1
Trends and Patterns in the Archaeology and Pre-modern History of the Gulf Region
D.T. Potts
Recent decades have witnessed an explosion of archaeological and historical literature examining the past 8,000 years of human occupation in the Gulf region, but only rarely have scholars taken a step back from their particular fields of expertise to look more broadly for trends that might be relevant to both the earlier and the later periods, and to the Iranian and Arabian sides of the Gulf. The aim of this chapter is to examine a number of issues from an explicitly holistic perspective, before turning to some illustrations of interconnectedness or centripetal interaction in the Gulf from pre-modern times. To begin with, however, a few words about the physical configuration of the Gulf are in order.
The Gulf as we know it today is a shallow, epicontinental sea – a trough c.1,000 km long, 200–350 km wide – bordered by eight countries: Iraq, Kuwait, Saudi Arabia, Bahrain, Qatar, the United Arab Emirates (UAE), Oman (Raʾs Musandam) and Iran. The creation of this trough has been explained by the theory of plate tectonics and attributed to the movement of two of its plates. Sometime between c. twenty-five million and twelve million years ago, the upthrust of the Arabian plate in the west against the Asian plate, followed by the downward warp of the eastern side of the Arabian Peninsula, created the floor of the Gulf. The formation of the Zagros Mountains c. 5 million to 2 million years ago caused the marine transgression that inundated the area, creating a vast sea. During the Late Pliocene (3.6 million to 2.6 million years ago), when sea levels worldwide were up to 150 m higher than they are today, the coasts of Arabia, Iraq and Iran were up to 100 km inland from where they currently are. Thereafter, cooler temperatures locked up more and more of the Earth’s water supply in the polar ice caps, and sea levels fell accordingly. In the Gulf the successive phases of marine regression that ensued are marked by marine terraces at 110, 70, 50–45, 40–38, 25, 18–15 and 10–7 metres above modern sea level. Dating these episodes is difficult, however, and seismic activity, resulting in both tectonic uplift and subsidence, complicates our understanding of the geomorphology of the Gulf’s coasts.
Between 70,000 and 17,000 years before the present, when worldwide sea levels were up to 120 m below their present levels, the trough between the southernmost ranges of the Zagros Mountains in Iran and the Arabian Shelf was not the shallow, epicontinental sea that it is today, but a river valley through which the combined effluent of the Euphrates, Tigris and Karun rivers ran southward, through the Straits of Hormuz, and into the Arabian Sea or Gulf of Oman. From a purely geographical perspective, this palaeo-river divided the valley floor into an Iranian and an Arabian hemisphere. Yet the human effects of the river’s presence were probably very different. Instead of separating the populations of these two hemispheres, the river almost certainly acted as a magnet, drawing late Pleistocene bands of hunters and gatherers to its banks in search of game, water fowl, fish and shellfish, and creating opportunities for interaction between groups exploiting the resources of the river’s catchment area, at least on a seasonal if not year-round basis.
Unfortunately, we have no way of retrieving either the material remains of the earliest inhabitants of this region or the evidence of their interactions because, around 17,000 years ago, worldwide sea levels began to rise, eventually transforming the Tigris-Euphrates-Karun river valley into the Gulf as we know it. Several generations of geomorphologists and hydrologists have documented the progressive infilling of the Gulf, the implications of which are wide-ranging from the standpoint of human geography and prehistoric settlement. As late as 10,000 years ago sea levels were still c.40 m lower than they are at present, after which they rose by fits and starts. The valley became covered, in part, by a discontinuous lake, and only gradually assumed its present form. Thus, for example, at 6000 BCE, Bahrain was not yet an archipelago; Qatar was not yet a peninsula; and the coast of Abu Dhabi was significantly further north than it is today. In other words, Bahrain, Qatar and many of what are today the offshore islands in the area of the Great Pearl Banks, including Dalma, Marawah and Sir Bani Yas, were part of the Arabian mainland. Thereafter, the sea level in the Gulf continued to rise, reaching a high point c.2–3 m above modern levels about 6,000 years ago, before receding gradually to its current level at a rate of c.0.4 m/1,000 years, according to Kurt Lambeck. To complicate the situation, however, tectonic uplift and subsidence must also be considered, and Lambeck has emphasized that, west of the Straits of Hormuz, towards the eastern end of the Zagros Mountains, late Holocene beach deposits have been observed at elevations of 30 m above sea level which cannot be explained by ‘eustatic and glacio-hydro-isostatic models’. Moreover, due to the effects of local topography, the infilling of the Gulf did not proceed smoothly or uniformly, affecting all parts of the region at precisely the same rate. As such, it would be unwise to make generalizations about the situation throughout the Gulf at a given point in time, and as Bernier and his colleagues showed, the detailed study of micro-environments along the Gulf coast is the only way to arrive at insights into the situation in a specific topographic zone. Be that as it may, it seems clear that the infilling of the Gulf had many consequences.
From the outset we should dispense with the notion that the Gulf, any more than its predecessor, the palaeo-river formed by the combined effluent of the Tigris, Euphrates and Karun rivers, ever acted as a barrier to communication. As we know from excavated prehistoric sites in Kuwait (e.g. al-Sabiyyah) and surface finds made at coastal sites in eastern Saudi Arabia (e.g. Dawsariyyah and ʿAyn al-Sayh), boats made of wood and/or reeds covered with bitumen were being used by the inhabitants of the Gulf’s coasts by 6000 BCE. Thereafter, sailing technology was certainly not static and the sailors of 3000 BCE, 1000 BCE or 1000 CE had, at their disposal, vessels, sails and rigging that must have been technically more advanced than those of their ancestors thousands of years earlier. But technological constraints did not hinder communication and, as a glance at Whittingham and King’s Reed’s Tables of Distances clearly shows, the distances between the coasts are modest. Thus Bahrain is only 175 nautical miles from Bushihr (Bushire), while Bushihr is only 150 nautical miles from Kuwait. Similarly, the distance from Bandar Abbas to Bandar Lingah is only 110 nautical miles. Compared to overland travel, either on the Iranian Plateau, particularly if the Zagros Mountains had to be traversed, or through the east Arabian desert, seaborne travel in the Gulf was relatively easy, both for the transport of goods and people.
While the holistic approach to the study of the Gulf encouraged by the above reading of its geomorphology and hydrology is appealing, it has found much less favour among archaeologists than historians. This is due to a constellation of factors and reflects a scholarly trajectory that needs to be dissected before it can be properly understood.
Dysfunctional archaeologies
Archaeological research in the Gulf began in the late 1870s and is thus much younger than in other parts of the Near East such as Turkey, Syria, Iraq, Egypt or continental Iran. By the early twenty-first century, hundreds of sites had been identified along the east Arabian coast, from Kuwait to Oman, and on many of the major islands fronting it (e.g. Faylaka, Tarut, Bahrain, Sir Bani Yas, Dalma, Marawah). Yet the evidence from the Iranian coast is largely limited to the areas around Bushihr and Siraf and to just a few of the Iranian Gulf islands (principally Kharg and Kish).1 The imbalance in the quantity of data available from the two sides of the Gulf has an obvious effect on our understanding of the past and may be attributed to a variety of factors. Fundamentally, this discrepancy reflects the relative intensity of survey and excavation along the Arabian seaboard, particularly in the past thirty-five years, and the much more desultory nature of the exploration that has occurred along the Iranian coast. In fact, the 1979 Islamic revolution, which ended decades of intensive archaeological research in continental Iran, prompted many scholars who had formerly worked in Iran to initiate survey and excavation in eastern Arabia (principally the UAE and Oman). In addition, the isolation of Iran since the 1979 Islamic revolution has made it difficult for scholars in the West to access what information has been published in Iran on research conducted along the Iranian coast, and for Iranian scholars to access publications concerned with sites in Arabia.2
But there are other reasons why it has been difficult to discern patterns and establish correlations between the Arabian and Iranian hemispheres; indeed, why very few attempts have been made to unite the two hemispheres into a coherent whole. First, the archaeology and early history of settlement in the Arabian hemisphere has generally been studied by a different set of scholars than those concerned with the Iranian evidence. Second, indigenous scholarship in virtually all of the countries bordering the Gulf – Oman, the UAE, Saudi Arabia, Qatar, Bahrain, Kuwait, Iraq and Iran – is overwhelmingly national in orientation and there is little engagement with work being done outside the nation in which scholars in these countries work, even when it is clear that the pre-modern archaeological remains found in one state are related to those of another.3 Although meetings of archaeologists and museum professionals from states of the GCC do occur, few of the participants have the overview of the corpora of contemporary and related archaeological finds from across the region that Western academics typically have. And finally, the politicization of the terms ‘Persian’ and ‘Arabian’ Gulf has obscured the common histories and archaeologies of the two hemispheres.
Despite these mitigating circumstances, however, it is obviously tempting – and timely – to try to view the Gulf region holistically. In the current, admittedly imperfect state of our knowledge about the Iranian coast, it certainly does seem as though this area was less densely occupied (density being a relative term) than the Arabian shore (including major islands like Faylaka, Tarut, Bahrain and Muharraq), particularly in the pre-Islamic period. A survey by Andrew Williamson and Martha Prickett (published by Seth Priestman and Derek Kennet) identified hundreds of sites along the coast, overwhelmingly Islamic in date, but the evidence of early sites is slim. It is frustratingly difficult, however, to interrogate the available evidence with a view to determining whether ...