Posts Tagged ‘Census’

Hinduism – 2: Getting to Terms with Religion

October 16, 2008

Continued from Hinduism – 1: What is ‘Hinduism’?

It’s time to remove the quotation marks around ‘Hinduism’.

It just adds to the confusion when one argues in this day that Hinduism is not a religion in the sense religion is understood in the Judeo-Christian tradition. It is better to explain that ‘religion’ has a wider scope.

See how religion is defined in the 1828 American Dictionary of the English Language:

Religion, in its most comprehensive sense, includes a belief in the being and perfections of God, in the revelation of his will to man, in man’s obligation to obey his commands, in a state of reward and punishment, and in man’s accountableness to God; and also true godliness or piety of life, with the practice of all moral duties.

If one starts with that definition it would be very hard to fit Hinduism into the mould.

However, one can take a modern perspective and understand religion as a “system of symbols (creed, code, cultus) by means of which people (a community) orient themselves in the world with reference to both ordinary and extraordinary powers, meanings, and values.” In this perspective, Hinduism is a religion but with characteristics very different from those of a Judeo-Christian religion.

One can immediately see the big difference between Islam and Hinduism in this context. The former placed a lot of emphasis on the ‘true’ word of God, quarreling even amongst fellow Muslims on correct interpretations of the true word. The latter placed much more emphasis on the practices of everyday life with a lot more room for variations from any externally prescribed way.

One can also now understand the puzzlement of the British when they got ready to conduct the first major census in the Punjab in 1868. The presence of Muslims and Sikhs convinced them that Punjab society could only be understood through the lens of religion.  But when they attempted to define Hinduism they were stumped: “Primarily and historically, it is the antithesis of Islam. Religion, in the etymological sense of the word, it is not, and never was.”

Denzil Ibbetson in the Punjab Census Report of 1881 wrote that “the books on Hinduism describe Hinduism as it ought to be, Hinduism as it once was, perhaps Hinduism as it now is among the pandits and educated Brahmans of the holy cities; but they do not describe Hinduism as it is in the daily life of the great mass of the population.”

And thus the rule that was adopted for the census definition of Hindus was “that the Native of India must be presumed to be Hindu unless he belongs to some other recognized faiths.”

This different nature of Hinduism as a religion was not trivial and one would need to keep it in mind as we proceed to explore its interaction with Muslims and the British in the posts to follow.

To be continued…

The definitions of religion are to be found in a thoughtful post in Religion Dispatches. The material on the Punjab Census is from In the Making: Identity Formation in South Asia by Kamaljit Bhasin-Malik, Three Essays Collective, New Delhi, 2007.

Back to Main Page


Building Democracy in Iraq

June 11, 2008

We have been discussing the census, electoral rules, and the nature of democracy in South and East Asian countries trying to draw lessons from events that happened between fifty and a hundred and fifty years ago.

It was therefore eerie to read a virtual replay that took place in Iraq only a few years back. We truly ignore history at our own peril.

The account is from the 2006 book by Rajiv Chandrasekaran (Imperial Life in the Emerald City: Inside Iraq’s Green Zone), an account of the American occupation of Iraq and the attempts to reconstruct the country. Here we shall reproduce just the bare essence that indicates the overlap with our earlier posts. Readers interested in the details should be able to obtain the book fairly easily.

From April 2003 to June 2004, Iraq was governed by the Coalition Provisional Authority (CPA), the American occupation administration, headed by Paul Bremer. Amongst the objectives of the CPA was to hold elections in Iraq and transform it into a democracy.

The first step was to select a Governing Council of Iraqis that would prepare the rules for the general elections to follow. Here is how Chandrasekaran describes what followed:

Their lack of experience led to a fundamental miscalculation. They tried to right Saddam’s wrongs by engaging in social engineering, favoring the once-oppressed Shiites and Kurds at the expense of the once-ruling Sunnis…. The result was a Governing Council that had strict quotas: thirteen Shiite Arabs, five Sunni Arabs, five Sunni Kurds, One Christian, and one Turkmen. To some Iraqis, who placed national identity over religious or ethnic affiliation… the Americans changed all that. They made a point of categorizing people as Sunni or Shiite or Kurd.

Chandrasekaran records the misgivings of one American adviser in Iraq who believed that ‘forming a democracy was easy, but forming a liberal, moderate democracy wasn’t’.

He believed that the CPA had committed a catastrophic error by establishing a quota for Sunnis, Shiites, and Kurds on the Governing Council. [He] believed that Iraqis hadn’t focused on ethnic and religious divisions before the war, and that it was the CPA’s quota system that had encouraged them to identify themselves by race or sect.

Bremer’s approach magnified rather than muted the very divisions that so many Iraqis rejected… The best Iraqis knew that they could not form one country, one democratic country, unless they were somehow able to get these categories behind them and look for leaders who, one way or another, would transcend these divisions. The best Iraqis… knew this. We didn’t.

We now move on to a discussion of electoral rules. ‘On June 1 [2004], United Nations envoy Lakhdar Brahimi announced the names of the members of the interim government that would assume power on June 30, when Bremer was scheduled to leave and the CPA [along with the Governing Council] was to be dissolved’.

The interim government… would remain in power for seven months, until January 2005, when elections would be held for a national assembly…. The biggest impediment to holding elections was the lack of an up-to-date census…. Without a census, there was no accurate way of knowing how many people lived in each province and, as a consequence, how to apportion seats in the assembly.

The United Nations team determined that there was no reasonable way to conduct a nationwide census before January 2005, the date by which the interim constitution required the first election to be held. The UN team, which put the goal of holding a perfect election over everything else, told the CPA that the only way to meet the deadline was to consider the entire country as a single electoral district. All Iraqis, no matter where they lived, would get to choose from the same list of candidates. The candidates could choose to run on their own, or they could band together with other members of their party and run as a slate. The number of votes a party received would determine how many members of its slate got seats in the assembly.

It was technically sound, if convoluted, but it had major flaws. The system, which required candidates to campaign nationwide, gave large parties a clear advantage over individuals and smaller parties. It would mean that the two dominant Kurdish parties and the two largest Shiite parties… would likely win a clear majority of seats, marginalizing moderates and secularists…. 

The single district wasn’t the only option on the table… a national database used to dole out monthly food rations could be used to provide a reasonably accurate estimate of how many people lived in each province….

Chandrasekaran quotes a senior CPA staffer:

There were plenty of us who said that a single-district election would be a disaster and that you could have a proportional representation system with the ration [database]… But Bremer and his keepers at the White House didn’t care. The type of election was a secondary issue to them. What mattered more than anything else was holding it on time. It was style over substance. On June 15, Bremer signed CPA Order 96. It stated that Iraq ‘will be a single electoral constituency’.

Chandrasekaran records the outcome:

Millions of Iraqis had headed to the polls in January 2005 for the country’s first democratic elections in decades…. All told, Sunni Arabs, who comprised about 20 percent of Iraq’s population, wound up with fewer that 8 percent of the seats in the legislature. Bremer’s single-district electoral law had shut the Sunnis out of the new government, depriving the Americans, and the Iraqis, of a valuable opportunity to win over Sunnis and weaken the insurgency.

All the key ministries were claimed by the Kurds and the Shiites, whose militiamen swept up legions of young Sunni men… with the acquiescence of the new government…. A civil war had begun.

Eerie, isn’t it?

Back to Main Page

Democracy in India – 5

June 10, 2008

In this series of posts we have thus far highlighted the following propositions: 

1. The census introduced by the British in India (around 1870) classified people by religion. This was unlike the practice followed by the census in Britain itself. 

2. Instead of using the religious beliefs as reported by the respondents themselves, the census classified them into the broad categories of Hindus, Muslims, Sikhs, Christians, etc.

3. A complex social reality that comprised of many mixed traditions, practices, and beliefs was simplified into set of broad overall categories.

4. When religious identity moved into the political domain with the adoption of separate electorates the rigid classifications assumed a new importance because one group could only gain at the expense of others. 

In this post we shall see with the help of Kmaljit Bhasin-Malik’s text how this new reality and realization affected the behavior of different groups and the impact it had on the nature and development of electoral politics in British India.

The existence of the census and the emergence of representative politics marked the emergence of competition between different social groups over access to employment, education and power. This competition was set off by the existence of the census and very naturally led to attempts to influence the census by sharpening social and religious distinctions.

Thus ‘Arya Samajists wanted all Hindus to return themselves as Aryas’ while Sikhs began to stress their distinct identity because till the 1901 census only Khalsa Sikhs were defined as Sikhs. The Singh Sabha came out with a tract titled Ham Hindu Nahin or ‘We are not Hindu’. 

The Arya Samaj launched a vigorous campaign to convert low-caste Muslims and Sikhs through its shuddhi movement spurred by a concern over a declining Hindu population revealed by the census – ‘any demographic decline was seen as having political consequences in an age of growing democratization’. Hence ‘Hindu reformers reacted violently to the suggestion that the 1910 census should both widen the category of ‘Sikh’ and eliminate Untouchables from the Hindu category.  Religion was being gradually politicized.

This purification extended to shared social practices that had marked pre-census life in India. Thus Arya Samaj writings sought to ‘convince our Hindu brethren that it is repugnant to their religious doctrines and authorities – nay it is a sin – to pay homage at [Muslim] tombs and shrines’. Similarly the Tat Khalsa ‘tried to root out the attachment of Sikhs to Hindu rituals, Brahmans and the worship of saints and pirs’.

Kamaljit Bhasin-Malik concludes that while the narrative sections of the census were quite comfortable with ‘fuzzy’ and ‘fluid’ religious boundaries and eclectic and syncretic religious traditions, the statistical tables of the census were unable to deal with these sorts of overlaps and ambiguities. The census employed religion as a fundamental category and mapped, counted and compared religious communities in endless tables dealing with issues ranging from population, civil condition, occupation, language, to infirmity.

Also, the census definition of what it meant to be Hindu or Muslim was something new. Muslim identity, for instance, was viewed as an ethnic category… all Muslims regardless of social status, geographic location, and linguistic orientation, were viewed as constituting a unified whole…. Similarly, the census established ‘Hindu’ as a unified category and imputed a rarefied unity to Indic religious orientations and practices that had not previously existed.

The data tabulated in the census became the basis for the distribution of political power and government patronage. Religious communities were always measured in terms of numbers and numbers were equated with political power. The fusion of religious identity and numerical strength as symbolized by the enduring categories of ‘minority’ and ‘majority’ and that in turn with political power was heightened by constitutional reforms. In 1910 the British introduced separate electorates for the Punjab Legislative Council…. This meant that while making claims for political power, individuals had to represent themselves as the ‘natural leaders’ of their particular community, and as speaking for the interest of the entire community.

While the census did not ‘invent’ the categories of ‘Hindu’ and ‘Muslim’ and ‘Sikh’ it changed their definitions, enumerated them, and also made them the basis for political power and patronage as symbolized by the separate electorates for Hindus and Muslims. The census not only invested religious categories with new forms of social, economic, legal and political meaning, but also exercised a growing catalytic effect. It established a cycle of description, action and change, followed by another description. The census was drawn into the world it described and became one more arena for conflict and manipulation. This can be illustrated by the campaigns of the Arya Samaj and the Singh Sabhas to influence census categories and definitions. The census grew increasingly intertwined with political decisions like the Partition of the subcontinent along religious lines. The boundaries of the two new nation states of India and Pakistan were drawn by relying on the data generated by the decennial census.

We will conclude this set of posts with what we hope would be an intriguing twist to the theme of political development in British India.

Back to Main Page  

Democracy in India – 4

June 6, 2008

A number of readers have expressed reservations about our comments on the first census in British India (Democracy in India – 3).

It is argued that disclosure of full information is always for the better and cannot but be helpful in the long run.

This misses the point. It is not always the case that pre-existing information is lying unobserved and a neutral process is involved in bringing this knowledge into the public domain.

With the first census in British India, knowledge was actually created. This is what Kamaljit Bhasin-Malik explains in her essays on the census:

The Punjab census illustrates that the census was not a passive data-gathering instrument. It did not merely count what is.  Census officials first had to create categories and define them. But this was no simple process and the realities that census takers encountered collided with their imperial taxonomies, which assumed Punjabi society was neatly divided into separate religious categories.

This conclusion follows from a close reading of the census documents. As an example, Kamaljit refers to Sir Denzil Ibbetson, the author of The Punjab Census Report of 1881:

While laying out the ‘Census definition of Hindus’, Ibbetson began with a preamble, which warned of the ‘absolute impossibility of laying down any definition or indicating any test by which we may distinguish him who is a Hindu from him who is not’. Having said this however, Ibbetson cheerfully proceeded to lay down a definition and offer an explanation for it. ‘Practically’, he wrote, ‘the rule we adopted was this’:

Every native who was unable to define his creed, or described it by any other name than that of some recognized religion or sect of some such religion, was held to be classed as Hindu…

Describing the category ‘Hinduism’, the report of the Census of India1891 said that the ‘clumsy name is only justified by convention…. Religion, in the etymological sense of the word, it [Hinduism] is not, and never was. The binding element is only educed by active opposition on the part of some other faith, such as Islam’.

In the chapter on the ‘The Religion of the People’, in the Punjab Census Report of 1881, Ibbettson wrote that ‘Creed is in the Punjab rather a social than a religious institution’, and that it is ‘so difficult in many cases to draw the line between one Indian creed and another; for the distinctions of faith, being based upon and attended by no deep spiritual conviction, are marked by a laxity and catholicity of practice which would be impossible to a bigot or an enthusiast’.

Kamaljit Bhasin-Malik notes that ‘Although the census officials of the Punjab gave detailed descriptions of what can be seen as the ‘syncretic’ nature of popular religion in Punjab, they were unable to conceptualize this syncretism, and took for granted the separate and monolithic religions of Islam, Hinduism and Sikhism’.

Today, when we are living the reality that was thus created, we find it impossible even to think that there could have been a category called ‘Hindu-Mahomadan’ that a Bombay census official had recommended but that was rejected as a viable solution to the existence of mixed allegiances.

All the material in this post is from In the Making: Identity Formation in South Asia by Kamaljit Bhasin-Malik, Three Essays Collective, New Delhi, 2007 (

Back to Main Page    

Democracy in India – 3

June 2, 2008

We pick up where we left off in the previous post (Democracy in India – 2) on the subject of the introduction of the census in British India and its implications for the trajectory of political developments.

The objective is to show that the use of an administrative mechanism like the census is not neutral but has a definite purpose, is based on prior prejudice, and can have severe unintended consequences. For this interpretation we rely on the most recent writing available on the topic – In the Making: Identity Formation in South Asia by Kamaljit Bhasin-Malik, Three Essays Collective, New Delhi, 2007 ( 

The census was symptomatic of the Victorian urge to ‘know’, ‘classify’ and ‘count’. Census operations were not reserved just for the colonies. They were instituted in almost all European countries in the 18th and 19th centuries and were usually prompted by concerns about depopulation and increasing poverty. In Britain, the census was instituted… in 1800… [and] stressed two objectives. Firstly, to know accurately the current size of the population, and secondly, to know the trend of the population… in order to provide ‘correct knowledge’ of increasing or decreasing demands of subsistence’.

Significantly… the census in Britain never recorded data on religion, apart from one survey in 1851. This was despite the fact that in early nineteenth century Britain, religious affiliation mattered intensely… Until well into the nineteenth century the state awarded the right to vote on the basis of religious affiliation. Yet, however much religion may have informed British life, it was never imagined… as having the power to shape the entire society into opposed ‘communities’. One could argue that the project of nation-building that Britain was involved in at this time encouraged an emphasis on homogeneity and themes that united Britain.

In direct contrast to this, India had come to be understood as a land of ‘many nations’, and of ‘various and varying races’, as Disraeli described its people… As an article that appeared in The Times proclaimed, the basis of a ‘right understanding’ of Indian government ‘must be exact knowledge of the population not only as a whole, but in its manifold ethnographic, communal and geographic divisions; and this can be obtained only by a full and careful periodic enumeration.

This colonial understanding of India and Indian diversity, which gave centrality to religious community and caste, was institutionalized in the census.

We shall continue in the next post by documenting how religious categories were defined and the outcomes of this assignment of centrality to religion in the Indian census.

Back to Main Page 



Democracy in India – 2

May 31, 2008

In the first article of this series (Democracy in India – 1) we had highlighted the importance of the introduction of elective governance in India by the British, the choice of separate electorates based on religion, and its negative impact on communal relations.

The following quote from the Indian Statutory Commission in 1930 showed how religion was turned from a social distinction into a political one that mattered in terms of who got what:

So long as people had no part in the conduct of their government, there was little for members of one community to fear from the predominance of the other. The gradual introduction of constitutional reforms, however, had greatly stimulated communal tension as it aroused anxieties and ambitions among many communities by the prospect of their place in India’s future political set-up.

Thus “while the goal of achieving independence from British rule was never a point of disagreement, the distribution of political power between the Hindu and the Muslim communities in a future, free India became a continuous ‘apple of disord’.” We quoted the historian K.M. Pannikar as saying that “the introduction of the principle of elected representation in public institutions actively promoted the rising of communalism in India.”

The question that was left unaddressed in the above exposition was the logically prior one of identity. How and why did individuals in Indian society come to see themselves primarily as members of religious meta-communities, how did they understand the implication of their numerical strengths, and why did they fixate on religion as the most important marker of their identities?

In short, in trying to understand the history of politics in India we have arrived at the absolutely critical question of the dynamics of identity formation. It is critical because the understanding of identity was instrumental in both the modalities of electoral politics when it was introduced, and in how individuals reckoned their potential interests in the subsequent electoral game.

It is here that we make a surprising discovery and see how a seemingly innocuous administrative instrument like the census had such a profound influence on Indian history.

In his very insightful book (The Idea of India) Sunil Khilnani mentions that the British introduced the decennial census in India in a limited form in 1871. He describes how the census “expanded perceptions of the social scale of communities: individuals and groups living in far corners of the country could now conceive of themselves as being members of a single, large community”:

This made it possible for the first time to imagine a common nation of Indians. But the enumeration and classification of individuals into categories of caste and religion, and the introduction by the Raj of electorates divided along communal lines, also solidified exclusionary identities… Hindus, Muslims, Sikhs, caste groups, paradoxically began to emerge as collective actors and to conflict with one another in the city itself, the putative arena of modernity.

Khilnani elaborates on this point:

After all, before the nineteenth century, no residents of the subcontinent would have identified themselves as Indian. There existed intricate, ramified vocabularies of common understanding, which classified people by communities of lineage, locality and sect; but ‘Indian’ would not have figured amongst its terms. Subcontinental society was hardly static, yet most people never ventured beyond their own or neighbouring localities. They knew little about each other and were uninterested in learning more, preferring to remain distant strangers in a land peopled in their imagination by marvelous and absurd ‘others’.

This is followed by an explanation of the emergence of reified meta-identities:

The Muslims of British India did not form a monolithic community with a single ‘communal’ identity or interest any more than the Hindus did. Class and region divided as much as religion might unite, and beliefs about community and interest varied between provinces where Muslims were in a majority and those where they were not. (The terminology of ‘majority’ and ‘minority’ was itself an inescapable imposition of the political accountancy of the Raj.)

It is interesting to think what the nature of national politics in India might have been without the census. And how might national politics have evolved if the census had not collected information on the religious preferences of the inhabitants of India.

This is a sobering thought for social scientists who think of survey instruments as harmless devices available to collect all sorts of useful information about a given set of respondents not fully informed about how that information might subsequently be used.   

We will continue our investigations into the role of the census in British India in a subsequent post. 

Back to Main Page