A view of the excavated city of Mohenjo Daro, April 2014. Photo: Usman.pg/Wikimedia Commons, CC BY-SA 3.0.
There is new evidence that the people of the Indus Valley civilisation (IVC) had a mixed diet, with meals possibly including pig, buffalo, sheep and goat meat. The evidence comes from a new study of fat residues in ancient ceramic vessels unearthed from northwest India and adds to our currently meagre knowledge of ancient food in South Asia.
A new paper in the Journal of Archaeological Science claims to be the first to investigate absorbed lipid residues in pottery from multiple Indus sites. This allows scientists to compare data from different IVC settlements across space and time time, Akshyeta Suryanarayan, one of the study’s authors and a postdoc at the Culture et Environnement, Prehistoire, Antique, Moyen Age, France. Suryanarayan was with the University of Cambridge when she conducted the study.
According to her, previous studies have examined the remains of plants and animals dated to the time of the IVC – but “lipid residue analyses enable us to get a glimpse into foodstuffs and the cultural use of pottery vessels,” Suryanarayan told The Wire Science. “This study provides chemical evidence for cooking of milk products, meat and possible mixtures of products or plants in pottery vessels.”
“Our paper contributes to a growing body of evidence that suggests people in the Indus Civilisation ate a mixed diet that included a wide range of plant products, as well as meat and also freshwater (and marine) resources,” she added.
Prabodh Shirvalkar, an assistant professor at the Deccan College Postgraduate Research Institute, Pune, and an expert in Harappan archaeology, said the study provides useful insights. However, he also cautioned that the study’s findings are sweeping. The researchers examined samples obtained from seven sites, and according to Shirvalkar, their findings are applicable only to areas with the same ecological and environmental characteristics.
“It is too early to extrapolate the findings to the entire Harappan civilisation,” Shirvalkar told The Wire Science. The researchers’ paper also says the same thing: that they need “a large number of samples from different regions to confirm the conclusions.”
The Indus Valley Civilisation was one of the first complex civilisations of the Old World, spread across large parts of modern Pakistan, northwest and west India and Afghanistan. For the study, a team of researchers from India, Spain and the UK selected seven sites on the semi-arid alluvial plains of northwest India, in modern-day Haryana and Uttar Pradesh. The region represents the ‘eastern domain’ of the civilisation.
The scientists said something similar to the climate of today – hot summers and cool winters – likely existed in the past as well. Animal bones found at Indus sites include those of cattle, such as water buffaloes, and of sheep and goat, deer, pigs, birds and fish. But the bones of cattle dominate.
The team analysed 172 pottery fragments, including rims of vessels where lipids of boiling food could accumulate, and shards recovered from rural and urban settlements. They analysed the lipids in organic residues in the Indus vessels. Lipid analysis allows researchers to distinguish between the meat of grass-eating, or ruminant, animals and the meat of non-ruminant animals.
The underlying technique is called mass spectrometry, which elicits the mass-to-charge ratio of ions in the sample. Based on this data, researchers can determine the masses of different atoms and molecules in the sample.
The results indicated that while the vessels did contain lipid residues, their concentrations were low. The team also observed that dairy products, ruminant carcass meat and either separate or mixtures of plants and fats of non-ruminant animals like pigs formed part of the cooking in Indus vessels.
“Much to our surprise, we found that most of the vessel residues demonstrated values for non-ruminant or omnivorous animal fats,” Suryanarayan said.
However, according to the team’s paper, there is “limited” evidence for direct plant-processing and dairy products, and the interpretation of a large proportion of the data is “presently ambiguous”.
Archaeologists have previously found remains – mostly bones – of omnivorous animals at Indus sites, including those of pigs, hares and birds. The bones of omnivores occurred in much smaller quantities relative to domestic ruminant animals like water buffaloes, sheep and goats.
The results also suggested that people of the IVC in both ‘rural’ and ‘urban’ parts used similar vessels.
In fact, also according to the paper, the results were “unexpected and challenging to interpret.” The data of nearly 60% of the vessels analysed overlapped with the range of non-ruminant fats – even as the lipid profiles of most vessels were characterised by fatty acids common in ruminant fats.
Second, the results don’t correlate with available animal remains in the region, such as bones. Only 2-3% of the identifiable specimens from the study sites comprised non-ruminant animals like pigs, fowls and hares.
Suryanarayan also said vessels may have been used repeatedly for multiple plant and animal products over time, thus rendering efforts to determine the origins of each set of compounds more difficult. The “evidence does not fit just one interpretation – so it is important to acknowledge the uncertainty in the results.”
Insights and limitations
Suryanarayan said a previous study had examined lipid residues in a single perforated vessel from Naushahro and suggested it had been used to hold dairy products – something her team’s paper hasn’t found evidence for. Another study investigated vessels from a Sorath Harappan site in Gujarat and reached conclusions similar to the present one, albeit with more evidence of dairy products as well.
Shirvalkar also said plants are harder to identify with this technique as they contain less fat – “there is a 90% chance that fatty acids in plants will not be detected” – so the findings as a result could be biased towards animal fats.
He added that lipid analysis as a technique could help differentiate lipids from the remains of ruminant and non-ruminant animals – but beyond that, it can’t say which species of animal the lipids could have come from. To bridge this gap, the researchers combined their lipid analysis results with findings from bones from previous archaeo-geological expeditions.
A major factor affecting the study of ancient food in South Asia, particularly in northwest India, is the degree of organic preservation at archaeological sites, according to the researchers’ paper. Fluctuations in temperature and moisture, pH levels and mineralisation negatively affect how well organic material is preserved in South Asian archaeological sites.
The researchers have indicated the need to build a database of reliable local isotopic references for fats and oils, to allow other scientists to quickly interpret their results without ambiguity. They also wrote that assessing changes over cultural and climatic periods will require scientists to sample more pottery from well-dated contexts.
The results demonstrate that the use of organic residue analysis in South Asia, combined with other bio-archaeological approaches, could reveal the full diversity of prehistoric South Asian food systems as well as the evolving relationship between pottery and foodstuff over time.
Note: This article was updated at 9:39 am on December 25, 2020, to correct Akshyeta Suryanarayan’s sex, to clarify the provenance of some quoted statements and to remove mention of “hair and hides” as animal remains in consideration.
T.V. Padma is a freelance science journalist.