Introduction

Over the past few years, archaeologists in Iran have discovered an increasing number of Middle Palaeolithic open-air sites, many reported in Antiquity (e.g. Alborz, Berrilon et al. 2007; Kashafrud, Jamialahmadi et al. 2008; Bardia, Conard et al. 2009; Boeen Zahra and Delazian, Vahdahti Nasab et al. 2008, 2009 respectively). Another new site at high altitude is reported in this issue (Eskandari et al. 2010), while in the present paper we offer a brief report of the largest so far at Mirak in Semnan Province. This new work offers a contrast to the explorations of a previous generation, which were largely focused on caves and rockshelters, the majority of them located in the Zagros Mountains (Coon 1951; Hole & Flannery 1967; Speth 1971; Roustaei et al. 2002, 2004; Otte et al. 2007; Jaubert et al. 2009). There is a common misconception that hominins spent most of their times in caves and rockshelters during the Upper Pleistocene. This idea has arisen largely from the difficulty of defining other types of site and their vulnerability today. But the new discoveries suggest that open-air sites were by far the more abundant.

Location and archaeological background

Approximately 5km south of the modern city of Semnan (220km east of Tehran) at the northern edge of the Iranian Central Desert lie several small mounds, known to local residents as Mirak (Figure 1). Mirak originally consisted of seven mounds of between 4-11m in height, separated from each other by a few hundred metres and forming a line 2-3km in length. Archaeological survey has located a vast lithic scatter here, extending over some 4ha (Rezvani 1990). Our knowledge concerning the geological reasons behind the formation of the mounds is limited, but they seem to be naturally formed and not all of them have lithic materials on the surface. Systematic survey in the region has resulted in the discovery of further mounds at Delazian, located c. 4km north of Mirak (Vahdati Nasab et al. 2008).

Figure 1.

Location of Mirak in Iran.

Figure 2.

View of Mirak.

Surveying strategy and sampling method

During the summer of 2009, Mirak was the subject of intensive survey aimed at mapping and sampling the Palaeolithic occupation (Vahdati Nasab. 2009). Given the size and the quantity of the artefact scatter, its content was explored by random sampling. Close to 9000 lithic fragments were collected from eight different loci, each 4 x 10m. We may conclude that over tens of thousands lithic pieces might be present at this site.

Apart from its enormous size, Mirak is spectacular in the quantity and quality of the artefacts that are spread all over it. They include diagnostic Mousterian and Levalloisian lithic types, never before reported from Iran in such numbers (Figures 3-9).

Figure 3.

Convergent Scraper made on Levallois flake.

Figure 4.

Mousterian point.

Figure 5.

Round Scraper.

Figure 6.

Levallois point.

Figure 7.

Levallois Core, with centripetal flaking scars.

Figure 9.

Levallois point.

Figure 8.

Levallois point.

Conclusion

Although analysis of the artefacts collected from Mirak is far from complete, given the abundance of Mousterian points, convergent scrapers, single and double scrapers, discoid cores and Levalloisian technique (cores, points and flakes) we propose that Mirak has been in use since the Middle Palaeolithic. The presence of elements of later periods (blades and bladelets) may suggest a continuation of occupation in this site. The preliminary examination of the artefacts suggests a large quantity of finished tools, especially points (Mousterian and Levalloisian), in comparison to debitages and cores. This may indicate that Mirak was visited repeatedly by groups of hunters during the Upper Pleistocene. Through ongoing analysis we hope to shed more light on the Palaeolithic settlement patterns beyond the Zagros Mountains and in the Iranian Central Desert.