Inscribed Unicorn Seals from Bagasra, Gujarat: A Comparative Analysis of Morphology, Carving Styles, and Distribution Patterns

Inscribed Unicorn Seals from Bagasra, Gujarat: A Comparative Analysis of Morphology, Carving Styles, and Distribution Patterns Gregg M. Jamison 1, Bhanu Prakash Sharma 2, P. Ajithprasad 3, K. Krishnan 3, Kuldeep K. Bhan 3 and V. H. Sonawane 3 1. Department of Anthropology, University of Wisconsin Waukesha, 1500 N. University Dr., Waukesha, WI 53188, USA (Email: gregg.jamison@gmail.com) 2. Department of Archaeology, University of Kerala, Kariavattom Campus, Thiruvananthapuram, Kerala 695 581, India (Email: bhanu1927@gmail.com) 3. Department of Archaeology and Ancient History, The Maharaja Sayajirao University of Baroda, Vadodara, Gujarat 390 002, India (Email: ajit.karolil@gmail.com; Krishnan.msu@gmail.com; Bhankuldeep007@sify.com; vhsonawane@ rediffmail.com) Received: 18 July 2017; Revised: 13 September 2017; Accepted: 01 November 2017 Heritage: Journal of Multidisciplinary Studies in Archaeology 5 (2017): 01 21 Abstract: Excavations at the Indus site of Bagasra (Gola Dhoro) conducted by the Department of Archaeology, The Maharaja Sayajirao University of Baroda uncovered significant quantities and varieties of material culture, including inscribed steatite seals. This paper presents a comprehensive analysis of this assemblage, including morphology, carving styles, and distribution patterns. Using complimentary research methods, it has been possible to identify variation in seal carving styles and techniques, as well as patterns that represent the products of distinct artisans and workshops. Taken together, the results highlight the significance of detailed studies of seals from individual sites to learn more about this important craft industry throughout the Indus Civilization as a whole. Keywords: Indus Civilization, Inscribed Steatite Seals, Bagasra, Gujarat, Technology, Style, Distribution Introduction Since their discovery heralded the announcement of a new ancient civilization, inscribed steatite seals of the Indus or Harappan Civilization (2600 1900 BCE) have captivated scholars and the general public alike. Today we understand that these objects were very important during the Indus period, serving multiple purposes and recording and transmitting culturally significant information. They have been found at many Indus sites and are now often used to identify its Harappan character. While a considerable amount of energy has been expended studying Indus seals over the last century, very few detailed analyses of seals from individual sites have been

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 undertaken. This research attempts to address this problem through an examination of the seals recovered from the site of Bagasra (Gola Dhoro) in Gujarat. Analyses conducted on seven seals engraved with the unicorn animal motif have yielded new insights into the distribution, metric attributes, and carving styles of these artifacts. Using basic descriptive and spatial analyses, we have determined that the sample of seals from Bagasra is actually quite variable in terms of metric properties and spatial distributions at the site. Formal analyses of unicorn iconography have revealed variability and the identification of a pair of seals that appear to represent a unique carving style. While the limited scope of the current study precludes robust interpretations about the relationships between seals recovered at Bagasra and other sites, the results do suggest that future research focusing on comparative analyses can be used to identify inter site patterns among seals recovered at various sites in Gujarat and beyond. These can be utilized fruitfully to investigate the nature and scale of variation in the corpus of Indus seals. Such studies will advance our awareness of Indus seal production and ultimately one of the world s earliest state level societies. Figure 1: Major Indus Sites and Interaction Networks (Courtesy: J. M. Kenoyer) The Indus Civilization The ancient Indus or Harappan Civilization represents the earliest manifestation of urbanism in South Asia (Kenoyer 1998), and today its material remains are dispersed over a vast area located primarily in modern Pakistan and northwestern India (Figure 1). Our understanding of the Indus has been informed almost exclusively through 2

Jamison et al. 2017: 01 21 archaeological investigations, today more than 1500 sites are known, distributed over one and a half million square kilometers (Kenoyer 1998). The remains of the Harappan Civilization include cities and villages, craft centers, river stations, campsites, fortified palaces, and probable ports. The Indus Tradition (Kenoyer 2008; Shaffer 1992; Table 1) is one of the chronological frameworks used to document the history of cultural adaptations in the region where the Harappan civilization flourished. While the sophisticated urbanism that characterizes the Indus Civilization reached its apex during the Integration Era of the Harappa Phase (2600 1900 BCE), the processes responsible for these developments have their antecedents during the Early Food Producing and Regionalization Eras. In addition to large, well planned urban centers, the Integration Era is also characterized by material homogeneity, although recent studies indicate that regional variation is present in many forms of material culture (Kenoyer 2008). Most of the published corpus of Harappan seals with known stratigraphic contexts has been dated to the Integration Era, and therefore it is the most important chronological period for this research. Table 1: Chronology of the Indus Tradition (after Kenoyer 2008; Shaffer 1992) Era Phases Time Period Foraging Era Mesolithic and Microlithic 10000 2000 BCE Early Food Producing Era Mehrgarh 7000 5500 BCE Regionalization Era Early Harappan Phases: Ravi, Hakra, Balakot, Amri, Sheri Khan Tarakai, Kot Diji, Sothi 5500 2600 BCE Integration Era Localization Era Harappan Phase: Harappa Site Period 3A Harappa Site Period 3B Harappa Site Period 3C Late Harappan Phases: Punjab, Jhukar, Rangpur 2600 1900 BCE 2600 2450 BCE 2450 2200 BCE 2200 1900 BCE 1900 1300 BCE Kenoyer (2000) has argued that the Indus cities were ruled by competing groups of elites, likely composed of merchants, landowners, and ritual specialists. Some of these elites would have been literate, using writing on pottery and inscribed seals to record economic transactions, important rituals, and communicate over great distances. This form of sociopolitical organization can be characterized by a city state or decentralized state model. These competing groups of elites, as well as the majority of the rest of society, were likely integrated at multiple levels, including economic interaction and a poorly understood ideological system (Kenoyer 2000). It is clear that the Indus represents a complex, urban civilization that is unique in comparison with contemporary polities in the Old World. The similarities in material culture and urban planning suggest the presence of a complex administrative body that would have integrated major urban centers and smaller settlements alike (Kenoyer 2008). This 3

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 integration would have been accomplished in part through the production and use of numerous craft items, including inscribed steatite seals. Figure 2: Harappan Stamp Seal with Unicorn Motif (Courtesy: www.harappa.com) Inscribed Steatite Seals Seals represent one of the most important forms of inscribed objects from the Indus Civilization (Kenoyer 1998; Rissman 1989). While there is a considerable amount of variation in the types and sizes of inscribed seals from the Indus Civilization, the most common are often referred to as stamp seals (Mackay 1931) because they were used to stamp wet clay, and this interpretation has been supported by the discovery of seal impressions and sealings. Within the category of Harappan stamp seals there are two principal types, square seals that usually depict a line or two of script with an accompanying animal motif (Figure 2), and rectangular seals or tablets that often depict only script (Rissman 1989). Square seals engraved with the unicorn motif are 4

Jamison et al. 2017: 01 21 the most common type found at Harappan sites, including Bagasra, and therefore will be the primary focus of this study. It is referred to as a unicorn because it appears to represent a single horned bovid, and the fact that it is represented in other forms of material culture including terracotta figurines suggests that it was meant to portray an animal that is distinct from twohorned animals that appear on other seals (Kenoyer 2010). In addition to the body and horn it is usually engraved with a number of other details, some of which are very small and finely carved, including the eye, ear, legs and hooves, tail, pizzle (penis), and standard, or ritual stand. Additional decorations include incising on the face and neck, halter (shoulder), and standard. Previous studies (e.g. Franke Vogt 1992; Rissman 1989) suggest that the ways in which these are carved vary considerably, an observation supported by the results of this study. Another important feature of unicorn seals is the presence of a short inscription, usually found at the top and consisting of a series of characters representing the Harappan script that has yet to be deciphered. The uses and function of the script engraved on seals would have been diverse and included recording various economic transactions, administrative activities, important rituals, and personal identification and ownership (Kenoyer 1998, 2006; Meadow and Kenoyer 2000). Traditionally they have been interpreted as static, uniformly produced and consumed administrative artifacts (Mackay 1931; Rao 1985; Vats 1940), but recent investigations at Harappa have uncovered changes in the styles of seals over time, evidenced through controlled stratigraphic excavations (Kenoyer 2004; 2006). These discoveries indicate that the use of seals and writing at Harappan sites was not static and unchanging over a period of 700 years. Moreover, discoveries from the Early Harappan or Kot Dijian Phase (2800 2600 BC) levels at Harappa have also uncovered evidence of an early writing system and the production of seals during this phase, indicating that the technologies of both seal production and writing have antecedents before the Integration Era (Kenoyer 2006). In light of these recent discoveries the need to re examine seals from multiple sites becomes clear, and the site of Gola Dhoro provides an excellent starting point to do so. Bagasra (Gola Dhoro) and the Indus Civilization The site is located about half a kilometer southeast of the village of Bagasra, Maliya Taluka of Rajkot District, Gujarat (Sonawane et al. 2003). It was discovered in the 1980s during an exploration conducted by personnel from Deccan College, Pune, and the Gujarat State Archaeology Department. Excavations were undertaken by the Department of Archaeology and Ancient History, The Maharaja Sayajirao University of Baroda in 1995 96 (Sonawane et al. 2003). Investigations carried out over nine seasons from 1996 2004 have revealed a significant amount of information regarding the size, layout, and material assemblage of this important regional site. The remains of the site rise approximately 7.5 meters above the surrounding plain and measure 160 x 120 meters in area (Bhan et al. 2005). It is one of several regional Indus sites located near 5

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 the Gulf of Kutch in Gujarat (Figure 3). Excavations have yielded evidence of four distinct cultural sequences affiliated with the Harappan Period in Gujarat (Table 2), comprising Classical and Sorath Harappan occupational debris, as well as Anarta materials (Sonawane et al. 2003). 6 Figure 3: Chalcolithic and Harappan Sites in Gujarat Salient features of the site include a massive mud brick fortification wall, numerous structural remains, and several spatially distinct craft activity areas (Figure 4). In addition to ceramics with Classical Harappan, Sorath Harappan, and Anarta Tradition affiliations, other recovered materials include stone beads, weights, tools, and debitage fashioned from a variety of raw materials (Bhan et al. 2005). Bagasra has yielded large quantities of shell objects and manufacturing debris, indicating that it was a production center. Faience production has also been attested at the site through the recovery of numerous finished and unfinished beads and a few bangles as well as manufacturing debris (Sonawane et al. 2003). Recent studies also indicate that copper production may have taken place here, albeit outside the fortified area of the site (Patel 2005). It is important to note that most of these craft activity areas have been identified within the fortification wall at the site, suggesting that the production of some goods within the settlement may have been restricted (Sonawane et al. 2003). Additionally, a wealth of other Harappan materials have been uncovered at the site, including copper tools and implements, terracotta objects and animal figurines, and important administrative materials including seals and sealings (Bhan et al. 2005).

Jamison et al. 2017: 01 21 Table 2: Chronological Sequence of Gola Dhoro (after Sonawane et al. 2003) Phase Diagnostic Association Cultural Affiliation I Early Urban Pre fortification Classical Harappan and Anarta II Urban Fortification Classical Harappan and Anarta III Late Urban Sorath Harappan (Rangpur IIA and IIB) Classical Harappan, Anarta and Sorath Harappan IV Post Urban Sorath Harappan (Rangpur IIC) Post Urban Sorath Harappan Figure 4: Plan View of Bagasra (Courtesy: B. A. Chase; M.S. University of Baroda) While the latter have not been recovered in large quantities, their presence suggests that some level of administrative and or political authority exercised power at the site. This interpretation is corroborated by the presence of a massive fortification wall and spatially segregated indicators of craft production, which previous studies have demonstrated can also represent administrative authority and control (Bhan et al. 2002; 7

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 Kenoyer 1989, 1992, 2000; Kenoyer et al. 1991, 1994). It is beyond the scope of the current study to thoroughly address the nature of socio political organization and control at the site, rather our aim here is to document and describe the unicorn seals found at Bagasra. We hope that in doing so we can contribute positively to our growing understanding of regionalism during the Indus Period in Gujarat. Methodology In this paper we are concerned with characterizing the basic metric attributes and stylistic properties of the unicorn seals and determining if there are any patterns or relationships among their spatial distributions. The first step in this research was to carefully measure and weigh the seals from Bagasra, using a standard digital caliper and digital scale. A total of seven seals were studied, and this total represents the full corpus of steatite stamp seals from the site. At least six are engraved with unicorn motif, the remaining seal is a fragment that contains only the standard, however this element is usually associated with the unicorn motif. Basic measurements including length, width, and thickness were recorded for each seal. Next we conducted a formal analysis of the carving styles and decorative techniques for selected attributes of the unicorn and standard, recording our observations for comparative purposes. After completing this we proceeded to record find spots within the site, including all available contextual information (Figure 4). Upon completion of recording these data for each of the seals a preliminary analysis was conducted to describe the basic descriptive statistics of the sample, their spatial distributions, and to search for patterns in carving styles and techniques. Results The results of this preliminary analysis will be described in two sections, one describing of each seal individually and another describing the sample as a whole. What follows is a brief discussion of the seven seals individually. BSR 2037 This object was uncovered in Trench El 15 during the 1998 field season. It is fairly small (approximately 1.8 cm square), and less than 1 cm thick. The boss has been drilled in a linear fashion rather than diagonally into the body of the seal. The seal is white in color and in addition to the engraved unicorn motif it also contains an inscription bearing two distinct characters of the Indus script (Figure 5). The inscription is linear in orientation and both characters are broadly carved and clearly discernable. No pizzle is visible at the base of the engraved body and facial details including mouth and nose are obscured or absent. The engraved surface has been damaged in several places; cracking is clearly visible near the head, horn, and rear legs. Both the neck and the body have been decorated with several engraved lines, as has the upper element of the standard in front of the animal. The latter feature can be described as being short and broad, both elements (upper and lower) are longer than they are wide in terms of maximum measurements. Overall the carvings are broad 8

Jamison et al. 2017: 01 21 and deep, and as a result both the inscription and animal motif are clearly visible. As a result of the carving style, however, finer details including the aforementioned facial features are not easy to distinguish. BSR 5555 A seal fragment was recovered from Trench Eg 2 during the 2003 field season and recorded as BSR 5555. The only iconographic element preserved intact is the standard, however a portion of the animal s lower neck and body can also be observed near the edge of the fractured surface (Figure 6). What little remains of the engraved body bears strong similarities to other unicorn seals, specifically the engraved lines often carved on the neck of the unicorn motif. Moreover, the engraved standard that constitutes most of the fragment is usually associated with this type of seal, with a series of vertical and horizontal lines that are commonly found on unicorn seals. Therefore, for the purposes of analysis we have included it within the subset of unicorn seals recovered from the site. Due to its fragmentary nature, it is difficult to glean much information about this object. The fragment measures approximately 1.4 by 0.8 cm and is 0.75 cm thick. It is yellowish gray in color and weighs only 1.6 grams. The stand itself is well carved and symmetrical, with both upper and lower elements measuring approximately 0.5 cm in length. Furthermore, both elements have been hatched or engraved with fine lines, and these can be seen running horizontally and vertically across the carved surface of the standard. The precision and detail visible in these features suggests that they may have been engraved by a skilled carver using a fine edged tool. Figure 5: BSR 2037 Figure 6: BSR 5555 (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) 9

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 Figure 7: BSR 6719 Figure 8: BSR 6952 (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) Figure 9: Interior Section of BSR 6952 Figure 10: BSR 7197 (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) BSR 6719 During the 2004 season this seal was recovered from Trench Eo 11. It is yellowish gray in color and measures about 2.6 cm square. It is almost 1.5 cm thick including the boss and 0.78 cm thick without it, with a weight of 13.6 grams. The boss is drilled in a linear direction, meaning that it is not drilled into the body of the seal. The inscription at the top of the engraved face consists of seven distinct characters, and although two of them have been partially obscured due to cracking, the others are well carved and clearly visible (Figure 7). The orientation of the inscription is roughly linear, however several of the characters on the left edge appear to be cramped due to the presence of the horn. The seal is fully intact and well preserved, but in addition to the damaged inscribed characters, cracks in the lower left and right hand corners have obscured the details of the standard and tail, respectively. The artisan(s) who crafted this object attempted to provide the face 10

Jamison et al. 2017: 01 21 with distinct features including an eye and nose, which are still present in the engraved surface. The pizzle is present on this sample, which, along with the finely carved legs, suggests that the carvers were concerned with attention to detail. The body is accented with four halter (shoulder) lines in a pattern that is seen elsewhere in the corpus of Indus seals (Franke Vogt 1992). The standard is hatched, or decorated with small incised lines, in the upper register, but these are not regularly spaced or symmetrical and appear to have been carved with some difficulty. Nevertheless this seal is one of the finest made and preserved from the site. BSR 6952 A second seal fragment was recovered during the 2004 field season from the western baulk of Trench Ea 12. It consists of the lower half of the body of an animal that appears to have been a unicorn, but the upper half of the body, including the neck, head, horn, and ear; as well as the full inscription, have been lost (Figure 8). The remaining fragment measures 3.45 by 2.38 cm, the former value representing the maximum width of the seal. In this respect it is the largest of the seals recovered at Gola Dhoro, and therefore not surprisingly the heaviest, weighing 15 grams. Fortunately the boss has also been preserved, and including this feature the seal is approximately 1.5 cm thick. At the risk of making value judgments about the quality of engravings, at first glance one is struck by how poorly carved this seal is in relation to others from this site and beyond. Although the full motif is not present, the remaining elements appear to be irregular in spacing and orientation, not carved deeply, and disproportionate in relation to each other. This observation applies to the visibly extent portion of the inscription as well. The standard in particular has been poorly executed, and it is unclear whether the cracks visible in the animal s rear flank have been caused by time or erroneous carving techniques. One of the most interesting and important features of this seal is the cross section visible along the fractured surface (Figure 9). While the exterior of the seal is yellowish gray in color, the interior is clearly of a different hue, in this case a grayish green. These differences are important in that they can provide information regarding technological aspects of seal production. Specifically, this sample suggests that either the surface of the seal was treated in order to whiten it during firing or that it was not heated thoroughly enough to transform the color of the steatite entirely. The potential for more detailed studies focusing on these problems will be addressed in future research. BSR 7197 This seal was also discovered during the 2004 season in Trench Eh 1. It measures 2.07 by 2.23 cm, meaning that it is slightly rectangular in shape. Maximum thickness including the boss is approximately 1 cm; without it, it is 0.56 cm. The seal is pale orange in color and weighs 6 grams. In addition to the animal motif and associated standard there is an inscription consisting of nine characters, the longest in the sample 11

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 of seals from Gola Dhoro. The orientation of the inscription is linear, and each character is engraved clearly and in a manner that allows easy identification. The body of the animal is engraved with nine lines, most of which are found on the neck. It is probable that these lines would have been carved with a fine edged tool, based on the fact that they are found within the confined area of the neck and also by the fact that they are narrow and deeply incised. The appearance of additional lines on the body (Figure 10) can actually be explained by the fact that the body has not been engraved to a uniform depth. The eye is large and carved in a distinct style, and the orientation of the horn appears to have been influenced by the inscription, suggesting that the latter may have been carved first. The pizzle is present, and the legs also appear to have been engraved with care. The standard is short and broad and hatched in both registers; however, these incisions have been damaged by cracking. Overall, the seal is well preserved and well made; it is also unique and distinct among the members of this sample, for reasons to be discussed in greater detail below. BSR 7368 Yet another seal was recovered during the 2004 investigations in Trench Eh 1. It is white and yellowish gray in color, and is the smallest of the intact seals from this assemblage, measuring about 1.4 cm square. It is less than 1 cm thick including the boss; the seal body itself has a thickness of 0.7 cm. It weighs only 2.3 grams, by far the lightest complete seal in the sample as well. The animal motif and standard are accompanied by an inscription consisting of five characters (Figure 11) and this inscription contains the only repetitive signs found in the seals from Gola Dhoro. The inscription is linear, the characters are regularly spaced and easy to discern, however two of them are damaged from a fracture in the upper right hand corner of the seal. Additional cracking has obscured the standard in the lower right hand corner of the engraved surface and the base of the rear legs and tail. The carver attempted to fashion facial details including an eye and nose, but these are broadly carved and therefore appear to be disproportionately large in relation to the rest of the head. The same can be said of the ear. The unicorn is decorated with six engraved lines, 4 or 5 of which are found on the neck, the remaining at the base of the animal s front flank. The dibble is present, as are fine details on the legs and hooves. Considering the small size of the seal the artisan(s) who crafted it have done a faithful job of representing most of the typical elements of unicorn iconography. The most impressive feature of this seal is the presence of a small compartment or chamber that extends vertically down the middle of the seal body (Figure 12). This would have originally been enclosed with a removable lid of some kind, which unfortunately has been lost. Seals of this type are extremely rare in the published corpus, even at the larger urban centers, which generally speaking display the greatest range of variation as regards morphology and style, so this discovery in and of itself is very important. Of further significance is the fact that all three sides of the seal are hatched, decorated with differing patterns of incised lines (Figure 13). To our knowledge no other seals like this have been found at any Indus sites, which raises further questions about the 12

Jamison et al. 2017: 01 21 importance and significance of the site of Gola Dhoro. A final note on the composition of the raw material itself is also in order here. Unlike most other seals with exposed interior sections, BSR 7368 is comparatively homogenous in terms of color and texture. It is also extremely light in weight (Appendix 1) at only 2.3 grams, indicating that it may be made of a material other than carved steatite. Of course this evaluation will require further testing, but ultimately may lead to new discoveries about the technologies involved in seal production. At any rate this seal is undoubtedly a unique and incredibly important discovery, one that warrants further consideration. Figure 11: BSR 7368 Figure 12: BSR 7368 Internal Compartment (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) Figure 13: BSR 7368 Hatched Faces Figure 14: BSR 8288 (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) 13

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 BSR 8288 This seal was found in Trench Ei 14 during the 2004 season. It measures 2.53 by 2.37 cm, and 0.72 cm in thickness without the boss, and about 1 cm with it. The weight is 7.2 grams and it is yellowish gray in color. The inscription at the top of the engraved surface contains six characters, all of which are unique and only appear once. Although the last character is damaged in the upper right hand corner, the others are evenly spaced and clearly distinguished from each other. The orientation of the inscription is roughly linear, albeit slightly slanting to the right (Figure 14). Additional cracking can be observed along the base of the engraved face, specifically where the front legs have been engraved, but beyond this the seal is well preserved. The body is engraved with eleven or twelve lines, most of which are concentrated on the neck. Lines visible in the rear flank may be due to weathering or created by multiple cutting strokes with the tool(s) used to carve the seal. The animal motif has not been engraved to a uniform depth; the head is engraved deeper than the neck, which is in turn deeper than the body. Facial details are obscured and difficult to distinguish without the aid of magnification, and the stand is short and broad, although it is clearly hatched. Proportionately, the legs are short and the pizzle is long, but both have been carved completely and with some level of detail. A prominent incision at the base of the front flank represents the boundary of the uneven engraved surface, with the flank being deeper than the rest of the body. Comparative Analysis and Descriptive Statistics One of the goals of this study was to determine whether there are any patterns in this dataset that may relate to uniformities or standardization in the production of seals, particularly in terms of metric attributes. In order to address this issue selected measurements were recorded (Appendix 1) and analyzed to identify patterns. Summary statistics for the whole sample were also generated and examined, and these tests were supplemented with a visual inspection of all of the seals to ascertain and evaluate whether any stylistic affinities were present. It is clear from the data that there is marked variation in all of the attribute classes under consideration here. As regards the seals, there are no strong correlations in terms of basic measurements including length, width, and thickness. Instead, a consistent range of variation is present, in other words the seals come in a variety of different sizes. Seal thickness is the least variable category, with slightly more than 0.2 cm in variation. While most of the seals in the sample are around 0.7 cm in thickness, two of them are less than 0.6 cm and none of them have the same value in this category. Seal length and width are much more variable in that none of the seals share the same dimensions, and several of them are not even square. Overall, there are few patterns visible in the data regarding seal length, width, and thickness. The same can be said of the metric attributes of the boss found on the reverse, unengraved face of the seals. Stylistically, there appear to be some similarities in this category, but statistically there are no strong patterns present. The six seals with intact 14

Jamison et al. 2017: 01 21 bosses demonstrate a range of variation of at least 0.5 cm in terms of maximum length and width, and almost as much (0.48 cm) in thickness. Three of the seals have a maximum boss length of around 1.1 cm and a width between 0.9 1.0 cm; however, there is no relationship between the dimensions of the boss and the seal as a whole. In addition, the other three seals with intact bosses have quite variable length and width measurements, and equally variable proportions among them. The thickness of the boss is also varied; although two of the seals measure 0.3 cm in this category the other four range from 0.2 0.7 cm. The one exception to this can be witnessed in the category of boss drill hole diameter, where five of the six samples have values between 0.27 0.33 cm, a range of only 0.06 cm. The one outlier, BSR 7368, is the smallest seal in the sample in terms of overall length, width, and thickness, and therefore it is not surprising that the hole drilled through the boss is smaller. The restricted variation observed in this category is noteworthy, and provides an excellent direction for future studies. Appendix 1: Absolute Values of Selected Attributes of BSR Seals Attribute/Seal Number 2037 7197 5555 6952 7368 6719 8288 Seal Maximum Length 1.87 2.07 1.37 2.38 1.39 2.62 2.37 (Cm) Seal Maximum Width 1.88 2.23.84 3.44 1.44 2.66 2.53 (Cm) Seal Maximum Thickness.57.56.75.76.7.78.72 (W/Out Boss) (Cm) Seal Weight (G) 3.6 6 1.6 15 2.3 13.6 7.2 Seal Color (Munsell) N9 10YR8/2 5Y8/1 5Y8/1w/ 5G5/2 N9w/ 5Y8/1 5Y8/1 5Y8/1 Boss Maximum Length (Cm) 1.18 1.11 n/a 1.5.8 1.3 1.16 Boss Maximum Width.93.9 n/a 1.11.83 1.36.99 (Cm) Boss Maximum Thickness (Cm).33.45 n/a.74.26.67.31 Boss Hole Maximum.27.3 n/a.3.14.28.33 Diameter (Cm) Seal Maximum Engraved 1.7 1.93 n/a n/a 1.2 2.46 2.21 Length (Cm) Seal Maximum Engraved 1.69 2.03 n/a n/a 1.35 2.51 2.41 Width (Cm) Seal Maximum Engraved.09.10.08.05.06.11.10 Depth (Cm) Trench El 15 Eh 1 Eg 2 Ea 12 Eh 1 El 15 El 14 Layer 3 1 10 5 Pit 1 n/a 4 Depth (Cm Below Datum) 100 26 300 70 75 223 66 105 15

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 Analyses of the stylistic properties of the unicorn motif in the Bagasra assemblage has revealed further evidence of variation. In spite of the small sample size, for the most part the unicorns have been carved with different styles and decoration techniques. This variation can be observed in most of the stylistic attributes studied (Figure 15). Some of the unicorns are engraved with fine details and decorations, others lack specific elements (including the pizzle) that are common throughout the corpus of published unicorn seals. Of particular interest is BSR 7368, which is already unique because of its compartment, but more so because of the hatched lateral faces. To date we have not identified these stylistic treatments on any other unicorn seals from any sites. In contrast, BSR 6952, even though it is fragmentary, depicts less detail and precision in the execution of the pizzle, engraved lines on the neck, halter, and standard than the other seals from the site, suggesting it may have been carved by a less experienced or skilled artisan. The identification of different carving styles of the seals from Bagasra supports other research (Franke Vogt 1992; Jamison 2017; Kenoyer 2009; Rissman 1989) concerning stylistic variability in Indus unicorn iconography. Figure 15: Selected Stylistic Attributes of the Unicorn and Standard (Jamison 2017) 16

Jamison et al. 2017: 01 21 On the other hand, we did uncover some interesting patterns regarding stylistic coherence. Two of the seals (BSR 7197 & BSR 8288) demonstrate strong parallels in the carving styles of multiple attributes, including heads, ears, horns, halters, tails, legs, hooves, pizzles, and both elements of the standards (Figure 16). Several of these are engraved with unique, uncommon carving styles, especially the heads, eyes, pizzles, and standards. The combinations of attribute styles present on both seals are distinct and may represent a local or regional carving style. The similarities in carving styles also suggest the possibility that the same artisans may have been involved in production (Jamison 2017). Future studies on the seals will allow us to determine if there are patterns in raw materials and tool marks that further support or refute these preliminary interpretations. Figure 16: BSR 7197 and BSR 8288 (Courtesy: Department of Archaeology, The Maharaja Sayajirao University of Baroda) Additional stylistic analyses have provided new insights into stylistic affinities among the Bagasra assemblage and unicorn seals from other sites. BSR 6719 has stylistic affinities with seals from Mohenjo daro, and BSR2037 is stylistically similar to a seal from Lothal. In both cases, unique attribute combinations that distinguish them from other examples in the published corpus of unicorn seals. These discoveries have already been thoroughly discussed elsewhere (Jamison 2017) and highlight the utility of formal stylistic analyses to uncover patterned variation in unicorn seal iconography. Such patterns are useful for understanding how seal production was organized and varied, and the methods employed in this study can be applied to materials from multiple sites and regions as new finds are made and published. Context and Spatial Distribution The best way to examine the spatial distribution of the seals from Bagasra is graphically (Figure 4). We can see that the seals have been recovered from different 17

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 areas of the site, inside and outside of the fortification wall. This is interesting because often seals are recovered within fortified areas at other sites, especially Harappa (Kenoyer 1998). While the majority (6 of 7) has been recovered inside of the fortification wall, one has been found outside of it. It is possible that it may have been accidentally discarded, however the fact that it is intact and was found near a probable locus of craft production suggests that it may have been used and discarded right there. Two additional seals have been found near the shell working area and have stratigraphic associations with these activities. Another two were recovered in the passageway of the southern gateway, and although they may have been re deposited through alluvial and or erosional activities (P. Ajithprasad, personal communication) they have still been found within the fortifications and were likely originally deposited there. Furthermore, our current understanding of the site suggests that craft production would have taken place throughout the fortified area, so even though the remaining two seals do not have direct stratigraphic associations with craft production we can infer an association based on their recovery inside rather than outside of the walled area. This interpretation is strengthened by the fact that the one seal found outside of the fortified area was also found in a craft production area. Although the seals were distributed throughout the site, in all cases they were found associated with gateways and or craft activity areas. An examination of the stratigraphic associations of the seals from Bagasra revealed evidence of considerable variation. They were recovered at varying depths, from 3 meters to 26 centimeters below datum, a range of nearly 2.75 meters. BSR 7197 was recovered at 26 centimeters below datum in the gateway, and likely represents a secondary deposit. Two of the seals (BSR 6719 & 6952) were recovered at depths of less than a meter (Figure 4, Appendix 1), while seals BSR 2037 and 8288 were found at approximately one meter. These objects were recovered in two adjacent trenches (El 14 & El 15) near the southeastern corner of the fortification wall, near a locus of faience manufacturing and lithic stockpiling. They provide the strongest evidence of contextual association among the seals at the site, being found at roughly the same depth and in the same area of the site, however, any relationship regarding use and function remains unclear at present. Another seal, BSR 7368, was found at a depth of 2.23 meters below datum in a pit sealed by Layer 1 in Trench Eh 1 (Figure 4), near the southern gateway. The last seal (BSR 5555) was found at a depth of 3 meters, in Layer 10 of Trench Eg 2. While there is a considerable amount of variation in terms of stratigraphic associations among the seals from Bagasra, It is possible that at least two of them may have been related, but further analysis is needed to confirm or refute this. At present we can say that there is a correlation between seals and craft activity areas and gateways at the site, because all of the seals were recovered in or near areas associated with these features. Current evidence suggests that Harappan elites used seals as symbols authority and power and may have been actively involved in craft production to generate wealth (e.g. Kenoyer 2000). These data seem to support this interpretation and further suggest that it occurred even at relatively small, regional sites like Bagasra. 18

Jamison et al. 2017: 01 21 Conclusion Despite the importance of inscribed steatite seals in the Indus Civilization, many questions remain regarding how and where they were made and how production would have varied between and within different sites. The first step in addressing these issues is to conduct detailed analyses of seal collections from individual sites, focusing on multiple aspects of iconography, metric proportions, and spatial and contextual analyses. In this study of seals from the site of Bagasra, we have outlined a method for studying these objects and generated important data regarding the variation present in the assemblage. Overall, the seals are diverse in terms of metric proportions, carving styles, and distributions. This supports current research on Indus unicorn seals and emphasizes the fact that they are not all standardized and uniform in size and style. This is likely due to the fact they were carved by different individuals and or groups, each with its own unique style and tradition. On the other hand, this study has uncovered patterned variation in carving styles that represent the products of distinct artisans and workshops that produced unicorn seals in the Indus. The fact that these patterns are present in a small collection of seals from a single site, and that at least a few of them demonstrate links with other sites and regions, provides directions for future research. Using a multi faceted and detailed research program such as the one described here, it should be possible to gain greater insights into the scale and nature of variation in Indus seals, as well as insights into how production was organized and varied. These techniques can be used to create comparative data sets, facilitating future collaborative and comparative studies. We are confident that this research will increase our knowledge of these essential objects and their role within one of the world s earliest and most important state level societies. Acknowledgments The Department of Archaeology and Ancient History, The Maharaja Sayajirao University of Baroda, Vadodara has provided the support and assistance that we needed to complete this research. We are also grateful to all team members of the Bagasra Excavations. Dr. Brad Chase of Albion College, Michigan, USA, provided us with some of the graphics used in this report, as did Dr. Rajesh SV, Assistant Professor Kerala University, Kariavattom Campus, Thiruvananthapuram. Their assistance is kindly acknowledged as well. The views expressed in this paper and all errors in fact or interpretation are solely our own. References Bhan, K. K., M. Vidale and J. M. Kenoyer. 2002. Some Important Aspects of the Harappan Technological Tradition. In Indian Archaeology in Retrospect, Volume 2 (Protohistory: Archaeology of the Harappan Civilization), edited by S. Settar and R. Korisettar, pp. 223 272. vol. 2. Indian Council of Historical Research, New Delhi. 19

ISSN 2347 5463 Heritage: Journal of Multidisciplinary Studies in Archaeology 5: 2017 Bhan, K. K., V. H. Sonawane, P. Ajithprasad, and S. Pratapchandran. 2005. A Harappan trading and craft production Centre at Gola Dhoro (Bagasra). Antiquity Vol 79 No 304, http://antiquity.ac.uk/projgall/bhan/index.html. Fairservis, W. 1986. Cattle and the Harappan Chiefdoms of the Indus Valley. Expedition 28(2):43 50. Franke Vogt, U. 1992. Inscribed Objects from Mohenjo Daro: Some Remarks on Stylistic Variability and Distribution Patterns. In South Asian Archaeology 1992, edited by C. Jarrige, pp. 103 112. Monographs in World Archaeology No. 14, Prehistory Press, Madison. Jamison, G. M. 2017. The Organization of Indus Unicorn Seal Production: a Diachronic Comparative Study of Style, Skill, and Sociopolitical Organization. PhD Dissertation, University of Wisconsin Madison, Madison, WI, USA. Joshi, J. P., and A. Parpola. 1987. Corpus of Indus Seals and Inscriptions: Collections in India Vol. I. of II Vols. Suomalainen Tiedeakatemia, Helsinki. Kenoyer, J. M. 2009. Indus Seals: an Overview of Iconography and Style. Ancient Sindh 9:7 30. Kenoyer, J. M. 1989. Socio Economic Structures of the Indus Civilization as reflected in Specialized Crafts and the Question of Ritual Segregation. In Old Problems and New Perspectives in the Archaeology of South Asia, edited by J. M. Kenoyer, pp. 183 192. Wisconsin Archaeological Reports. vol. 2. Dept. of Anthropology, U.W. Madison, Madison, WI. Kenoyer, J. M. 1992. Harappan Craft Specialization and the Question of Urban Stratification. Eastern Anthropologist 45(1 2):39 54. Kenoyer, J. M. 1998. Ancient Cities of the Indus Valley Civilization. Oxford university Press, Karachi. Kenoyer, J. M. 2000. Wealth and Socio Economic Hierarchies of the Indus Valley Civilization. In Order, Legitimacy and Wealth in Early States, edited by J. Richards and M. Van Buren, pp. 90 112. Cambridge University Press, Cambridge. Kenoyer, J. M. 2003. Uncovering the Keys to the Lost Indus Cities. Scientific American July 2003:68 75. Kenoyer, J. M. 2004. Steatite and Faience Manufacturing at Harappa: New Evidence from Mound E Excavations 2000 2001. Journal of the National Museum, Karachi, in press. Kenoyer, J. M. 2006. The Origin, Context and Function of the Indus Script: Recent Insights from Harappa. Proceedings of the Pre symposium of RIHN and 7 th ESCA Harvard Kyoto Roundtable, edited by O. Toshiki, pp. 9 27. Research Institute for Humanity and Nature (RIHN), Kyoto. Kenoyer, J. M. 2007. Personal Communication Kenoyer, J. M. 2008. Indus Urbanism: New Perspectives on its Origin and Character. Presented at the Arthur M. Sackler Colloquium of the National Academy of Sciences, Early Cities: New Perspectives on Pre Industrial Urbanism, in press. 20

Jamison et al. 2017: 01 21 Kenoyer, J. M., and M. Vidale. 1992. A New Look at Stone Drills of the Indus Valley Tradition. In Materials Issues in Art and Archaeology, edited by P. B. Vandiver, J.R. Druzik, G.S. Wheeler, and I. Freestone, pp. 495 518. vol. III. Materials Research Society, Pittsburgh. Kenoyer, J. M., M. Vidale, and K. K. Bhan. 1994. Carnelian Bead Production in Khambhat, India: An Ethnoarchaeological Study. In Living Traditions: Studies in the Ethnoarchaeology of South Asia, edited by B. Allchin, pp. 283 306. Oxford and IBH Publishing, New Delhi. Mackay, E. J. H. 1931. Chapter XXI: Seals, Seal Impressions, and Copper Tablets, with Tabulation. In Mohenjo daro and the Indus Civilization, edited by Sir J. R. Marshall, pp. 370 405. Probsthain, London Meadow, R. H., and J. M. Kenoyer. 2000. The Tiny Steatite Seals (Incised Steatite Tablets) of Harappa: Some Observations on Their Context and Dating. In South Asian Archaeology 1997, edited by M. Taddei and G. De Marco, pp. 321 340. Istituto Italiano per L Africa e L Oriente, Rome. Patel, A. 2005. Copper Artifacts from Bagasra (Gola Dhoro), a Harappan Site of Gujarat, Western India. Puratattva 36:220 231. Possehl, G. L. 1998. Socioultural Complexity Without the State: The Indus Civilization. In Archaic States, edited by G. M. Feinman and J. Marcus, pp. 261 292. School of American Research Press, Sante Fe. Possehl, G. L. 2002. The Indus Civilization: A Contemporary Perspective. Altamira Press, Lanham, MD. Possehl, G. L. 2003. The Indus Civilization: An Introduction to Environment, Subsistence, and Culture History. In Ethnobiology and the Indus Civilization, edited by S. Weber and W. R. Belcher, pp. 1 20. Lexington Books, Lanham, ML. Rao, S. R. 1985. Lothal: A Harappan Port Town (1955 62), 2 Vols. Memoirs of the Archaeological Survey of India 78. ASI, New Delhi. Rissman, P. C. 1989. The Organization of Seal Production in the Harappan Civilization. In Old Problems and New Perspectives in the Archaeology of South Asia, edited by J. M. Kenoyer, pp. 159 170. Wisconsin Archaeological Reports. vol. 2. Department of Anthropology, University of Wisconsin, Madison, WI. Shaffer, J. G. 1982. Harappan Culture: A Reconsideration. In Harappan Civilization, edited by G. L. Possehl, pp. 41 50. Oxford and IBH Publishing, New Delhi. Shaffer, J. G. 1992. The Indus Valley, Baluchistan and Helmand Traditions: Neolithic Through Bronze Age. In Chronologies in Old World Archaeology (3rd Edition), edited by R. Ehrich, pp. 441 464. 3rd ed. vol. 1. 2 vols. University of Chicago Press, Chicago. Sonawane, V. H., P. Ajithprasad, K. K. Bhan, K. Krishnan, S. Pratapchandran, A. Majumdar, A. K. Patel, and J. Menon. 2003. Excavations at Bagasra 1996 2003: A Preliminary Report. Man and Environment XXVIII(2):20 50. Vats, M. S. 1940. Excavations at Harappa. Government of India Press, Delhi. 21