Prasanta Chandra Mahalanobis is remembered as the father of Indian statistics – but his academic work in other disciplines is not very well known.
For example, today, we don’t speak about Mahalanobis’s work on meteorology, astrology or – for that matter – racial biometrics.
India is today grappling with difficult questions about identity and belonging, and trying to stave off the reductive solutions its government has advanced. Against this backdrop, Mahalanobis’s analysis of caste, religion and national identity illustrate what he got wrong and, more importantly, what our government is getting wrong.
The evidence that Mahalanobis’s early work was steeped in questions about racial biometrics comes in the form of two tools he designed: one statistical, and the other mechanical.
He developed the statistical tool when working on the biometric data of Anglo-Indians, and used it to measure what he called “caste distances”.
The statistical methods that scientists had at the time could indicate whether an individual was part of a population. That is, given a person’s height and head circumference, scientists purported to be able to determine which racial group they belonged to.
However, these were just ‘yes’ or ‘no’ questions. Mahalanobis wanted to know how much more likely they were to belong to that group than to another group. In other words, he was looking for the degrees of relatedness between races.
“How are … Anglo-Indians … related to the different caste groups of Bengal?” he wrote in a 1925 paper entitled ‘Analysis of Race Mixture in Bengal’. “Are they more closely allied with the Hindus or with the Mahomedans?”
To answer this question, Mahalanobis reasoned that he would first have to compare individual physical traits separately, before combining them to produce a relative value. He called this number, or ‘caste distance’, the D2 statistic.
“Two castes which resemble each other very closely will have a small caste distance,” he wrote in the 1925 paper. “Castes which are widely different in character will have [a] large caste distance.”
For example, Brahmins, Mahalanobis wrote in 1949, “have medium nasal length, [high] nasal depth, and narrow nasal breadth.” He also observed that Basti Brahmins had deeper noses and longer faces than other Brahmins. Tribal groups had short, shallow, and wide noses; Muslims and Chattris had similar facial features; and Chamars had broader faces than Kahars.
Mahalanobis’s description of the castes of Uttar Pradesh
Mahalanobis built the mechanical tool known as the photographic profiloscope to accurately measure people’s facial features. The device was essentially a modified camera attached to a chair and captured people’s silhouettes.
His interest in racial biometrics wasn’t just a matter of scientific curiosity. He didn’t build statistical and mechanical tools merely as an intellectual exercise. Instead, he was obsessed with figuring out the physical relationships between social groups.
Projit Bihari Mukharji, an associate professor at the University of Pennsylvania, has studied the profiloscope in detail. To explain Mahalanobis’s actions, he situates the construction of the profiloscope in the political turmoil of early 1930s’ British India.
At this time, B.R. Ambedkar and M.K. Gandhi were bitterly divided on whether Dalits should fall within the Hindu fold to form a unified electorate. “This was also the time,” Mukharji said, “when intractable political questions about who belongs to the nation and who doesn’t were being raised.”
The Bengali Dalit leader Jogendranath Mandal, Mukharji wrote in a 2015 paper, “argued for an alliance of the lower castes with the Muslim League and its programme of Muslim nationalism.”
Later in the same text, Mukharji continued: “Mahalanobis … sought to clarify the contours of the nation… The profiloscope promised to provide a clear outline of the national races available [in] India.”
In other words, Mahalanobis wanted to answer questions like ‘who is Indian?’ and ‘what constitutes an Indian identity?’
Today, the Citizenship (Amendment) Act (CAA) 2019, and the National Register of Citizens (NRC) have forced us to reckon with these questions once again.
The CAA and the NRC are respectively efforts to redefine and codify an Indian identity. Since this cuts to the heart of what constitutes a nation, and the people that populate it, they are clearly nationalist activities. Mukharji observed that Mahalanobis’s biometric projects, including the profiloscope, were also nationalist projects, all collected under the umbrella of biometric nationalism.
And like Mahalanobis, the current government mistakenly believes that a technocratic fix can address problems of belonging. “The idea that biometrics is somehow a guide to answering these questions,” Mukharji said, is a shared fantasy.
A schematic diagram of a profiloscope together with examples of the device in use
Such technological solutionism only offers a delusion. But even when operating within this fallacy, Mahalanobis’s biometric projects differ from the current government’s efforts in some significant ways. To understand how, it is important to dig deeper into the features of biometric nationalism itself.
“Biometric nationalism has two strands,” Mukharji explained, and what separates them is key to understanding why Mahalanobis’s ideology doesn’t converge with the Bharatiya Janata Party’s.
The first strand is called pan-Indian nationalism, the idea that communities across India are horizontally connected. That is, all upper-caste Hindus across the country, the thinking went, shared a common identity that distinguished them from, say, all Christians in the country.
Contemporaries of Mahalanobis who were engaged in anthropological studies – like the Bengali intellectual Atul Krishna Sur, and Brija Shankar Guha, the first director of the Anthropological Survey of India – subscribed to this school of thought. It also resonated with the right-wing political apparatus operating in Bengal.
In 2017, Mukharji wrote that in 1942, the Hindu Mahasabha published a pamphlet entitled ‘Bengalir Nritattwik Porichoy’ (Bengali for ‘the anthropological identity of Bengalis’) that attempted to prove using biometrics that upper-caste Bengalis were closer to the upper-caste peoples in neighbouring regions than to the province’s lower-caste members.
Images of individual silhouettes obtained by use of the profiloscope
However, Mahalanobis was firmly in the second camp. “He wanted to prove that Indian populations were vertically integrated,” Mukharji said. “So, say, a Bengali brahmin is closer to a Bengali person of lower caste, than, say, to a Tamil brahmin.”
“Both strands try and make cases about who belongs and who does not belong to the nation, and what the shape of that nation is,” he continued. “Mahalanobis was on one side, which was closer to a linguistic and regional nationalism. And he opposed people who were increasingly making the case for a pan-India kind of nationalism.”
This distinction is the principal difference between the government’s efforts to define an ‘Indian identity’ and Mahalanobis’s search for something similar. The government’s nationalism is pan-Indian, religious and exclusionary. Mahalanobis’s nationalism was regional and/or linguistic and more inclusive.
Mahalanobis wanted to use the profiloscope to prove that a pan-Indian identity had little biometric meaning, and in doing so, he pointed the way to unite everyone in an incipient nation. Racial identities, Mahalanobis reasoned, were mutable and therefore could be influenced by location. In this paradigm, caste, race and religion don’t have to come in the way of constituting a national identity.
The CAA makes the opposite claim – that Afghan Hindus, for example, stake a higher claim to an Indian identity than Afghan Muslims. Here, regional identities play second-fiddle to religious identities, and this is precisely why the CAA must be opposed.
The NRC promises an apparently clear definition of a national identity – a technical fix for a problem of belonging. It is an act that outsources a moral and political debate to the bureaucratic machinery, and thus a tool that will only facilitate discrimination based on religion or documentation.
The scientific history of Mahalanobis’s profiloscope, along with his other biometric projects, offers us a helpful lesson in how to synthesise the concept of a national identity. It is unfortunate that our collective memory of him hasn’t accommodated this dimension of his diverse interests. His signature contribution to the field of statistics – the D2 coefficient – is today called the Mahalanobis distance. Its racial origins, much like the origins of the profiloscope, have been buried by the sands of time.