What was really going on at the so-called stupa mound in Mohenjo-daro? This important paper by Giovanni Verardi and Federica Barba challenges long-standings interpretation of the so-called Stupa Mound at Mohenjo Daro as a Buddhist site dating to the 2nd century AD. Through careful analysis of original excavation reports, photographs, and recent site visits, the authors argue this structure on what excavators called Site 1 was actually, beneath any later Buddhist structures, fundamentally an Indus period stepped temple comparable, in their opinion, to Mesopotamian ziggurats.
Dr. Verardi first raised doubts about the Buddhist interpretation in 1987, but here he and Dr. Barba build their case by examining R.D. Banerji's original excavation reports (largely unpublished until 1984, now available online). They also critically reassess John Marshall's interpretations in his 1931 book Mohenjo-daro and the Indus Civilization (below is a link to the pages in Marshall's report which endorse the Buddhist temple interpretation, but also includes some doubts on his part). They analyze photographic evidence, and incorporate observations from a 2007 site visit (the photographs above are from 2025). Dr. Banerji, the first excavator at Mohenjo-daro, originally described Site 1 as "a high platform or tower" with rooms surrounding a courtyard, containing a "solid platform of burnt bricks" in the center. On this platform rose what Banerji called "a hollow stupa" built of sun-dried bricks. Marshall later endorsed the Buddhist interpretation, creating a narrative of a stupa-cum-monastery complex for the entire edifice that has persisted. Marshall did this while, the authors argue, suppressing some of the evidence Banerji presented that showed the showed the whole complex to be much older.
The authors have a number of solid arguments: the stupa's rectangular structure is not square as plinths for stupas usually are and yielded far fewer Buddhist objects than similar sites. Much of the construction around the stupa is from the Indus period, and "thousands" of votive offerings in Indus pointed vessels and larger pots seem to have been found by Bannerji in Indus-period rooms around it, suggesting a possible holy site. Bannerji interpreted these "funeral urns" (vessels containing ashes and bones) as burials but the authors reinterpret as votive deposits related to religious offerings. Marshall did not properly discuss these, "omitting the evidence for which he was unable to find an explanation," in their view (p. 271). They write:
"Repositories similar to those of Site 1 have been identified in other cultural contexts as votive deposits, from ancient Mesopotamia to the Greco-Roman world. The main types are ‘obliteration deposits’ or favissae, ‘foundation deposits’ and ‘propitiation deposits’ (Bonghi Jovino 2005: 35-41). The favissae are pits or sealed rooms containing votive materials or ex-votos, which could not be disposed of when they fell into disuse because of their sacred nature. They usually contain votive statues or large quantities of pottery carefully arranged near the precincts of a temple. The other repositories have usually the shape of a pit or a jar containing terracotta figurines and miniature pots – votive objects intended to obtain the favour of the god. The ‘foundation deposits’ are generally sealed as they are built underneath or within a wall with the purpose of consecrating a structure to a god. Conversely, the 'propitiation deposits’ are often concealed and visible at the same time: for example, they can be covered by stone slabs, which conceal them indicating at the same time their presence" (p. 272).
Through architectural analysis and artifact examination, the authors identify the structure as a stepped monument with multiple terraces, accessed by front and side stairways, some of which are partially visible today (see image 4 and 5 above). The double stairway identified by Banerji (barely visible anymore, see top of Image 4) could not have led to a stupa's circumambulation area, they argue, because there was not enough space on top if had been a stupa. In fact they say that in general the building techniques match those of the protohistoric town and other areas in the citadel. The so-called "monastic cells" (actually rooms) around are also highly irregular compared to typical Buddhist monasteries. Instead they surmise that the structure shows a stepped design clearly visible in Banerji's photographs.
They conclude this was the central religious building of Mohenjo-daro, purpose-built for ritual activities, and propose that Site 1 and the Great Bath should be considered together as forming the sacred area of the ancient city. In fact Banerji's initial interpretations were that this was an Indus holy site, perhaps an island, around which the river flowed, but Marshall discounted this interpretation. They also argue that many scholars have regarded the "Great Bath" adjacent to Site 1 as some sort of public ritual structure, and for a temple to be adjacent to it would make much sense.
The paper doesn't include results from new extensive excavations that could further validate their claims, and whether or not a temple of sorts existed where the Buddhist structure is would need to be shown, if possible, through new excavations beneath the area today. That said, it is common in the subcontinent for a new religion to build atop the buildings of an older one, and this may well be why Buddhists put up a temple of some sort here (Banerji did find the crumbling remains of what he called a Buddha in one alcove). The main Buddhist temple from the 2nd-3rd centuries ACE in the region was a few kilometers away. The German excavator, Dr. Michael Jansen, who put together much of the data from earlier excavations at the site in the 1980s, supported this new reading of the stupa mound and citadel area although he wanted more decisive proof. Given the way stepped ziggurat temples were the nucleus of ancient Mesopotamian cities it is not unreasonable to think the same might have been the case in the Indus Valley; temples then could control land outside and craft areas within the city. The Site 1 area is closely connected to the L Area adjacent to it where specialized craft activities were carried out. Dr. Verardi's drawing [Image 2, above] is his best guess from the data and knowledge of ziggurats as to what one such version of a temple at Mohenjo-daro might have looked like (the comparison with Mesopotamian ziggurats, while intriguing, requires more research to determine if it represents cultural exchange or independent development, and we really do not know how old Mohenjo-daro was).
While initially classified as a Buddhist religious monument from a much later period than the original city, this paper posits that the structure may have far more ancient origins and religious significance within the Indus civilization itself, and brings together a variety of points that reach back to the first excavator's work to support many of its arguments. "There is little doubt that the monastery-cum-stupa is an Indus building, except the mud-brick structure, which alone emerged when Banerji started the excavation. The strata attributed to the so-called Buddhist period yield Indus material and, as is shown by the stratigraphy, belong to the late phase of the protohistoric town" (p. 271-2). If they are correct, and there arguments do seem compelling, this could help lead to many more insights about the hitherto little understood Indus culture and belief system.
Images: 1. Mohenjo-daro view of Stupa mound at sunset. 2. 3D rendering by G. Veneroso in black and white, later colorized. 3. Corner view of stupa. 4. Steps leading up to stupa. 5. Top of steps leading up to stupa. (All photographs 2025).
