Revisiting the Joshipur Rock shelter Complex (Sehore District, )

Parth R. Chauhan1,2,3, K. Krishnan4, Ketika Garg1, Shubham Pal1, Bharti Singh1 and Avinandan Mukherjee4

1. Indian Institute of Science Education and Research, Mohali, Punjab, (Email: [email protected]) 2. Stone Age Institute, Gosport, IN, USA (Email: [email protected]) 3. Department of Anthropology, Indiana University, Bloomington, IN, USA (Email: [email protected]) 4. Department of Archaeology and Ancient Indian History, Faculty of Arts, The Maharaja Sayajirao University of Baroda, Vadodara ‐ 390002, , India (Email: [email protected])

Received: 27 August 2015; Accepted: 19 September 2015; Revised: 06 October 2015 Heritage: Journal of Multidisciplinary Studies in Archaeology 3 (2015): 105‐127

Abstract: This paper describes a rock art complex near Joshipur north of the Narmada River in the Sehore District of Madhya Pradesh. First reported several decades ago (IAR 1975‐76), this cluster of rock shelters and overhangs forms a part of the larger complex of rock shelters in the central India. Preliminary surveys indicate that the Joshipur complex comprises at least ten rock shelters, many of which preserve rock art, microliths and bone specimens. This paper broadly describes and illustrates some of the more prominent ancient paintings and associated evidence and contextualizes its broader regional significance.

Keywords: Central Narmada Basin, Joshipur Rock shelters, Rock Art, Microliths, Bones, Lithics, Sediments

Introduction The Indian Subcontinent has a lengthy history of prehistoric occupation stretching back to at least 1.5 Ma (Pappu et al. 2011) and possibly even up to ~2 Ma (Dennell, 2009). This entire record represents multiple biological and technological dispersals into and within the region, presumably all from the western direction ‐ although the northeastern zone should not be ruled out; as dispersal corridors in both directions. Recent debates have centered on the timing(s) and environmental contexts of modern human arrival into the Subcontinent and associated technologies (e.g. Chauhan et al. 2015; Groucutt et al. 2015). In the case of multiple dispersals of moderns into within the last 100 kyr (Dennell and Petraglia, 2012) it is difficult to pinpoint which modern human population(s) first created rock art and art in general. This also applies to the uneven pattern of symbolic behavior in the South Asian archaeological ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

record (James and Petraglia, 2005). This is because the older rock art occurrences are presumably heavily faded and not well preserved and hence undatable and often undiagnostic. In the Indian Subcontinent, no well‐preserved paintings have yet been recovered in deep caves as known from Europe for example. Such deep caves are rare in South Asia and the known ones do not appear to have been inhabited or significantly exploited during prehistoric times; most of the paintings are in rock shelters and thus steadily exposed to the natural elements. South Asian rock paintings are predominantly spatially associated with microliths more than any other stone technology. Though claims have been made of some rock art to be Lower Palaeolithic in age (e.g. Bednarik, 2013), they remain controversial and no comparable supporting material has come forth from other contemporary sites. Especially at multicultural rock shelters that preserve lengthy cultural sequences, it is highly difficult to associate the rock art with a specific cultural horizon since the rock art has not been adequately dated. Thus at some rock shelters, there may be variable temporal gaps between the rock art and the spatially associated lithic assemblages (regardless of whether they are Palaeolithic or Mesolithic). Therefore it is provisionally understood that the makers of the vast majority, if not all, of the South Asian rock paintings were the same modern human groups that produced the often‐associated microlithic assemblages. Since there are very few absolute dates for any of the South Asian rock art, its age bracket is relatively wide and individual estimates disputable. South Asia’s rock art occurrences have generally been chronologically bracketed using attributes, such as, style, superimpositions, patina, theme, colours, subject matter, and archaeological evidence associated with the rock paintings. Recently, a petro glyph in the Billasurgam Cave in was dated to the mid‐Holocene (Tacon et al. 2013) and attempts have also been made to date the rock art at Bhimbetka (Bednarik et al. 2005). Recently, some Southeast Asian rock paintings was dated to ~40 ka (Aubert et al. 2014), making them broadly contemporary with their European counterparts and invoking tantalizing possibilities of independent or parallel development of symbolic behavior. It is also broadly contemporary with the earliest microliths in South Asia (~45ka) known from Mehtakheri, Madhya Pradesh (Mishra et al. 2013). Although cultural and genetic connections between South Asia and Australasia appear to have been established rather late, the earliest art in South Asia may be older if assuming modern humans arrived here first. In any case, all visible rock art is currently thought to be no older than the Upper Paleolithic period.

South Asian rock art basically includes paintings using mixed pigment types, abstract engravings, etched or pecked designs and cupule marks (e.g. Pandey, 1993).The themes of the paintings and bruisings were diverse and include animals, abstract designs and depictions of humans engaged in different activities ranging from hunting to dancing. These paintings have been reported to occur in a range of colors including green and yellow pigments with red and white paintings being the most common. While pecked designs and etchings (i.e. petroglyphs) are restricted to the northernmost and southernmost geographic zones of the Subcontinent, paintings in rock shelters ‐ and on large rocky surfaces ‐ have been reported throughout the Subcontinent

106 Chauhan et al. 2015: 105‐127

(Kennedy, 2000; Neumayer, 1983; Pandey, 1993). The densest occurrences of rock art are concentrated in central and southern India with other examples of isolated but rich sites at numerous zones in between (e.g. Mathpal, 1992). The best known rock shelter complex in the Indian Subcontinent is the World Heritage Site of Bhimbetka south of Bhopal in central India. Located further south of Bhimbetka are additional well‐known painted rock shelters such as Adamgarh, the Saru Maru complex and the Pachmarhi occurrences. These sites are a part of the Narmada Basin, most famous for its rich preservation of vertebrate fossils, known since the 19th century and studied by various scholars since then (e.g. de Terra and Paterson, 1939). This river basin is also the region which has yielded the only‐known pre‐modern Homo fossil(s) in the entire Indian Subcontinent and appears to have a lengthy fluvial history spanning at least ~800 kyr (Rao et al. 1997)along with the archaeology (Patnaik et al. 2009). Prehistoric sites are found here in the form of surface scatters in sedimentary and bedrock contexts as well as in stratified occurrences. The aim of this paper is to briefly describe one of the lesser‐ known rock art complexes nearby at Joshipur and outline its regional archaeological potential and significance.

The Joshipur Rock shelter Complex The concerned rock shelter complex (N 22°47’54.26” E 77°43’36.94”) is located in the southeastern part of the Sehore District of Madhya Pradesh and about 4 to 5 km slightly northeast of Hoshangabad town across the Narmada River (Fig. 1). The nearest village is Joshipur situated on the northern bank about of the Narmada River directly across Hoshangabad. As a result of previous work, at least 45 rock shelters have been documented in the immediate region between Shahganj and Pangoraria. In the 1970s, the Prehistory Branch of the Archaeological Survey of India reported numerous archaeological discoveries from both Hoshangabad and Sehore Districts of Madhya Pradesh (IAR, 1974‐75; IAR 1975‐76), following upon the earlier prehistoric surveys of de Terra and Paterson (1939). Led by K.D. Banerjee, team members included B.P. Bopardikar, P.R.K. Prasad, A.J. Nambhiraju, S.K. Gulrandhe, K.S. Venkataramaiah and B.K. Rudra. Other rockshelter groups known from this region include Mou, Talpura, Unchakhera, Bayan and Pangoraria, the latest studied being the Saru Maru complex (Simte, 2015). Shaik and Chauhan (2013) also reported a painted rockshelter at Bandarjari (10 km northeast of Saru Maru) which may have been along a trade or transit route and thus linking the rock shelters near the Narmada River with those in the Bhimbetka area. Many of these rock shelter complexes represent multi‐cultural sites indicating repeated visits and/or occupation(s) following the prehistoric occupation. For example, the site‐complex of Saru Maru also includes Buddhist stupas and may have been a monastic complex accommodating small groups of priests (Simte, 2015). At Bhimbetka, the cultural‐stratigraphic sequence ranges from the Acheulean to the Mesolithic (Misra, 1985) and at Adamgarh, up to Early Historical times (Joshi, 1978). At many of these South Asian rock art sites, prehistoric, protohistoric/Chalcolithic, Early Historical and Medieval paintings variably overlap. Such lengthy and diverse archaeological histories at individual locations reflect the

107 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

geographic, ecological and socio‐cultural importance these locations had for multiple cultures over time. For example, one Joshipur rock shelter (JRS‐1) at the base of the Vindhyan Hills had been turned into a modern Hindu shrine (Fig. 2) and contains saffron flags, tridents, offered bangles, coins and ‘Jai Shree Ram’ (and names of other deities) painted in Devnagri script amongst the prehistoric paintings (Fig. 3). Likewise, one of the rock shelters at Bhimbetka has served as long‐term Hindu temple for decades and another location below a painted rock shelter at Digamber (8 km north of Hathnora) has now been turned into a Hindu temple and local ashram.

Figure 1: General location of the Joshipur rock shelter complex in relation to the Narmada River and other known rock art sites near Hoshangabad town

In their report in IAR (1975‐76:28‐29), the investigators mention a total of eight rock shelters at Joshipur, five on the slopes of the Vindhyan Hills and three at the top of the range. They mention that paintings occur here in red, orange, yellow and white, with many superimposed upon one another. One of the more common scenes depicted in the paintings is “… deer‐hunting. Besides, elephants, spotted deer in white ochre, antelopes with in filled geometric and non‐geometric designs are the animals represented. Hunters, clad in skin‐cloth, appear to be warriors of late periods. Superimposed drawings are in different styles. In some of the shelters Lower Palaeolithic and microlithic tools unrelated to the paintings were found.” They compare the Joshipur stratigraphy to an area southwest and closer to the Narmada River near Mou, from where they recovered microliths (i.e., flakes, flake‐ blades and cores) from a “red earth at the top of the hill near the rock‐shelter” and note similar stratigraphy at other rock shelters including Pangoraria and Unchakhera. A stream section near Mou comprised (from bottom to top) bedrock, a gravel‐bed and red clay; from the interface of the latter two, they reported “a number of small to medium unifacial chopper‐scrapers of pebbles, a broken ovate, discoidal scrapers, flakes and cores”.

108 Chauhan et al. 2015: 105‐127

Figure 2: Modern Hindu shrine outside JRS‐1

109 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 3: Modern Painting of Devnagri Script in JRS‐1

Our basic documentation at the Joshipur rock shelters was carried out over two separate summer field seasons in 2014 and 2015, respectively. During the first visit, we documented the rock art occurrences at the bottom of the Vindhyan Hills and during the second season, we documented the occurrences at top of the Vindhyan Hills. All of the Joshipur rock shelters discussed in this paper are situated within an area of 3 km2. The plains at the foot of the Vindhyan Hills are at approximately 300m AMSL and the highest point of the Vindhyan Hills at the Joshipur complex is approximately 580m AMSL. The local geology comprises Vindhyan quartzites as the bedrock, with outcrops of Deccan Trap (basalt) exposed at several locations to the northeast of Joshipur near Shahganj (Tiwari and Bhai, 1997). The local landscape also preserves recent structures in the form of rock piles and an observation tower (Figs. 4 & 5) ‐ both presumably made by the Forest Department. The main hill on which the Joshipur rock shelters are situated also includes several key locations that probably served as vantage points for game scouting for example, and from the highest point on the hill, one can clearly see the Narmada River 5 km to the south. The Quaternary sediments at the base of the Vindhyan Hills comprise eroded fluvial deposits alternating between pale brown sandy silts with calcrete (calcium carbonate) nodules and reddish‐brown silts. These sediments presumably represent a combination of fluvial activity of the Narmada River and/or local seasonal streams near the rock shelters and colluvium moving down slope from the hills to the north. These sediments variably yield typologically Middle Paleolithic stone artifacts as well as microliths; similar geological contexts along the Vindhyan foothills also appear to preserve faunal material, albeit fragmentary, in association with calcrete‐rich deposits. At Joshipur, lithics appear to be eroding out of these sediments or are at least geologically associated with them in surface context. Additional surveys and test‐pits are required to see if these occurrences on the plains at the foot of the Vindhyan Hills are contemporary with the ones associated with the regional rock shelters. The latter is the most common contextual occurrence of lithics around Joshipur.

110 Chauhan et al. 2015: 105‐127

Figure 4: Stone pile as a boundary marker, possibly made by the Forest Department

Figure 5: Watchtower at the top of the Vindhyan Hills (maker and age are unknown)

In fact, microliths are the primary stone tool types found within the rock shelters or near their entrances, i.e. in the vicinity of the drip line. At all the surveyed shelters, the surface occurrences ranged from a single specimen to over several dozen and one

111 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

shelter contained no lithics although future excavations may yield otherwise. In other words, almost all large ‘boulders’ (often a result of rock fall or weathering from the Vindhyan hills) variably preserved behavioural evidence in the form of microliths, rock art or both. Thus far, only less than a dozen of these isolated ‘boulders’ preserved no archaeological evidence, suggesting that these ‘rock falls’ have been occurring since, during and following prehistoric human occupation of the region. However, extensive geomorphological and petrological studies are required to confirm such processes and their timings. Also, many of the rock shelters and isolated ‘boulders’ had rough and unsuitable surfaces due to intense chemical and geological weathering and thus possibly explaining the absence of paintings on such contexts. Such locations also rarely yielded any microliths, possibly indicating the variable use of such locations and during the prehistoric human occupation of this region. However, distinguishing between true behavioural absence and a complete absence of paintings due to weathering will prove to be difficult. Nonetheless, different locations/shelters/boulders on the slopes of the Vindhyan Hills probably served respective functional purposes, whether one was used exclusively for making art, manufacture and/or utilization of microliths or simply serving as temporary shelter from rain and heat during daily movement across the regional landscape. Many of these isolated boulders lie directly on the bedrock (beds and exposures of Vindhyan quartzite and quartzitic sandstone) or preserve shallow deposits of fine sediments including slope‐wash, colluvium and in situ weathering of the bedrock. Several of the boulders preserved evidence of ant or termite mounds (Fig. 6).

Figure 6: Termite/ant nest under one of the fallen boulders of Vindhyan quartzite

112 Chauhan et al. 2015: 105‐127

On a broader scale, our surveys for rock shelters and lithic scatters ultimately extended up to the eastern boundary of the Sehore District, several kilometres north of the Narmada River. At one location, local villagers warned us that it was difficult to access some of the shelters (long known to them) due to the dangerous presence of numerous bees. This demonstrates that many more rock shelters and related prehistoric sites remain to be studied and documented in this region. For example, our general observations in this part of the Vindhyan Hills revealed that isolated rock shelters with paintings are more common than previously thought and require proper documentation to understand the cultural landscape at multiple levels. On occasion however, we had difficulty in identifying specific painted rock shelter sites as being previously reported/discovered or not. This is primarily because most of the older publications that have reported rock art sites have not included adequate location and descriptive details such as GPS coordinates, local or regional landmarks, multiple overview photographs of the shelters and paintings and so forth. As a result, we don’t always know if specific rock shelters have been previously discovered (e.g. Shaik and Chauhan, 2013) including at Joshipur. However, technological aids such as Google Earth, for example, are making it easier for archaeologists to provide more specific and detailed information following the discovery of new sites. The remainder of this paper provides a general description of the Joshipur rock shelters, associated rock art and the archaeological evidence. A more comprehensive study of all of these aspects is in progress, which will include measurements of the individual paintings, more detailed descriptions as well as reconstructions of panels in schematic black‐and‐white line drawings.

The Rock Art of the Joshipur Rock Shelter Complex Although the previous investigators reported eight shelters, we identified and documented ten individual shelters. This discrepancy between our study and the previous one may be a result of the physical nature and uneven shapes of the monoliths, making it difficult to always demarcate one rock shelter from another when they are lithologically connected. We counted those rock shelters as individual if they were physically separated from each other and without paintings in the intermediate areas, focusing on discrete clusters of painted panels/walls/surfaces. Due to their proximity and the low number of paintings in them, we have combined two individual rock shelters and labeled them collectively as Joshipur rock shelter 5 (JRS‐5).Some of these rock shelter floors variably preserve evidence of pottery fragments (Fig. 7) and lithics (Fig. 8), occasionally both within the same shelter. Most of the pottery shards on the surface appear to be of recent age although archaeological specimens may be buried and recovered in the future (e.g. Simte, 2015). The microliths are made on quartz and varieties of crypto‐crystalline silica; chert, and chalcedony. In addition to sediments, the rock shelter floors are generally comprised of dry vegetation, rubble or fragments from weathered bedrock or shelter walls and large fallen slabs of rock from the shelter ceilings/walls. Under one of these large boulders of Vindhyan quartzite, modern bones were also observed, representing carnivore/scavenger activity in the

113 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 7: A potsherd found on the floor of JRS‐1

Figure 8: Rich scatter of microliths outside JRS‐1 area (Fig. 9). This hints at a future potential difficulty in distinguishing between modern or recent bones and sub‐fossils in association with archaeological/natural horizons within the rock shelters. Unless dated directly or through their respective strata, such datasets will prove to be difficult to interpret and describe particularly highly fragmented faunal materials. Unlike the deeper cave contexts which better preserve the faunal evidence such as Kurnool Caves in Andhra Pradesh and Badatomba‐lena in Sri Lanka (see Chauhan, 2008), the central Indian rock shelters are largely ‘shallow’ and thus the contained paleoanthropological material is usually exposed to the natural elements. Our general observations indicate that proper comprehensive surveys are required, and the number of rock shelters and their area of

114 Chauhan et al. 2015: 105‐127 occurrence in this region may increase significantly. When combined, we noticed a total of 410 individual paintings out of which 79 were indiscernible paintings or patches/clusters of paintings, their preservation being affected by the natural elements and other processes such as exfoliation and chemical weathering. In some of the rock shelters, the paintings are either so faded or so clustered and superimposed, that they are difficult to identify and count individually. Many more paintings probably existed here before slowly fading away over time due to these various processes. Nonetheless, we have counted, classified and tabulated all the visible individual paintings in Table 1 and include human figures, animal figures, plants/trees, abstract designs, modern graffiti, a chariot, a trident and other images. We feel that future publications reporting new rock art sites for the first time should conventionally include such tables and other basic data that is not generally recorded or included. At this stage, we have deliberately avoided separating the individual paintings into Mesolithic, Chalcolithic and Historical categories as we are in the process of revising the traditional criteria to make such ‘chrono‐cultural’ separations. In addition, we are in the process of proposing numerous new attributes which need to be recorded at rock art sites in general. JRS‐7 yielded the largest number of paintings (n = 156), followed by JRS‐1 (n = 115) which is the largest stand‐alone shelter in the cluster. Human figures not holding anything (n = 67) appear to be the most common theme in all the rock shelters, followed by human figures with weapon/and/or shield (n = 66) and human figures with bow‐and‐arrows (n = 52); some human figures are as small as 8 cm in length (Fig. 10). A few human figures are remarkably similar to the human figures found in typically stylized form only in the Mahadeo Hills around the Pachmarhi area (e.g. Neumayer, 1993). The least number of paintings come from JRS‐2 (n = 4), JRS‐3 (n = 2) and JRS‐5 (n = 4). Subjects that we were not completely sure about have been noted with a question mark, and some categories have been recognized for the first time or are rarely known from other reported rock art sites. One example is a human figure inside a structure with a triangular roof (Fig. 11) in JRS‐7; similar depictions of a hut have been reported from Lakhajoar (Neumayer, 1993).

Figure 9: A modern bone left by a carnivore or scavenger in JRS‐5

115 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 10: Various types of human figures from the Joshipur rockshelter complex

Figure 11: Structure with a human figure inside (top) and two archers (bottom) in JRS‐7

Two of the most unique paintings are that of a goat‐like creature standing on two legs (Fig. 12) and an ithyphallic figure with a harp? (Fig. 13), both from JRS‐9. For animals which we could not clearly identify, we counted them under the category of unidentifiable animal figure as we did with unidentifiable bird and an unidentifiable figure with possible scales/feather/design. We have also separated specific subjects into dual categories: horse with and without rider; deer with and without horns/antlers; cattle with and without humps; and elephant with and without rider/gear. JRS‐4 is the only shelter to have visibly preserved a possible leopard and JRS‐9 preserves a chariot with a rider and pulled by a draught animal (Fig.14). Other interesting depictions include 18 unidentified felines with 13 of them clustered together in JRS‐7 alone and two clusters of beehives in a tree, in JRS‐4 and JRS‐9 respectively. The painting of the beehive cluster in JRS‐4 is over one meter in horizontal length and shows at least one dozen beehives (Fig.15). Another equally large painting is that of the unidentifiable figure with possible scales/feather/design from JRS‐7 (Fig.16).

116 Chauhan et al. 2015: 105‐127

Figure 12: Painting of an upright ‘goat‐like’ creature superimposed on a wavy line (snake?) in JRS‐9

117 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 13: Painting of a possible harp player in JRS‐9

JRS‐1 is represented by a large block of Vindhyan quartzite forming a very suitable shelter facing northeast and is about eight meters in vertical height and over ten meters on horizontal length (Fig. 17). It is slightly isolated from the other rock shelters and very close to the plains at the foot of the Vindhyan Hills. It also appears to be (thus far) the only shelter in the Joshipur area with a modern Hindu shrine as well as modern English graffiti (Fig. 18). In addition to yielding most of the listed painted subjects, this shelter also shows a double‐humped camel? (or a camel with riding gear?), a crane? and a

118 Chauhan et al. 2015: 105‐127 tiger? (Fig. 19). It also shows a dancing scene and along with JRS‐6 and JRS‐7, this shelter also illustrates a battle scene. The absence of specific subjects and styles at the Joshipur complex (Table 1), but which are generally found at other sites, is also notable: e.g. intricate/geometric patterns, specific types of human figures such as ‘S’ and stick shaped and so forth. It should also be highlighted here that many subjects or scenes in the paintings as interpreted by archaeologists may not be what the original artists intended to portray (e.g. Selvakumar, 2011). The same can be said for this study: our interpretations and descriptions may not accurately reflect the original reasons, motives and perceptions of the ancient painters at Joshipur. Therefore, great caution is required when interpreting and describing rock art, regardless of their age and degree of visual clarity.

Figure 14: Painting of a chariot being pulled by a stylized bovine (?) in JRS‐9

Figure 15: Painting of a cluster of beehives in JRS‐4

119 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 16: Painting showing feathers/scales/design in JRS‐7; other faded paintings on the bottom

120 Chauhan et al. 2015: 105‐127

Figure 17: General view of JRS‐1 at the foot of the Vindhyan Hills

Figure 18: Modern English Writing in JRS‐1

121 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Figure 19: Paintings of various wildlife in JRS‐1: a‐ possible crane; b‐ double humped camel?; c‐ possible tiger

122 Chauhan et al. 2015: 105‐127

Table 1: Documented painted subjects and their frequencies at the Joshipur complex

1 2 3 4 5 6 7 8 9 ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐

SUBJECT No. JRS JRS JRS JRS JRS JRS JRS JRS JRS Single human figures 26 ‐‐ 9 ‐‐25 3 4 67 Human figures with bow‐and‐arrow 7 ‐‐‐‐‐44 1 ‐ 52 Human figures with weapon and/or 18 ‐‐ 1 ‐ 21 26 ‐ ‐ 66 shield? Human figures carrying object(s) on 1 ‐‐ ‐‐ 1 2 ‐ ‐ 4 shoulders (load‐bearer) Human figure inside structure? ‐‐‐‐‐1 ‐ ‐ 1 Battle scene 2 ‐‐ ‐‐ 1 2 ‐ ‐ 5 Ithyphallic figure with harp? ‐‐‐ ‐‐ ‐ ‐ ‐ 1 1 Dancing scene 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Horse without rider ‐‐‐‐1 ‐ ‐ ‐ 1 Horse with rider 7 ‐‐ 2 1 2 5 ‐ ‐ 17 Deer with antlers/ horns (male?) 2 ‐‐ ‐‐ 1 1 1 4 9 Deer without antlers (female?) 1 1 ‐ 1 ‐ 1 1 1 6 Cattle without hump 3 1 ‐‐‐‐‐ 1 8 13 Cattle with double hump 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Double‐humped camel? 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Felines 1 ‐‐ ‐‐ ‐13 ‐ 4 18 Peacock/peahen ‐‐‐1 ‐ ‐ ‐ ‐ ‐ 1 Unidentifiable bird 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Ostrich or ostrich‐like bird 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Crane? 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Elephant 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Tamed elephant (with rider and/or ‐‐‐ 1 ‐ ‐ ‐ ‐ ‐ 1 gear) Tiger? 1 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 1 Leopard? ‐‐‐1 ‐ ‐ ‐ ‐ ‐ 1 Beehives on tree ‐‐‐1 ‐ ‐ ‐ ‐ 1 2 Unidentifiable animal figure 6 1 ‐ 7 ‐‐4 3 16 37 Plant with leaves? ‐‐1 ‐ ‐ ‐ ‐ ‐ 1 Chariot with single rider (draught ‐‐‐‐‐‐‐ ‐ 1 1 animal cattle) Trident (trishul) ‐‐‐‐1 ‐‐ ‐ ‐ 1 Abstract subject/design 1 ‐‐ 1 ‐‐3 ‐ 1 5 Unidentifiable figure with possible ‐‐‐‐‐‐1 ‐ ‐ 1 scales/ feather/design Erect “goat‐like” creature (mythical?) ‐‐‐‐‐‐‐ ‐ 1 1 Unidentifiable/faded patches/cluster/ 24 1 2 9 2 8 28 3 2 79 paintings English graffiti 5 ‐‐ ‐‐ ‐ ‐ ‐ ‐ 5 Hindi graffiti 3 ‐‐ ‐‐ ‐ ‐ ‐ 2 5 TOTAL 115 4 2 35 4 36 156 12 46 410

123 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Discussion and Conclusions This work is preliminary and qualifying rock shelters require further investigations through test‐pits, lithic and faunal specimen plotting and collecting as well as geochronological applications. The study of other important attributes of the paintings (e.g. dimensions of each painting; color and style of each painting; discrete location of each painting; direction in which the shelter is facing) are still being studied and compiled, and will be published through more detailed papers representing the Joshipur rock shelter complex. Collected lithic assemblages are also under study as well as similar descriptive analyses of other rock shelters in our study area. Each encountered painting and even entire rock walls where paintings are heavily faded require digital scanning and extensive photography and videography. Ideally, entire shelters – where feasible – should be scanned using 3D laser methods, thus preserving the morphology of the shelters and the precise locations of the paintings. Future comprehensive surveys using high‐resolution documentation methods may reveal subtle nuances in regards to patterns in spatial overlap of certain types of paintings as well as their geographic distribution patterns across the Vindhyan Hills. The areas between the rock shelters at Joshipur also need to be surveyed for lithic scatters and faunal material. Despite the fact that there was down slope movement of archaeological specimens, their presence, and especially discrete concentrations, can reveal key information regarding the use of the immediate landscape and movement or transport of material between the rock shelters. Plotting and geological studies of such occurrences, i.e. off sites and non‐sites, have rarely been carried out on microlithic sites and can reveal important new information regarding pre‐ and post‐depositional processes. The age of the paintings and the spatially associated lithic scatters is not clear at this stage, although both are presumed to be broadly contemporaneous and thought to belong to the Mesolithic period. The South Asian Mesolithic period is difficult to define as the oldest microliths (45 ka) are now known to overlap with ‘Upper Paleolithic’(currently ambiguous) assemblages and the youngest microliths were still being made in the 19th century. Likewise, the rock art also probably has a similar age range with the oldest paintings being either completely faded and the younger paintings being made by extant tribal groups. The younger paintings appear to have been created by various tribal groups and in some regions, painting designs and specific motifs has continued as seasonal or annual rituals at specific locations. Although high‐resolution studies are lacking in relation to seasonal prehistoric movements across the Vindhyan landscape, changing accessibility to both water and rock resources must have played an important role. For example, our surveys during the summer revealed almost complete absence of water at many rock shelter and open‐ air sites, a situation which is probably opposite during the peak and post monsoon months. One exception is the type of rock shelter sites which preserve water all year round in shallow depressions or semi‐deep pools which have formed through weathering or fluvial erosion, i.e. bases of seasonal waterfalls. It is highly likely that intensive occupation within dense rock shelter complexes led to various behavioral innovations through the sharing of ideas, resource locations and socio‐cultural

124 Chauhan et al. 2015: 105‐127 interactions. Multiple groups and families probably shared the landscape at Joshipur and interacted with groups from neighboring regions.

Acknowledgements We thank the Archaeological Survey of India for the permits to carry out this work and the Director, IISER Mohali for approving the research funds. We thank Prateek Chakraborty for assistance in the field and Shaik Saleem for suggestions in interpreting some of the individual figures. Finally, useful comments by the reviewer, Urmi Ghosh Biswas, are greatly appreciated and have been taken into account during the revision of the original draft.

References Aubert, M., Brumm, A., Ramli, M., Sutikna, T., Saptomo, E.W., Hakim, B., Morwood, M.J., van der Bergh, G.D., Kinsley, L. and A. Dosseto. 2014. Pleistocene Cave Art from Sulawesi, Indonesia. Nature 514: 223–227. Bednarik, R.G. 2013. Lower Palaeolithic Rock Art of India and its Global Context. Indian Journal of Physical Anthropology and Human Genetics 32(1): 113‐142. Bednarik, R.G., Kumar, G., Watchman, A. and Robert, G. 2005. Preliminary Results of EIP Project, Rock Art Research 22 (2). Chauhan, P.R. 2008. Large Mammal Fossil Occurrences and Associated Archaeological Evidence in Pleistocene Contexts of Peninsular India and Sri Lanka. Quaternary International 192(1):21‐42. Chauhan, P.R., S. Ozarkar, S. Kulkarni. 2015. Genes, Stone Tools and Modern Human Dispersals in Center of the Old World. In: Proceedings of the Symposium on the Emergence and Diversity of Modern Human Behavior in Palaeolithic Asia. November 29‐December 1; Tokyo, Japan. Kaifu Y, and Goebel T, editors. Texas A & M University Press. Pp. 94‐113. de Terra, H. and T.T. Paterson. 1939. Studies on the Ice Age in India and Associated Human Cultures. Washington: Carnegie Institute 493. Dennell, R.W. 2009. The Palaeolithic settlement of Asia. Cambridge: Cambridge University Press. Dennell, R.W. Petraglia, M.D. 2012. The Dispersal of Homo Sapiens across Southern Asia: How Early, How often, How Complex? Quaternary Science Reviews 47: 15‐22. Groucutt, H.S. Petraglia, M.D. Bailey, G., Scerri, E.M.L., Parton, A., Clark‐Balzan, L., Jennings, R.P., Lewis, L., Blinkhorn, J., Drake, N.A., Breeze, P.S., Inglis, R.H., Devès, M.H., Meredith‐Williams, M., Boivin, N., Thomas, M. G and A Scally. 2015. Rethinking the Dispersal of Homo Sapiens out of Africa. Evolutionary Anthropology 24: 149‐164. Indian Archaeology – A Review. 1974‐75. Indian Archaeology – A Review. 1975‐76. James, H.V.A. and M.D. Petraglia.2005. Modern Human Origins and the Evolution of

125 ISSN 2347 – 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 3: 2015

Behavior in the Later Pleistocene Record of South Asia. Current Anthropology 46: S3‐S27. Joshi, R.V. 1978. Stone Age cultures of Central India. Poona: Deccan College Post‐ graduate Research Institute. Kennedy, K.A.R. 2000. God‐apes and Fossil‐men: Paleoanthropology of South Asia. Ann Arbor: University of Michigan Press. Kumar, G. 1996. Daraki‐Chattan: A Palaeolithic Cupule Site in India, Rock Art Research 13 (1): 38‐46. Mathpal, Y. 1992. Rock Art Studies in India, in Rock Art in the Old World M. Lorblanchet (Ed.). 1992. Indira Gandhi National Centre for Arts, New Delhi, Pp. 207‐14. Mishra, S., Chauhan, N and A.K. Singhvi. 2013. Continuity of Microblade Technology in the Indian Subcontinent since 45 ka: Implications for the Dispersal of Modern Humans. PlosOne8(7). Misra, V.N. 1985. The Acheulian Succession at Bhimbetka. In: V.N. Misra and P. Bellwood (ed.) Recent Advances in Indo‐Pacific Prehistory, Oxford and IBH Publishing Co., Delhi. Pp. 35‐47. Neumayer, E. 1983. Prehistoric Indian Rock Paintings. Oxford University Press, New Delhi. Neumayer, E. 1993. Lines on Stone: The Prehistoric Rock Art of India. Manohar Publishers, New Delhi. Pandey, S.K. 1993. Indian Rock Art. Delhi: Aryan Books International. Pappu, S., Y. Gunnell, K. Akhilesh, R. Braucher, M. Taieb, F. Demory, and N. Thouveny. 2011. Early Pleistocene Presence of Acheulian Hominins in South India. Science 331:1596‐1600. Patnaik, R., Chauhan, P.R., Rao, M.R., Blackwell, B.A.B., Skinner, A.R., Sahni, A., Chauhan, M.S. Khan, H.S. 2009. New Geochronological, Palaeo climatological and Palaeolithic data from the Narmada Valley Hominin locality, central India. Journal of Human Evolution 56: 114‐133. Rao, K.V., Chakraborti, S., Rao, K.J., Ramani, M.S.V., Marathe, S.D., Borkar, B.T., 1997. Magnetostratigraphy of the Quaternary Fluvial Sediments and Tephra of Narmada Valley, Central India. In: Udhoji, S.G., Tiwari, M.P. (Eds.), Quaternary Geology of the Narmada Valley‐A Multidisciplinary Approach. Special Publication No. 46 of the Geological Survey of India. Geological Survey of India, Calcutta, pp. 65‐78. Selvakumar, V. 2011. Rock art as Cognitive‐Functional Representations: Methodological Issues in the Study of Rock Art in India. Man in India 91(2): 363‐376. Shaik, S. and P.R. Chauhan. 2013. Bandarjari: A New Rock Art Site near Hoshangabad, Madhya Pradesh. Heritage & Us 2 (2): 32‐37. Simte, L.L. 2015. Rocks, relics and paths: Tracing places in the Early Historic landscapes of the southern Vindhyas. Unpublished Ph.D. Thesis. Jawaharlal Nehru University. New Delhi.

126 Chauhan et al. 2015: 105‐127

Tacon, P.S.C., Boivin, N, Petraglia, M., Blinkhorn, J., Chivas, A., Roberts, R.G., Fink, D., Higham, T., Ditchfield, P., Korisettar, R., J‐x Zhao. 2013. Mid‐Holocene Age obtained for Nested Diamond Pattern Petroglyph in the Billasurgam Cave complex, Kurnool District, Southern India. Journal of Archaeological Science 40(4): 1787‐1796. Tiwari, M.P. and H.Y. Bhai. 1997. Quaternary Stratigraphy of the Narmada Valley, in S.G. Udhoji & M.P. Tiwari (ed.), Quaternary Geology of the Narmada Valley: a multidisciplinary approach (Special Publication 46 of the Geological Survey of India): 33‐63. Calcutta: Geological Survey of India. Udhoji, S.G. and M.P. Tiwari (Eds.). 1997. Quaternary Geology of the Narmada Valley: a multidisciplinary approach (Special Publication 46 of the Geological Survey of India). Calcutta: Geological Survey of India.

127