One way to understand the connections between ancient Indus sites, often separated by enormous distances, is to look ever more closely at some of their smallest artefacts – in this case unicorn seals, usually merely 5 cm square. While the results of this cross-site analysis of seals are preliminary, they do point to some significant connections among the seal manufacturing and/or using populations. They also seem to suggest that Mohenjo-daro, where by far the largest number of unicorn seals (and seals in general) have been discovered, played a similarly outsized role in determining and making seals used elsewhere. The evidence also seems to suggest that one can distinguish changes in seal carving styles over 700 years of manufacture across sites.
The 350 unicorn seals selected for comparative analysis are examined across two axes: stylistic elements, like the eye or ear of the single-horned creature; and the proportions of its compositional elements, including the all-important “standard” found below the head of almost all such seals (see Image 2). “Seals found to share at least seven attribute carving styles and five metric proportions,” write the authors, “were classified into groups and subjected to further visual analyses, based on the results of earlier research including experimental and ethnoarchaeological replication studies (Jamison 2013, 2017). We argue that these represent the products of individual artisans and workshops that produced Indus seals at Mohenjo-daro and elsewhere, recognizing that since they were carved by hand and usually small (less than 5cm in any dimension), identical copies are uncommon and unexpected,” (p. 23).
About 1/7th of the total number of seals analyzed fell into some 14 groups which share sufficient stylistic and metric attributes. All have at least one seal from Mohenjo-daro (a 100 seals were also included from Harappa). These allowed the authors to declare that “collectively, all identified seal groups represent stylistic links between Mohenjo-daro and other Indus sites and regions,” (p. 25). Some the groups suggest that the seals were manufactured in Mohenjo-daro and then made their way to smaller settlements like Allahdino or Chanhu-daro in Sindh, Kalibangan in the Ghaggar-Hakra region, or Lothal and Bagasra in Gujarat. Most interestingly, “Group #8 is comprised of three seals, one each from the major sites of Mohenjo-daro, Harappa, and Dholavira. It is the first and only example of a stylistic group comprised of seals from three Indus urban centers and provides evidence of links among them. All three are large, well made, and remarkably preserved, suggesting they were valuable and cared for by their owners. Though minor variation is present in the carving styles of a few compositional elements and proportions, they also demonstrate a high level of coherence, which is likely not coincidental,” (p. 27). What might this say about ties between such widely separated urban centers in what remain today three very distinct regions (and language groups) in the subcontinent? Another group (#7, Image 3) with similarities among three seals from Harappa and one from Mohenjo-daro seems to emphasize what must have been significant cross-urban connections and continuities.
The authors also discuss possible chronological variations in carving styles across regions, and with the inter-regional connections allow them to conclude that these all “support the inference that unicorn seal production and use was widespread and not limited to just the major urban centers,” (p. 30). They are fully aware of the fact too that this is just the first set of steps that need to be followed up with far more detailed – if not microscopic analysis – of the seals themselves. In short, a significant and important step has been taken to examining the nature of ancient Indus society by looking ever more deeply into the apparently important sinews like seals that bound it and its peoples together.
Two asides: While not part of the analysis, and possibly because the samples in each group were so mostly so small, the seals in the stylistic/metric groups do not seem to suggest any similarities in the signs carved; one wonders if this could have been addressed more directly, and even whether finding partial sign sequence similarities offers another approach to examining or grouping seals for further stylistic/metric analysis (it is likely that AI could help sort and find patterns to present to archaeologists for further study). Secondly, the map of Mohenjo-daro included in the article (Image 4), with excavated areas in the 250 hectare site drawn in, reminds us of how much of this major Indus center remains unknown. How many thousands of seals are left for future archaeologists to find?
Images:
1. Indus Unicorn Seal.
2. Stylistic and Metric Attributes used in Formal Analyses of Unicorn Seals.
3. Stylistic Seal Groups #3, 5, 6, 11, & 12 From Mohenjo-daro and Sites in Gujarat.
4. Plan Map of Mohenjo-daro.