A majority language and a minority cause: Bengali and the question of belonging in Pakistan

On 10 April 1952, Pakistan’s first Constituent Assembly voted 41 to 12 to postpone a decision on recognising Bengali as a state language. Seven weeks earlier, police had fired on students demonstrating for Bengali in Dhaka. The story of the language question is usually told from that perspective, as a quarrel between the eastern wing and the western one. Recognition in 1956, when the second Constituent Assembly wrote both Bengali and Urdu into the constitution, did not end it.

The Assembly’s own record carries a less familiar story. Most of the 12 members who opposed delay were Hindu representatives, drawn mainly from East Bengal. The Hindu opposition sat as the Pakistan National Congress in a chamber controlled by the Muslim League. For these members, delay was never only a question of timing. Bengali was the language of Pakistan’s majority and also a test of whether the state would make room for citizens who were not Muslim. Their support for Bengali repeatedly coincided with votes for broader representation and constitutional safeguards. Their speeches show the conditions under which such a case could be made.

Adaad’s analysis draws on Aiwan-e-Jamhoor’s digitised Hansard, which is forthcoming: a catalogue of 15 named divisions in the first Constituent Assembly between 1948 and 1953, containing 623 individual votes, and the complete debates on Bengali on three days, in February 1948, April 1952 and February 1956. A division is a recorded vote in which the printed record names every member on each side. The catalogue is not a census of every division the Assembly held, and a missing name is treated as missing rather than as an abstention or a vote against. The division lists do not label members by party or religion. Community and regional annotations were added afterwards from the Hansard speaker headings, the speeches and three historical accounts. Reading the divisions alongside the debates connects members’ stated arguments with their recorded choices.

Language as a test of belonging

Was the Bengali vote a one-off? Figure 1 takes the two sides from the 1952 language vote and tracks how each voted on six other questions between 1948 and 1953. These covered who could sit in the Assembly, whether the state would have a religious character, detention without trial, how East Bengal’s seats would be filled, and whether non-Muslim personal laws would be exempt from a requirement that laws conform to the Quran and Sunnah. If the same groups held together across such different issues, the Bengali vote was part of a larger disagreement about the kind of country Pakistan would be.

The broad alignment holds across all six rows. Among members recorded in both divisions, opponents of delay overwhelmingly favoured opening East Bengal’s general seats to every community, making detention reviewable and removing the sovereignty and sacred trust paragraph from the Objectives Resolution. Few supporters of delay backed those proposals. The direction reversed on the motion that narrowed who could sit in the Assembly: most supporters of delay favoured the restriction, while most opponents rejected it. The position favoured by the opponents of delay lost in the full House on all six occasions.

Who were these members? Ten of the 12 who opposed delay are identified in the source annotations as Hindu representatives for East Bengal or Sindh. Two were Muslim members from the western provinces, Shaukat Hyat Khan and Asadullah Jan Khan. Two Hindu members who supported delay, Akshay Kumar Das and Dhananjoy Roy, had voted with the opposition on the sovereignty paragraph in 1949. The votes suggest an alignment between government and opposition, with the language question falling largely along it. They cannot tell us whether community, party loyalty or conviction drew that line.

Early voting blocs in the Assembly

Figure 1 began with the Bengali vote and worked outward. Does the pattern also appear when that vote is excluded? Figure 2 uses twelve other recorded votes on seven sitting dates. It connects members whose voting agreement meets an 80 per cent threshold across at least three shared dates, giving each date equal weight. A community detection algorithm then identifies groups of members who regularly voted together, without being told how many groups to find.

To compare members fairly, the figure includes only those with recorded votes on at least four of the seven sitting dates. Thirty-six members qualify, including 29 of the 53 who voted on the language motion. The algorithm places them in groups of 13 and 23, with no links meeting the chosen threshold between the two groups. The same partition appears at agreement thresholds of 75, 80 and 90 per cent. Holding these 36 members fixed, omitting any one sitting also leaves the partition unchanged.

Only after the groups were formed was the Bengali vote added. Of the smaller group, ten members opposed postponement, two supported it and one was not recorded in that division. All 17 recorded language voters in the larger group supported delay; the other six were not recorded. Both groups drew most of their members from East Bengal, so the divide ran within East Bengal’s representation as well as across the Assembly. The smaller group was largely Hindu and the larger mostly Muslim.

The split over Bengali thus followed a broader voting alignment in the recorded divisions. Figures 1 and 2 draw on overlapping votes, so they show the same pattern in different ways rather than providing independent evidence for it.

Connecting language to other issues

The vote on 10 April tells us who opposed delay, not who actually wanted Bengali recognised. The motion was about timing, so a member could support Bengali and still vote for delay. To better understand the grounds on which members supported recognition for Bengali, we must consult their speeches. Ten speakers in that day’s debate argued clearly for recognition. These ten are identified by what they said: eight opposed delay, one supported it and one was not recorded in the language division. Figure 3 follows them through three votes: the language vote itself, and the later divisions on common seats and non-Muslim personal law.

Positions were read in the complete record of the debate, one position per speaker per proposition, and linked to the page on which it appears. A reported position, a procedural remark or a silence is not counted as an endorsement.

Of the ten clear advocates, six later voted for the personal law exemption, three voted against it and one was not recorded in that division. The one is A. K. Fazlul Huq, who told the House he wanted Bengali recognised but urged it to postpone rather than force a division it would lose. [2] His name appears in neither list on 10 April, nor in the two later divisions. The personal-law comparison therefore covers nine recorded voters among the ten advocates. The common-seats comparison covers eight, because Asadullah Jan Khan was also unlisted in that division.

The three who voted against the personal law exemption were Shaukat Hyat Khan, Asadullah Jan Khan and Nur Ahmed. Shaukat supported Bengali recognition while suggesting a common Quranic script for Bengali and Urdu. He argued that delay could foster the belief that West Pakistan sought to dominate East Bengal, endangering Pakistan’s solidarity. His speech accommodated Bengali within a national and Muslim cultural argument. [2]

Nur Ahmed shows this most clearly. He proposed the resolution to recognise Bengali, then voted to postpone it, and later voted against both common seats and the personal law exemption. In all three divisions he voted with the government majority. Yet he was among Bengali’s advocates. The votes shown in Figure 1 divided the Assembly into two camps. However, members on both sides argued for recognition of Bengali, even if they voted differently on timing and minority rights.

Language and national unity

What the minority members had to say to make that case is in the debates, and it begins before the 1952 vote. In February 1948, Dhirendra Nath Datta sought to add Bengali to the Assembly’s working languages, arguing that the language of Pakistan’s majority deserved a place at the centre. Liaquat Ali Khan defended Urdu through Pakistan’s Muslim nationhood and the need for a common language between its eastern and western parts. Sris Chandra Chattopadhyaya challenged the implications for non-Muslims’ right to help frame the constitution. He described Pakistan as a state belonging to its people, Muslim and non-Muslim alike. [1]

By April 1952 the ground had shifted, and Chattopadhyaya’s speech shows how. He warned that Hindus would again be blamed for the language agitation. He distanced his community from protest outside the legislature and described the Assembly as “our only forum”. He then cited Muslim editors who supported Bengali and invoked Muslim patronage of the language. Later in the speech he made the constraint explicit: “We are in an absolute minority there”. Participation in agitation, he feared, would expose his community to accusations and detention. [2]

“We are in an absolute minority there.” Sris Chandra Chattopadhyaya, 10 April 1952

This gives his appeal to Muslim history a political purpose as well as a cultural one. Establishing Bengali’s Muslim credentials answered the allegation that its recognition served an alien or hostile interest. It could help a minority representative secure a hearing for a demand that would widen his community’s place in the state. His speech supports that reading directly, although one speech cannot establish a general strategy among all minority members or separate calculation from conviction.

Kamini Kumar Datta made a related argument through Bengali’s shared Hindu and Muslim culture. He emphasised Muslim contributions to its development and rejected the accusation that the movement had been engineered from outside. He also grounded the demand in Pakistan’s federal character: recognition of both Bengali and Urdu would allow the two wings to understand each other. The claim offered a definition of national unity built through accommodation. [2]

Those arguments complicate a simple distinction between religious and secular language. References to Muslim heritage could defend a culturally plural Pakistan, answer suspicion or affirm a speaker’s own convictions. The policy being sought, the audience being addressed and the speaker’s position all matter. Minority advocacy could work within the majority’s account of Pakistan while pushing for stronger protections for minorities.

After recognition

By the second Constituent Assembly’s debate on 16 February 1956, recognition had brought another question into focus: how soon would Bengali and Urdu become languages of everyday state work? Figure 4 records the positions of the eight speakers who explicitly endorsed both as state languages. Five wanted English’s continuation shortened to ten years or less, two backed the 20-year provision and one expressed no clear timetable preference.

The second Assembly’s record for this sitting gives the result of each amendment but no division list, so the figure reports what speakers said rather than how members voted.

Mahmud Ali and Sheikh Mujibur Rahman also sought explicit support for Punjabi, Sindhi, Pushto and Baluchi. Hamidul Huq Choudhury defended the longer transition and argued that existing protections and provincial representation could support other languages. The adopted clause recognised Bengali and Urdu while allowing English to continue for 20 years, with provision for extension. Provinces could replace it sooner. Recognition itself was not postponed for 20 years. [3] The argument had moved from securing recognition to making it effective.

What the record suggests

Across these debates, demands for Bengali recognition repeatedly raised the question of who could claim Pakistan and on what terms. The divisions show a recurring alignment between resistance to delay and support for broader representation and constitutional safeguards, though it was not unanimous. The position favoured by opponents of delay lost each of the six votes examined here. The speeches reveal how minority representatives argued for a place in a state whose national identity was being defined in Muslim terms.

For the minority advocates discussed here, Bengali’s recognition was bound up with their own place in Pakistan. Chattopadhyaya and Kamini Kumar Datta defended the language through its Muslim history and shared culture, making a case for accommodation in terms the Assembly’s majority could recognise. To argue that Bengali belonged to Pakistan was also to claim their own right to shape it.

Zoya Sameen is an Assistant Professor of History in the Faculty of Arts and Sciences at the Aga Khan University. Her work examines gender, empire and decoloniality in modern South Asia and the British Empire, and she is working on history, politics and pedagogical nationalism in Pakistan. She received her PhD in History from the University of Chicago, where she taught for two years before joining AKU.