poorly documented with only two radiocarbon dates for the early Siahbid phase that provide a range of ca. 4880-4490 cal. BC (Henrickson 1985: 73; Voigt and Dyson 1992: table 2). This suggests that the late Siahbid phase might have lasted into LC 1, but this is not presently supported by radiocarbon dates or parallels in material culture.

Drawings by S. Renette based on original drawings in the Mahidasht Survey Project archive at the Royal Ontario Museum

26The 4^th millennium occupation of the western central Zagros remains elusive in the present available data, since both Chogha Maran and Tepe Siahbid were not occupied during this time. However, survey data has shown that during the 4^th millennium BC, at least some of the Godin VII-VI plain wares spread westward and even a few painted Godin VI sherds have been documented from Deh Savar (Levine and Young 1987: 39). In addition, BRB are well-documented at many sites in the western central Zagros, sometimes at high quantities and with possible additional Mesopotamian-style ceramics (e.g., pinched rim cups with a string-cut base) that might reflect the spread of the Uruk network into this region (Henrickson 1994; Renette 2018). Still, at present, there is insufficient data to assess the material culture of the western central Zagros during the 4^th millennium BC.

Recent fieldwork results from the Zagros region

Upper Little Zab basin

29This basin, which stretches from the Ushnu-Solduz plains to Marivan along the eastern side of the chaîne magistrale that encompasses the modern-day border between Iraq and Iran, is connected to the Zagros Piedmont and the Tigris region via only a few passes (Hajji Omran, Keleshin, Peshdar Plain). Until fairly recently, this region was terra incognita on archaeological maps, but now Iranian survey and excavation projects have begun to fill in this gap. Two sites, in particular, provide evidence for the Chalcolithic period: Lavin Tepe near Piranshahr in the north and Tepe Namashir near Baneh in the south. These sites show that the Dalma cultural tradition, including both painted and impressed wares, spread into this region, but not further south or west. The Pisdeli period is present at both sites, but rather poorly represented. Only a few black-on-buff sherds from these sites can be attributed to the Pisdeli tradition, while no black-on-red pottery has been reported (Nobari et al. 2012: fig. 7.2; Saed Mucheshi et al. 2017: fig. 12.16). These sherds probably date to the early part of the Pisdeli period, which is supported by the presence of late Ubaid painted conical bowls (Saed Mucheshi et al. 2017: fig. 11.2, 12.11). At Tepe Namashir, there was continued occupation at least into LC 1, although it is not clear whether there was a stratigraphical break from the previous period. The ceramic assemblage of level III at the site shows both burnished, red-slipped ware on typical Pisdeli shapes (Saed Mucheshi et al. 2017: fig. 8) and buff wares on typical LC 1 shapes familiar from the Zagros Piedmont and Tigris region (Saed Mucheshi et al. 2017: fig. 9-10).

30For the later part of the Late Chalcolithic, the Lavin Tepe project reported BRB at four sites in the area (Nobari et al. 2012: 103). A recently published report of a 2007 survey in the Sardasht district of the Little Zab basin documented an additional six sites with Late Chalcolithic sherds. Material collected on the surface at the site of Tepe Badamyar Rabat is almost identical to that from LC 3-4 sites in the Trans-Tigridian region, demonstrating that, in contrast to the southern Lake Urmia area, during the 4^th millennium BC this area was culturally integrated with the Zagros Piedmont zone (Abedi et al. 2019).

Marivan – Sanandaj – Mahidasht

31Further south, a series of small valleys in an otherwise rugged, mountainous landscape connect the northern Zagros to Kermanshah to the south and the Shahrizor and Tanjaro plains of Sulaymaniyah in the west. On the Iraqi side of the border in Penjwen, excavations at Gird-i Rostam have recently produced good evidence for LC 1 occupation (Potts et al. 2019). The plain wares at this site were often burnished, which fits well with widespread Zagros Plain wares of this period (Potts et al. 2019: 114). The site also produced a few painted sherds, one of which fits well within the Pisdeli/Seh Gabi black-on-buff tradition with a band of geometric motifs above a painted band or horizontal lines on a small hemispherical bowl (Potts et al. 2019: 119). The motif of alternating upward- and downfacing triangles has a long history from at least the late Ubaid and Pisdeli periods (Dyson and Young 1960: fig. 2.3) to the Godin VI period (Levine and Young 1987: fig. 17.6), as do dots as filling motif (Dyson and Young 1960: fig. 2.2; Levine and Young 1987: fig. 10:50.1 and 11.8; Zamani Dadane et al. 2019a: fig. 5.5).

32A little further southeast, near Sanandaj, the material culture of Tepe Gryashan is much more representative of the Zagros region with many black-on-buff painted sherds of the Pisdeli/Seh Gabi tradition (Mohammadi Ghasrian et al. 2019: fig. 3) and red-slipped plain wares that frequently have finger-impressed surface treatment (Zarei et al. 2017: fig. 14; Mohammadi Ghasrian et al. 2019: fig. 11). Mixed in with this typical northern Zagros LC 1 tradition are several wide flower pots, which are chaff-tempered, straight-sided bowls with scraped bases, typical of the northern Mesopotamian LC 1-2 (Mohammadi Ghasrian et al. 2019: fig. 8). Gryashan seems to have remained occupied at least through the end of LC 2 as evidenced by the presence of red-slipped plain ware more similar to Godin VII and the presence of chaff-tempered vessel types with comparanda in the plain wares of Godin VII-VI as well as the LC 1-2 assemblage at Kani Shaie (Mohammadi Ghasrian et al. 2019: fig. 9 and 11; Rothman and Badler 2011: fig. 4.45; Renette et al. forthcoming: fig. 8-9 [see note 1]).

33At Qaleh Naneh, a few sherds suggest occupation during LC 1-2: a vat with a grooved rim is paralleled in LC 2 levels at Kani Shaie and Logardan in southern Iraqi Kurdistan (Vallet et al. 2017: fig. 18.7; Saed Mucheshi et al. 2018: fig. 9.4; Renette et al. forthcoming: fig. 9.12 [see note 1]), while two sherds of black-on-red painted ware have parallels in the Mahidasht and beyond (Henrickson 1985: 74, fig. 20; Saed Mucheshi et al. 2018: fig. 10). The site continued to be occupied at least until the middle of the 4^th millennium BC based on the presence of BRB, hammerhead bowls, and a distinct grey ware, all of which have parallels in LC 3-4 contexts in Iraqi Kurdistan (Saed Mucheshi et al. 2018: fig. 8-9). BRB were also found at the nearby sites Tepe Rasha and at Sarghal’eh together with other possible LC 3-4 ceramic types (Saed Mucheshi et al. 2012: fig. 6-8; Zamani Dadane et al. 2019b: fig. 8).

34Similar sherd material has been retrieved from several sites in the Mahidasht/Kermanshah region during the 1975-1978 Mahidasht Survey Project, led by L. Levine. In her summary of the Late Chalcolithic material from this survey, Henrickson identified Godin VI related ceramics (Henrickson 1994). However, sherd drawings from the original survey records suggest a stronger western orientation, similar to the material found around Marivan and the Trans-Tigridian region, with only broad connections to the Godin VII-VI assemblages (fig. 4). Almost none of the characteristic Godin VII shapes are present in the archive’s drawings, but final analysis of the complete survey collection remains to be done (Young and Levine 1974: fig. 13; Levine and Young 1987: fig. 16).

Drawings by S. Renette based on original drawings in the Mahidasht Survey Project archive at the Royal Ontario Museum

Eastern Zagros

36Further north, Tepe Kalanan (Saed Mucheshi et al. 2011) and Tepe Gheshlagh (Motarjem and Sharifi 2014; Sharifi and Motarjem 2018) illustrate the local variations on the wider Chalcolithic themes. Both sites were occupied during the Godin VII period based on the presence of typical red-slipped ware with moulded application under the rim (Sharifi and Motarjem 2018: fig. 15) and coarse ware basins with finger-impressed lugs (Saed Mucheshi et al. 2011: fig. 9; Motarjem and Sharifi 2014: fig. 17b) that have direct parallels in the Godin Tepe sequence (Levine and Young 1987: fig. 16). In addition, the local assemblages from these two sites include vessel decorations that are unparalleled at Godin Tepe: appliqué stylised goat horns (Saed Mucheshi et al. 2011: fig. 18; Sharifi and Motarjem 2018: fig. 11), red-slipped ware with a single M (perhaps a stylised mountain?) painted in white near the base of the vessel (Sharifi and Motarjem 2018: fig. 10), and greyish buff ware with incised zigzags (Saed Mucheshi et al. 2011: fig. 18-19; Sharifi and Motarjem 2018: fig. 14). A set of radiocarbon dates from Tepe Kalanan and thermoluminescence dates from Tepe Gheshlagh both provide a secure absolute date between 4000-3750 cal. BC, equivalent to late LC 2-3.

Toward a synthesis of the Chalcolithic of the northern and central Zagros region

37Given the lack of long-term, extensive excavations and detailed surveys in the Zagros region during the past four decades, writing a synthesis of the Chalcolithic period necessarily needs to focus on ceramic development and chronology. Such a ceramic chronology, anchored by absolute dates, allows the correlation of the Zagros region with Chalcolithic developments in neighbouring regions. Here, we outline an updated chronological framework, highlighting remaining gaps in our knowledge, and we reevaluate how the Zagros Chalcolithic traditions correlate with developments in neighbouring Mesopotamia and the Iranian highlands. Previous syntheses could only rely on three geographically remote and distinct stratigraphic sequences (Pisdeli, Seh Gabi/Godin Tepe, Tepe Siahbid/Chogha Maran), which amplified their differences. With new data from regions in-between, this synthesis considers the Chalcolithic of the northern and central Zagros to be largely a unified process, based on continuous interaction between dispersed communities, that is expressed in different ways at the local level.

38Since Henrickson first developed the three-fold division of the Zagros Chalcolithic (Early, Middle, Late; Henrickson 1985), the terminologies and absolute dates for Mesopotamia and the central Iranian Plateau have changed drastically resulting in problematic discrepancies. Adapting the Zagros chronological framework to these neighbouring regions is a difficult task, not only because of the ongoing lack of detailed datasets. While the Santa Fe scheme for Greater Mesopotamia divides a long Late Chalcolithic (ca. 4600-3100 BC) into five subperiods (LC 1-5) following the Ubaid period (Early Chalcolithic) of the early 5^th millennium BC, the now current chronological scheme for the central Iranian Plateau utilises a very long Transitional Chalcolithic (ca. 5200-4300 BC) followed by short Early, Middle, and Late Chalcolithic periods of 300 years each (table 1). Given such diverging chronological schemes, it is impossible to tie the Zagros sequence to both simultaneously in a single framework, while developing a third independent scheme is undesirable if the goal is to facilitate closer integration of archaeological datasets between these regions in recognition of the high degree of interaction during the Chalcolithic. Still, without the availability of a robust absolute chronology based on multiple stratigraphically anchored sequences, the Zagros sequence needs to be correlated using similarities in ceramic developments.

39The Zagros region underwent cultural influences from both directions, often fluctuating in the degree of influence from west or east and with differences from valley to valley, while largely maintaining its own distinct cultural traditions. For example, given Godin Tepe’s position at the eastern end of the Zagros on a major route leading to the eastern highlands, this site’s sequence relates very closely to the central Iranian Plateau in the 4^th millennium BC, while the Marivan and Mahidasht regions experienced more influences from the Mesopotamian world. Overall, we argue that the majority of the northern and central Zagros aligns more closely with the Mesopotamian sequence based on the stylistic similarities with the Ubaid and LC 1 painted traditions of northern Mesopotamia, the subsequent spread of CFW-related materials through much of the western Zagros, and the lack of distinct pottery types (e.g., monochrome painted ware) that characterise the Late Chalcolithic of the central Iranian Plateau (ca. 3700-3300 BC, equivalent to Godin VI) throughout the majority of the Zagros with only the exception of the eastern zone (east of the Kuh-i Sefid). We maintain the division of the Zagros Chalcolithic based on ceramic traditions as laid out originally by Henrickson, but using new evidence we update the absolute chronology (fig. 5), align the development with the Santa Fe LC scheme, and argue for a stronger cultural integration of Zagros communities than previously assumed.

Early Chalcolithic (ca. 5000-4600 BC)

41In Henrickson’s framework, next came the Middle Chalcolithic with the introduction of the Dalma ceramic tradition. In order to align the Zagros chronology with both Mesopotamia and the central Iranian Plateau, as well as to reflect similar cultural developments in the Zagros as its neighbours, we here rename this period the Early Chalcolithic (previously Middle Chalcolithic 1), roughly dated to 5000-4600 BC.

42During the Early Chalcolithic, the Dalma tradition, consisting of a relatively uniform assemblage of painted, impressed, and plain wares, spread across a region from Lake Urmia to the eastern central Zagros (Kangavar and Nehavand valleys) along the eastern flanks of the Zagros. Data from Lavin Tepe and Tepe Namashir demonstrate that this tradition also spread through the upper Little Zab basin (along the border with present-day Iraq) just north of the Marivan region.

43The western central Zagros (Kermanshah and Marivan regions) was oriented toward Mesopotamia during this period. Communities in this zone developed a distinct local variant of the black-on-buff painted ware closely related to the late Ubaid tradition, in addition to a red-on-buff version, forming the Siahbid tradition (Levine and Young 1987: fig. 12 and 15). At the same time, Dalma impressed ware was widely adopted in this western zone and the plain wares were probably very similar throughout the entire Zagros region.

Late Chalcolithic 1 (ca. 4600-4200 BC)

44In Henrickson’s framework, this was the Middle Chalcolithic 2 period, which encompasses the Pisdeli tradition in the north, the Seh Gabi tradition in the eastern central Zagros, and the late Siahbid and Maran traditions in the western central Zagros. The date of the beginning of this period is uncertain. There is good evidence for influence from the late Ubaid in northern Mesopotamia on the development of Pisdeli painted ware, but overall the Pisdeli material finds its best parallels in Gawra XII-XI that span LC 1 and perhaps into early LC 2 (Henrickson 1986: 121). As with the transition from late Ubaid to LC in northern Mesopotamia, the development from Dalma to Pisdeli happened gradually, while in the western Zagros CFW and mass-produced bowls also began to be produced.

46At Pisdeli and Seh Gabi, there is also evidence for social developments associated with the wider Chalcolithic world. Stamp seals and impressions were found that have good parallels at both Tepe Gawra in northern Mesopotamia and Tepe Sialk on the central Iranian Plateau (Voigt and Dyson 1992: 161). The Ushnu-Solduz region, where Pisdeli is located, was undoubtedly integrated in the exchange network that connected northern Mesopotamia with northern Iran and Central Asia at this time, as evidenced most obviously by the presence of lapis lazuli from northern Afghanistan at sites like Tepe Gawra. Further to the east on this route, a pot sherd with the impression of a stamp seal has been recovered during survey work in the Abharroud Basin in eastern Zanjan province (Alibaigi et al. 2011).

47Unlike most of the Zagros, the western zone around Marivan and the upper Little Zab basin increasingly began to adopt influences from the Trans-Tigridian region. Evidence from Tepe Namashir and possibly Gird-i Rostam shows that burnished, red-slipped Zagros plain wares were in use, while painted wares seem less common than in the rest of the Zagros. At the same time, vessel shapes identical to those from the Zagros Piedmont and Tigris region were adopted, aligning this region more with the northern Mesopotamian LC tradition where painted ware also phased out of use early in LC 1. Much of the LC 1-2 ceramic assemblages from Tepe Namashir, Tepe Gryashan, and Gird-i Rostam is virtually indistinguishable from those from Kani Shaie in Iraqi Kurdistan. This Mesopotamian tradition does otherwise not spread into the Zagros, except for occasional sherds at Pisdeli and in the western Kermanshah region where they stand out within the local assemblages.

Early Late Chalcolithic 2 (ca. 4200-4000 BC)

48Previously labeled as Middle Chalcolithic 3, this period is very likely the last phase when Zagros Chalcolithic painted traditions were dominant. Even though the absolute chronology is uncertain, there is good evidence to suggest that these painted traditions went out of use some time during the early part of LC 2, which can be considered a transitional phase. Therefore, we have opted to divide the LC 2 into an early and later part that reflects better the current Zagros phasing. However, as more radiocarbon dates in association with stratigraphically anchored assemblages become available, the transition from the painted traditions of LC 1 to the chaff-tempered Godin VII tradition might need to be pushed further back even more.

Late Chalcolithic 2-3 (ca. 4000-3700 BC)

49The period subdivisions for the 4^th millennium BC cannot be anchored with absolute chronology. The periodisation laid out here is therefore to a large extent speculative, but based on radiocarbon dates for the Godin VII period and the CFW tradition in northwestern Iran, in addition to perceived relative lengths of occupation at Godin Tepe VII-VI:3-1.

51The transition to Godin VI:3 is certainly gradual since the difference between the two periods is based mainly on the frequency of plain ware vessel shapes, in addition to the disappearance of a few distinct Godin VII shapes and the introduction of small amounts of painted pottery (see LC 4). Only at Tepe Gheshlagh is there one sherd that could fit with the Godin VI painted ware (Sharifi and Motarjem 2018: fig. 13), but thermoluminescence dates from the site give too wide a range to situate this within LC 3 and instead it could reflect continued occupation into LC4.

52In the western central Zagros, the Mahidasht survey project identified chaff-tempered plain wares of the Godin VII-VI period and possibly material related to Sargarab ware in southwestern Iran, which indicates that the region remained inhabited into the 4^th millennium BC despite the current lack of direct evidence for the LC 2-3 period from excavations. Small bits of information come from a rescue excavation near Taq-e Bustan in Kermanshah where a graveyard and some architectural remains were discovered associated with Late Chalcolithic pottery (Alibaigi et al. 2017; Mirghaderi and Hozhabri 2017). Assigning a precise date to this material is difficult, but the absence of BRB suggest a possible early date in late LC 2-3.

53Further to the north, the narrow zone encompassing the Little Zab basin and the Marivan area became fully integrated within the Zagros Piedmont tradition. Typical LC 2-3 pottery was identified at Tepe Gryashan and Qaleh Naneh near Marivan and at Tepe Badamyar Rabat in the Little Zab basin. CFW sherds have also been documented in the Marivan survey project at a few sites, which at least demonstrates that this typical LC chaff-tempered ware was spreading into the western part of the northern Zagros (Zamani Dadaneh et al. 2019a: fig. 6).

54In northwestern Iran, the chronology of the development of CFW is now better documented thanks to studies of old excavations at Geoy Tepe and new work at sites such as Dava Göz and Kul Tepe (Burton-Brown 1951; Marro 2010 [for the southern Caucasus]; Abedi 2017; Abedi et al. 2014, 2015). These have produced solid radiocarbon dates placing the beginning of this tradition already ca. 4200 BC and lasting until ca. 3700 BC when the Kura-Araxes culture spread southward from the Caucasus. Unfortunately, south of Lake Urmia, there is no data available for this period. Survey in the Ushnu-Solduz region was not able to detect any occupation after the Pisdeli period (perhaps mid-LC 2) until the end of the 4^th millennium BC (Voigt 1989: 287; Danti et al. 2004: 595).

Late Chalcolithic 4 (ca. 3700-3400 BC)

55The main reason at present to divide LC 4 from the preceding phase is the change at Godin Tepe with the introduction of painted wares that are closely related to the central Iranian Plateau. These painted wares generally did not spread further west, although occasional sherds do show up throughout the western central Zagros, especially in the Kermanshah region. However, when taking the Godin VI painted ware, which makes up only a small percentage of the Godin Tepe assemblage, out of consideration, the central Zagros region is remarkably homogeneous with a shared chaff-tempered tradition that is related, yet distinct, to the LC 3-4 ceramic tradition of the Trans-Tigridian region and Khuzistan, consisting of buff wares, coarse wares, and smoothed red-slipped wares that are technologically distinct from earlier red-slipped wares. The orientation of external influences in the eastern zone of the region toward the central Iranian Plateau, where Mesopotamian influence was also growing during this period, and the western zone toward Mesopotamia and southwestern Iran is a factor of geographical proximity and natural communication routes.

56The Marivan and Little Zab zone had already been largely integrated in the Mesopotamian Late Chalcolithic tradition, which is perhaps most clear from the early occurrence of BRB at sites such as Qaleh Naneh, with a sequence spanning LC 2-4, Tepe Rasha, Sarghal’eh, and Tepe Badamyar Rabat. However, the lack of distinct Uruk types at these sites, such as conical cups with a string-cut base, nose-lugs, impressed bands, and undercut band rim jars as well as “local” LC 4 ceramic types known from Kani Shaie, Logardan, Girdi Qala, Gurga Chiya, and Bab-w-Kur just across the border in Iraqi Kurdistan, could suggest that these sites were abandoned by LC 4. Alternatively, the spread of Uruk ceramic types might not have reached into the Zagros Mountains valleys at all, although limited unpublished data from the Mahidasht survey project suggests otherwise (Henrickson 1994). At the same time in northwestern Iran, around Lake Urmia (except for the Ushnu-Solduz plains on its southern shore) the Kura-Araxes culture spread and displaced the CFW tradition around the middle of the 4^th millennium BC (Abedi et al. 2015).

Late Chalcolithic 5 (ca. 3400-3200/3100 BC)

57By the final centuries of the Late Chalcolithic, the spread of the Uruk network reached the Iranian Plateau. The site of Godin Tepe is frequently used as evidence for southern Mesopotamian presence also in the Zagros Mountains, but again, this site, which is located at the eastern edge of the Zagros, is not necessarily representative. Rather, Godin Tepe and the Hamadan region had become culturally affiliated with the central Iranian Plateau where southern Mesopotamian pottery types were also introduced.

Discussion

59Summarising these developments, we can identify three major phases of the Chalcolithic period: Ubaid/Dalma (Early Chalcolithic); LC 1-2 black-on-buff and black-on-red with red-slipped, burnished plain ware; and LC 2-5 chaff-tempered plain wares. First, during the Early Chalcolithic, the Ubaid horizon spread throughout Mesopotamia, which also influenced the otherwise indigenous Siahbid painted ceramic tradition in the Mahidasht/Kermanshah region. At the same time, the northern Zagros and the eastern flanks of the central Zagros adopted the completely distinct Dalma tradition.

60By the mid-5^th millennium BC, both of these cultures were gradually replaced. However, while Mesopotamian communities largely abandoned painted ceramics during LC 1-2 in favour of chaff-tempered plain wares, Zagros communities shared a potting practice characterised by red-slipped, often burnished, plain wares and a new painted tradition consisting of simple black designs on a buff to white background. While similar in design, the technological execution differed strongly from region to region. In the eastern part of the central Zagros, pale buff vessels were decorated with black, shiny, almost vitrified paint (Seh Gabi). In the western Mahidasht/Kermanshah region, vessels with a reddish fabric were given a thin white wash before applying black painted designs (Maran red, white, and black ware). The northern end of the Zagros, around Lake Urmia, applied black painted designs either on buff vessels, often with a white slip, or on dark red-slipped vessels (Pisdeli). This black-on-red variant had a very different distribution, spanning from the eastern Khabur region to the western edge of the central Iranian Plateau, probably reflecting the emergence of an exchange network that connected northern Mesopotamia with Central Asia through northern Iran.

61By 4000 BC (LC 2), these painted traditions had gone out of use and were replaced by chaff-tempered plain wares, in parallel with the Mesopotamian world. The northwestern zone of the Zagros, encompassing the Little Zab basin and the Marivan region, were strongly influenced by the Trans-Tigridian region to the west. The central Zagros, on the other hand, adopted its own variant assemblage of chaff-tempered wares similar to Sargarab ware (Wright 2013: 52-57), while influence from the northern central Iranian Plateau, with chaff-tempered, red-slipped wares quickly reached the Hamadan region and Godin Tepe, and continued a gradual spread through the central Zagros. Throughout the 4^th millennium BC, Mesopotamian influence (mainly BRB) and Iranian Plateau influence (Godin VI painted ware) increased steadily, but the Zagros region appears to never have become fully integrated within the growing Uruk network.

Conclusion

63Lack of data on the Late Chalcolithic from the northern and central Zagros Mountains have resulted in an exclusion of this region from wider discussions of the major developments during the 5^th-4^th millennia BC. Only the important discovery of southern Mesopotamian-related ceramics at the site of Godin Tepe became a commonly referenced data point for the Zagros. The resulting assumptions were that either there are multiple outposts similar to Godin Tepe spread along the natural routes through the central Zagros or that Zagros communities were largely excluded from the Late Chalcolithic world.

64In recent years, Iranian archaeologists have collected important new data through surveys and small-scale excavations. However, the results from their fieldwork have not yet been included in wider Near Eastern scholarship. Synthesising these results reveals that the communities inhabiting the Zagros Mountains followed similar trends of the Chalcolithic as their Mesopotamian neighbours, while developing their own traditions in response to interregional developments. Strikingly, but perhaps not surprising considering the fragmented landscape, technological and stylistic ceramic practices were often highly localised in the Zagros while at the same time reflecting a continuous interaction sphere that resulted in shared decorative schemes, choices in potting practices (temper, surface treatment), and similar vessel shapes. With limited data available in the past, local variation in ceramic assemblages skewed archaeological interpretation toward a model of separate cultural traditions. Filling the map with new fieldwork results highlights commonalities between Zagros communities. This allows the emergence of a new understanding of the Chalcolithic period in the Zagros Mountains as an internally varied, yet unified cultural zone that existed in parallel to Mesopotamia and the central Iranian Plateau.

65The local variation of the Zagros Chalcolithic reflects the dramatic, mountainous landscape that severely restricts communication. Not surprisingly, the northwestern zone of the Little Zab basin and the Marivan region more rapidly integrated northern Mesopotamian influences facilitated by a few important passages through the chaîne magistrale. The western Mahidasht/Kermanshah zone frequently adopted influences from central Mesopotamia and southwestern Iran, while the eastern zone around Godin Tepe and Hamadan were naturally connected to the central Iranian Plateau and northwestern Iran. The fragmented landscape and the continuous influx of external influences were strong enough to prevent a real cultural unification of the mountain peoples of the Zagros during the Chalcolithic. Nevertheless, despite these challenges, these communities maintained a loose shared identity, distinct from their more unified neighbours.

66As work in the Zagros continues, new research will be able to move beyond issues of ceramic zones and chronology. It has become increasingly clear that interregional interaction was a major factor in the emergence of complex societies in the ancient Middle East. The Zagros Mountains served as an interface for the exchange of goods, ideas, and peoples between Mesopotamia and the Iranian Plateau. Mapping different influences in material culture will eventually allow the identification of the formation of trade routes, as well as the emergence of ethnic identities and local polities as communities developed a need to claim ownership over their land in response to increased contact with the outside world.