The debate on the caste census must be rescued from the immediacy of electoral pragmatics and party passions and relaunched as a dialogue among citizens on the very nature of nationhood and democracy that we would like to bequeath to the next generations. In this sense, the decennial census must be positioned as a cartographic exercise that dynamically mirrors India’s immense cultural diversity and socio-economic inequality, including caste. The postcolonial interruption in the colonial practice of counting caste in the census by dropping it is paradoxical. Mainly because the background assumptions against ‘caste’ enumeration—that it promotes caste populism, competitive casteism, social divisions, votebank politics, and so on—are severely undercut by the subsequent retention of ‘religion’, a category that is vulnerable to similar fallouts. In the imagination of the ruling elite, it seems, all socio-cultural divisions are equal, but some are more equal than others. The elite repression of caste enumeration—encapsulated in what Gail Omvedt refers to as the “three monkeys” policy: see no caste, hear no caste, speak no caste—stands in sharp contrast to the valorisation of religion as a category. I will return to this point later.