Exploitation, Transport and Circulation of the Rohri Hills Chert (Sindh, Pakistan) during the Indus Period

1. Introduction

This paper discusses some aspects of the Bronze Age Indus Civilization which flourished in the north-western regions of the Indian Subcontinent during the 3rd millennium cal BC [1]. The paper focuses on the Indus chert mines of the Rohri Hills, in Upper Sindh, where the Italian-Pakistani Joint Rohri Hills Project carried out surveys and excavations in the 1990s [2]. The Rohri Hills have yielded the largest chert mining areas of the Indian Subcontinent whose exploitation took place during the Bronze Age. The main scope of this paper is to define the way the chert mined from the hills was transported, how, where, when, and why. To reach the goal, the landscape around the Rohri Hills, the way the Indus and Hakra/Ghaggar (henceforth Hakra) Rivers changed their courses [3,4,5,6,7], and the distribution of the Indus sites so far known in the area will be considered [8].

W.T. Blanford was the first to write a description of the Rohri Hills in 1880. In his seminal volume on the geology of western Sindh he reported the characteristics of the Kirthar limestone which constitutes the hills [9]: 45. “These hills form a rocky area, surrounded by all sides, except the south, by the alluvial plain of the Indus, and slightly broken up at their northern extremity, where they are intersected by the river. To the south the rock is covered up by blown sand, but the latter, farther to the southward, appear to rest on alluvium” [9]: 101. Another point of great interest in Blanford’s description is “an escarpment along the western side of the hills, and the beds dip thence with a steady gentle slope, rarely exceeding 2°, and in many places less than 1°, to the eastward” [9]: 102. The presence of chert artefacts is reported in the following page: “The flints weather out and cover the surface throughout a large area; cores and the flakes split form them being scattered about in abundance in some places” [9[]: 103.

The first knapped stone artefacts were discovered near Sukkur a few years before by Lieutenant-General G. Twemlow “three feet below the rock in the bed of the river (Indus)” [9,10,11]. The hills did not attract the attention of other geologists until 1937, when H. De Terra and T.T. Paterson visited Sukkur. These authors were the first to suggest a chronology for the lithic assemblages discovered near Rohri [12], provide a schematic profile of the hills around Sukkur and Rohri with the location of the knapped stone artefacts in situ, and describe their stratigraphic position on the top of a silt deposit above which “lie migrating dunes” [13]: Figure 193. In their paper they report the presence of “conical cores and very thin long blades reminiscent of the stone industry of Mohenjo Daro” [13]: Plates XLIX and L, which are characteristic Indus Civilization artefacts.

The Indus chert workshops became even more important after the publication of the knapping floors discovered on the terraces that rise south of the city of Rohri by B. Allchin [14]. More recently two chert mines excavated along the western edge of the Shade Shaheed terraces were AMS-dated to the Mature Indus Civilization, more precisely to the middle of the 3rd millennium cal BC (RH-862: 3880±70 BP: GrA-3235; RH-59: 3990±24 BP: GrM-21237) [2,15].

Environmental Setting

The characteristics of the landscape surrounding the Rohri Hills, the presence of two important rivers with very different courses and behaviours, the Indus, in the west, and the Hakra, in the east, and the thick alluvial Indus Plain [16]: 83, will be considered before discussing the way chert artefacts and other lithic material, among which are limestone blocks, were transported to the Bronze Age Indus centres (Figure 1).

Blanford reports the presence of “an escarpment along the western side of the hills” [9]: 102, ca 35-40 m high, east of the shrine of Shadee Shaheed (Figure 2). The most important groups of chert mines were discovered in 1986 on the top of this terrace (ca 135-140 m a.s.l.), which gently slopes towards the east down to the Thar Desert dunes and the Nara Canal which flows partly in an old bed of the Hakra River (ca 65 m a.s.l.).

The satellite and aerial photographs show that the surface of the Rohri Hills is deeply incised by Pleistocene streams which flew down to the Nara Canal and the westernmost Thar Desert fringes [17]. At least seven superimposed seams of good-quality chert are embedded in the almost vertical nummulitic limestone escarpment of the Shadee Shaheed Hills (Figure 3). However, only the uppermost one has been exploited by the Indus miners to extract nodules.

Today the Indus River flows ca 40-50 km west of the westernmost edge of the hills. The alluvial plain consists of deposits at least 6m thick [5]: 295, 8-15m at Mohenjo Daro [18]: 376, which accumulated between the end of the Pleistocene and the recent Holocene [19]: 257. Its surface is incised by ancient river bends the chronology of which has never been defined.

The Indus River changes, inundations [20,21], and the location of freshwater resources in relation to ancient settlements [22]: 50, have been debated for decades with different results [16,22,23,24,25]. The available data show that the territory around the hills was very different from that of the present a few centuries ago [26]: 321, and also during the 3rd millennium cal BC [27,28]. They suggest that the plain has never been an easy landscape to cross because of several natural obstacles: “ease of traversing will depend on topography, on the presence of barriers or impediments, and also on the mode of travel” [29]: 393. We know that when the Mature Indus Civilization flourished the climate was slightly more humid than that of the present, and the changes that led to the present arid conditions began to establish around the end of the same millennium [30,31].

2. Materials and Methods

2.1. The Bronze Age Indus Sites

The distribution of the Indus sites around the Rohri Hills can help us to interpret the way chert was transported, supplied, and where lithic tools were manufactured. Four settlements located within a radius of ca 70 km from the Shadee Shaheed mining area, have been considered for this work. Two are located west of the Indus (Lakheen-Jo-Daro and Mohenjo-Daro), two, east of the river (Kot Diji and Naru Waro Dharo). Unfortunately, our knowledge of the precise quantity and typology of knapped stone artefacts collected from the aforementioned sites is limited by several factors the most important of which are the retrieving techniques employed during excavation. According to the excavation reports, none of the four Indus sites has been water-sieved to fine mesh, with the exception of the bead manufacturing workshop of the Moneer South East Area (MNSE) at Mohenjo-Daro [32].

The Mature Indus city of Lakheen-Jo-Daro is located in the industrial area in the north-western suburbs of Sukkur, close to the right (western) bank of the Indus River (Figure 1, n. 1). Lakheen-Jo-Daro is a large urban settlement whose size has been recently compared to those of the Bronze Age cities of the Greater Indus Valley [33]: 125. The excavations carried out at Lakheen-Jo-Daro yielded many impressive burnt brick structures, among which is a wide road, a small part of which has been excavated [34]. One ordinary radiocarbon date obtained from charcoal shows that at least part of the settlement flourished during the Mature Indus phase, probably around the middle of the 3rd millennium cal BC or slightly later (3960±140 BP: GrN-23123) [28]. According to the excavators, different areas of the site yielded a few knapped stone artefacts made from Rohri chert. They consist of one subconical blade core, unretouched blade and bladelet fragments, a few flakes and crested blades [34].

Mohenjo-Daro is the largest ever excavated and best preserved Bronze Age metropolis of the Greater Indus Valley [35]. Its location and existence, along the right (western) bank of the Indus (Figure 1, n. 2), has been linked with “new means of transport by ship, necessitating settlements close to the river” [36]: 16. Despite its importance, impressive architecture, and the great variety of brick features, among which are hundreds of habitations with superimposed structures [37,38], the chronology of the site is based on only six radiocarbon dates, marking the “poor archaeological control over the remains of the Indus Civilization” in general [35]: 36. They were obtained from burnt wood and charred straw sampled from the excavations carried out by G. Dales [39] at the western edge of the HR mound [40]. The dates fall between 3913±64 BP (P-1179) and 3702±63 BP (P-1182A). The results are generally accepted, while the traditional relative chronology based on sequences of structural units is considered inconsistent [41]: 68, because “no more than an approximate stratigraphical interpretation can be attempted on the results to date” [42]: 163.

A generic description of the Mohenjo-Daro chert artefacts has been provided by J. Marshall [43]: 458. This author illustrated only 12 typical Indus blades, 4 subconical blade cores, and 3 large, retouched chert blocks that he interpreted as agricultural implements [43]: Plate CXXXI: 1-19). G. Dales’s excavations yielded 115 knapped stone artefacts [44]. Thanks to the accurate mapping and water sieving of the deposits [32,45] ca 4000 lithic artefacts, debitage and debris flakelets have been recorded from the bead manufacturing workshops of the “Moneer South East Area (MNSE) of the town, part of a more general surface survey of the craft areas of the ancient urban settlement” [46].

Kot Diji is a mound with a complex sequence located ca 15 km south-west of the shrine of Shadee Shaheed [47]: 43. (Figure 4). The site has been radiocarbon dated between 4421±141 BP (P-196) and 3925±134 BP (P-195) from charcoal samples collected from four different layers [48]. Kot Diji lies very close to the westernmost fringes of the Rohri Hills in an area of ancient river bends (Figure 1, n. 3). The excavations carried out in the 1950s yielded a few knapped stone artefacts among which are cores, blades, sickles and leaf-shaped, bifacial arrowheads [49]: P-62. The lithic assemblages have never been studied in detail. Consequently, it is not possible to distinguish the artefacts of the Early Bronze Age Kot Diji phase from those of the Mature Indus period.

Naru Waro Dharo is a large site that rises from the alluvial Indus Plain ca 8 km north of Kot Diji close to an ancient meander bar of the Indus (Figure. 1, n. 4). The excavations carried out in 1955 did not yield structural remains. Though the site is undated, it has been attributed to the Indus Civilization according to the typological characteristics of the pottery assemblage. The excavator has reported the presence of “chert blades and cores” [50]. The site has been interpreted as a temporary Indus camp due to the absence of structural remains, of a well-defined archaeological horizon, and the presence of ceramic potsherds scattered over a large area [51]: 267.

The distribution of the archaeological sites in this region of Upper Sindh shows that the territory that extends between Shikarpur, in the north, and the Nara Canal, in the east and south, was almost unsettled during the 3rd millennium BC [52]: 179. However, apart from those discussed in this paper, a few more unexcavated and unpublished Indus sites are known in the region [52]: 181-183.

2.2. The Rohri Hills Chert Sources

In December 1975 the Cambridge Archaeological Mission discovered many chert working floors on a few limestone terraces located near Sukkur and Rohri. They were later attributed to the Indus Civilization due to the typological characteristics of the knapped stone artefacts and cores [14,53]. In the spring of 1986, chert mines of the same phase were discovered on the mesas east of the shrine of Shadee Shaheed, in the central-western side of the hills [54]. While the discoveries made by the Cambridge Archaeological Mission consisted of small working floors, those of the 1986 revealed the existence of impressive groups of chert mines characterised by different groups mining pits, heaps of limestone blocks and chert debitage flakes [55,56]. The investigation of the Shadee Shaheed area started a few years later by the Ca’ Foscari University of Venice (I) and Shah Abdul Latif University, Khairpur (PK) “Italian-Pakistani Joint Rohri Hills Project” [57].

The surveys and excavations carried out in the Shadee Shaheed Hills showed the importance of chert in the economy of the Bronze Age Indus Civilization. Before the 1980s, knappable raw materials had been very rarely considered by most archaeologist working in the Greater Indus Valley [58], with very few exceptions [59,60]. The 1993 surveys led to the discovery of thousands of chert mines, workshops [61,62] and other features [57] (Figure 5), very few of which were partly excavated in the following years [2,63,64]. They showed the importance played by knappable stones in a complex Bronze Age urban civilisation and the way chert was extracted from open-air pits to extract good-quality nodules embedded in the uppermost seam of the mesa (Figure 3). Only two Shadee Shaheed mine pits have been radiocarbon dated [15]. The results showed that they were exploited around the middle of the 3rd millennium cal BC, when the Mature Indus Civilization flourished in the region. Unfortunately, radiocarbon dates are not available from the working floors discovered near Rohri and Sukkur by the Cambridge Archaeological Mission in the 1970s due to the absence of organogenic material.

Millions of chert artefacts and impressive heaps of limestone blocks, debitage and waste flakes were visible on the surface of the Shadee Shaheed Hills in the 1990s (Figure 6). The discovery of the Rohri Hills mines has opened many questions regarding the way chert was exploited (Starnini and Biagi, 2006), raw material, finished and unfinished products were transported to the Indus centres, technology and characteristics of the products [65], relationships between producers and consumers [61], the way they were transported, and “what were the most frequent activities in which chipped stone tools were used” [66]: 99. The latter point is very important because we still know very little of it due to the absence of finished tools in the Rohri mining areas and workshops, which is another important argument of discussion [67]. In contrast, specific implements, among which are locally made straight perforators and micro-drills [64] have been recovered from the bead manufacturing areas of Mohenjo-Daro [32], Chanhu Daro [69] in Sindh.

Other urban centres have yielded different types of chert implements among which are notched sickles [70], blades used in agricultural activities [71] and pottery manufacture at Nausharo in Balochistan [72], long end scrapers [46], very few other types of tools [73], and subconical blade cores. Were only finished chert artefacts transported to the Indus centres or both finished and unfinished products and also cores and nodules? What was the market demand like [74], and how it “determined the amount of time and energy expended” in the hills “opposed to other economic and noneconomic activities” [75]: 65? Unfortunately, the only available data come from workshop 480 in the Rohri Hills, from which at least 2000 entire blades were manufactured and exported after removing the percussion bulb [76]. This datum is important to have an idea of the impressive amount of released debitage flakes (Figure 6) and the relationships between production and transportation [77]: 7. Was chert mining a seasonal activity due to the severe, desert summer heat and the monsoon precipitations that followed? We know that today limestone mining takes place during the winter months due to harsh climatic conditions.

2.3. Land and Fluvial Transportation

How was chert transported to the Indus sites located around the hills? Some important characteristics of the central-western side of the Rohri Hills, namely the region close to the shrine of Shadee Shaheed, have been provided by W.T. Blanford [9]: 102. This part of the hills, where the most impressive groups of chert mines are located, has very steep sides, the terrace which is uneasy to reach on foot, is inaccessible by wheeled wagons or carts. We can suggest that other routes were followed to transport lithic material according to “a balance struck between production and transportation costs” [77]: 7.

For sure, the workshops discovered in 1975 by the Cambridge Archaeological Mission in the northernmost part of the hills, where the Indus flows close to the archaeological area, were easier to access. This means that water transport was employed to move chert and limestone blocks to the city of Lakheen-jo-Daro. This is also the case for Mohenjo-Daro, which is located along the right (western) bank of the Indus. In both cases a large river with unique characteristics and dangerous waters in some periods of the year had to be crossed [36,78]. We can assume that flat-bottomed, sliding boats (Figure 7) similar to those of the Indus fishers or Mohanna [79,80] were used to complete such a difficult task also because we know how versatile they are [81,82]. The way the distribution took place and was regulated, is not known. However, a recent analysis of the Indus seals would suggest that these specimens were utilised also to control trade activities which “would have been in the hands of specialist merchants” [83]: 179, although so far we do not have direct evidence of signs related to transport of chert or limestone blocks [84].

The sites of Kot Diji, which is located along the western fringes of the hills, and Naru Waro Dharo, were probably accessed by land. However, the complex landscape of the Indus Plain (Schuldenrein, 2002) and the presence of river bends and forests undoubtedly made its crossing an uneasy task [85,86]. We know that during the Indus phase different types of wheeled transport (carts and chariots) were utilised. This is shown by many complete and fragmented ceramic toys which have been uncovered from the Indus Civilization sites distributed throughout the entire Greater Indus Valley [87].

Following the observations reported above, we can assume that lithic material transportation and distribution towards the Indus centres located around the Rohri Hills was made either by land or water, or both. A more complex system can be suggested for the sites located further away. We do not know the precise radius of distribution of the Rohri chert, although production system is undoubtedly tied to transportation [77]. For sure the distribution of some types reached Harappa in the Punjab and Nausharo in Balochistan [88] in the north, Badin and the Hab River mouth in Lower Sindh [89], and some areas of the Gujarat (India), in the south [90,91]. When did this happen given that the lower Indus Valley is inaccessible for land transportation for some five months a year? [92]: 124.

3. Results

“The identification of the types and sources of the raw materials is not easily achieved and will always leave scope for some ambiguities unless it is backed by precise and extensive scientific evaluations” [93]: 123. This statement is important for the interpretation of mobility and transportation patterns. However, we have to point out that the importance of knappable material, chert in particular, has always been underrated by most archaeologists working in south Asia also in the recent past [57]. This observation is important if we consider the way other materials, pottery in particular, but not only pottery, precious and unique items for example, have always been given a special role not only in constructing cultural and sometimes ephemeral chronological seriations, but also economic strategies and trade patterns. Apart from south Asia, this view has been put forward in other regions of Eurasia especially as regards the Bronze and other metal ages [15]. Reverting to the Indian Subcontinent in general, and Sindh in particular during the Indus phase, attention has sometimes been turned more towards transoceanic and Central Asian contacts than the problem of internal request, movement and production of different types of goods [94].

Some of the problems related with lithic mining, transportation and utilisation of knappable material have been recently re-discussed in the wider context of the importance played by different varieties of knappable raw materials in some countries of Eurasia during the Bronze Age, their exploitation, meaning and social dimension [15]: 422. The research underway in the Greater Indus Valley has changed the obsolete idea of the almost irrelevance of the “lithic factor” during the Bronze Age [95,96]. This fact has been given more relevance thanks to the discovery of the Rohri Hills chert mines, which unfortunately have been almost completely destroyed mainly during the last decade making an irreparable damage to the study of the Indus Civilization and its economic strategy [2,97].

Though we still know almost nothing of road systems in the Indus Valley during the Bronze Age, they have been recovered in other countries where the problem of transportation has been extensively discussed also recently [98].

4. Discussion

The difficulty of defining transport methods and routes followed during the 3rd millennium cal BC Indus communities has already been emphasised by R. Law [88]: 301, who nevertheless remarked the “enormous interaction system – especially in regard to its rock and mineral artefacts”. However, our knowledge is still very scarce at least as regards the movement of chert and limestone blocks all over Sindh. Apart from the exploitation of chert for the activities reported above, we know that this material was utilised also for the manufacture of different types of cubic weights [99]: Plate 113, though so far we do not have evidence of any chert weight workshop either in the Rohri Hills or in the Indus cities of Sindh. As far as we know, squared limestone blocks were employed to cover one of the curved brick drains of the Great Bath in Mohenjo-Daro [41] and to reinforce the basement of brick wall structures (Figure 8). We can suggest that these blocks were mined from the Rohri Hills, which are located less than 50 km east of the city, as the crow flies.

Our knowledge regarding other regions of Pakistan, the Punjab for example, is even scarcer. We know that the Cholistan Desert around Derawar is very rich in Bronze Age sites [100,101], some of which are urban centres of an impressive dimension, Ganweriwala for example (Figure 9) [102], that were built along the banks of the Hakra River which changed its course around the end of the Bronze Age [3]. We can suggest that also the Hakra River was exploited for transporting lithic material from the Rohri Hills to the cities of the northern regions. In this respect, it is important to point out once more that the Rohri Hills terraces are inclined to the east, towards the ancient course of the Hakra River which, in Indus times, lapped the eastern side of the hills.

If our knowledge of the way chert was supplied to the Bronze Age cities of this part of Upper Sindh is still insufficient, we have to admit that we know even less of other regions of the Greater Indus Valley. In her paper on the transportation systems employed during the Indus phase and historical periods, Miller [103] discussed several aspects of land and water transport. Very recently, scientific analyses have shown some evidence of land, bovid-driven, short and also long-distance transport [104]. However, to improve our knowledge the number of isotopic analyses would need to be multiplied.

5. Conclusions

The problems of resources, technology, trade and transportation have been largely debated by the specialists of the Bronze Age of Europe [105]. The same can be said of the presence of wheeled chariots and carts during the 3rd millennium cal BC, and the complex structure of some of these vehicles already when they made their appearance in several countries of Eurasia around the beginning of the metal ages [106,107], roughly same period when they started to be utilised in the Indian Subcontinent [87]: 6.

V.G. Childe was the first to distinguish between two and four-wheeled wagons, to discuss where and when they were in use. Regarding the Indus Civilization, he was also the first to suggest that carts and wagons were utilised “from a considerable distance over land on which carts would easily travel” [108]: 178, “played the same role in provisioning the great Harappan cities” [108]: 183, “transport was facilitated by wheeled vehicles and boats” [109]: 176, and “in Egypt, where all cultivable land lies close to the Nile, boats took the place of carts” [108]: 177. All these observations, which were written more than seventy years ago, are still actual today, and are part of the debate on the transport of materials within the Greater Indus Valley.

The transportation problems put forward by the aforementioned authors are basically the same that we are facing today, despite our knowledge has improved especially with regard to the characteristics of the territory where the Indus Civilization flourished, the chronology of the most important events, raw materials exploitation and trade, and mining activities.

Though the present paper has been centred mainly on an underrated problem, chert and its transportation, the general impression is that it can be extended to other aspects of great economic or subsistence importance whose study has not improved during the last decades, plant use and agriculture for example [110,111] because “surges in agricultural productivity follow the growth of cities” [112]: 7.

It has been observed that after over a century of the discovery of the Indus Civilization [113,114], many problems have not been solved [115], one of which is for sure transportation.