This is an extremely important paper, just published in Nature. It completely scrambles timelines around South Asia and the development of agriculture in the region. Using new radiocarbon analyses of human tooth enamel from 23 Neolithic burials at Mehrgarh, the study demonstrates that the aceramic Neolithic occupation (Period I) began between 5223 and 4914 BCE and ended between 4769 and 4679 BCE, lasting between 186 and 531 years. This is a much later and shorter timeframe than previously thought, which had suggested a start as early as 8000 BCE and a duration of up to three millennia. In other words, Mehrgarh was developed two to three thousand years later than previously thought.
The new dating - which corrects less accurate dating methods used by earlier excavators from charcoal samples – also implies, through a comparison with other sites in the region towards the west along a number of dimensions, that agriculture in the Indus Valley emerged as a result of a relatively late diffusion of farming populations and practices from the west (the Fertile Crescent and Iranian Plateau), rather than through local, independent development at Mehrgarh. "These new analyses were performed on [dental] samples from exceptional, primary contexts with no stratigraphic disturbance" (p. 3). The evidence supports a model of agricultural diffusion from the west, consistent with bioarchaeological data showing West Asian domesticated animals at Mehrgarh from the outset. However, the authors note that the process was likely multidirectional and staggered, with possible local innovations and adaptations.
Pottery at Mehrgarh (Period IIA) appeared only after 4650 BCE, which is later than previously believed. This suggests that the earliest Neolithic communities at Mehrgarh were aceramic, even as pottery was already widespread in neighboring regions. Nonetheless, the absence of pottery at Mehrgarh during Period I, despite knowledge of the technology and availability of resources, suggests deliberate cultural choices rather than technological incapacity.
Furthermore, the thick Neolithic deposits at Mehrgarh were formed much faster than previously assumed. The nine cemetery levels, alternating with architectural phases, accumulated over a relatively short period, challenging previous assumptions about the tempo of site formation and occupation cycles.
More research into the reasons behind the delayed adoption of pottery at Mehrgarh, and the social and economic dynamics of these early farming communities, could provide deeper insight into Neolithic developments and movements of people and/or technologies. The study highlights significant gaps in the archaeological record between Mehrgarh and contemporaneous sites in Iran and Central Asia. Systematic surveys and excavations in these understudied regions could clarify the routes and mechanisms of Neolithic diffusion. Applying the latest radiocarbon techniques (especially on tooth enamel and other secure contexts) to other South Asian and Iranian sites will help refine regional chronologies and test the generalizability of the Mehrgarh findings.
The findings necessitate a major revision of the chronology for the onset of agriculture in northwest South Asia; most books and articles will have to be updated. Mehrgarh, long considered one of the earliest farming sites in the world, is now shown to be significantly younger than previously thought. It also highlights the importance of robust, contextually secure radiocarbon dating in archaeological chronology. That said, much more research is also needed to the east of the Indus Valley, and more secure datings from sites there as well, so the story of the development of agriculture and the rise of cities in the Indus basin and surrounding areas is far from complete.
Images: 1. Map with location of Mehrgarh and additional sites mentioned in this article. 2. Result of the v1.1 Bayesian model of Mehrgarh Period I.