The States Reorganisation Act 1956 is usually described as the statute that gave India linguistic states, and the description is wrong in the one way that matters. No government of India ever adopted language as the principle on which the internal map would be drawn. Every official body that examined the question before 1956 advised against doing so, and the commission whose work the Act carried into law said in terms that language could be one consideration among several and the sole consideration under no circumstances. The map that took effect on the first of November 1956 looks linguistic because the claims that could no longer be resisted were linguistic claims, pressed by movements that had made particular boundaries politically unavoidable. The Act is therefore best read not as the application of a principle but as a settlement: a negotiated resting point between a doctrine the state refused to endorse and a set of demands it could no longer refuse, reached by applying four tests of which language was only the second.

That distinction explains everything about the settlement that otherwise looks inconsistent. It explains why two of the largest linguistic questions in the country, Bombay and Punjab, were left unsolved in 1956 and had to be settled by separate Acts in 1960 and 1966. It explains why a Marathi-speaking region, Vidarbha, was denied the separate state the Commission had recommended for it, and why a Telugu-speaking region, Telangana, was merged into a larger Telugu state against the Commission’s preference for keeping it apart. It explains why the Hindi-speaking belt was left as a handful of very large states rather than redrawn, and why the map contains units, such as the bilingual Bombay of 1956, that no theory of language would have produced. A principle applied consistently does not leave its two hardest cases for later. An accommodation does, and the history of the Act is the history of how that accommodation was reached, what it cost, and what it left behind.

The map the republic inherited

The internal map that the Constitution of 1950 recorded was not drawn by anybody for the purposes of the republic. It was the residue of two separate processes of accumulation. The provinces of British India had been assembled over a century and a half for revenue, policing and military convenience, and their boundaries reflected the order in which territory had been acquired rather than any quality of the people inside them. Madras Presidency held Tamil, Telugu, Malayalam and Kannada speakers inside a single administration because the East India Company had acquired their districts at different times for different reasons. Bombay Presidency held Marathi, Gujarati and Kannada speakers for the same accidental reason. Alongside the provinces stood the princely states, more than five hundred of them at the transfer of power, which acceded to India in 1947 and 1948 and were then grouped, merged or kept intact in a hurry, in a process directed by Sardar Vallabhbhai Patel and V. P. Menon that was concerned with accession and viability rather than with language or identity.

Why was the internal map of 1950 not drawn by language?

The map of 1950 was inherited, not designed. Its units were the provinces the British had assembled for revenue and control and the princely states that had acceded in 1947 and 1948, grouped by the Constitution into four classes. Nobody chose those boundaries for language, administration or identity.

The four classes into which the First Schedule divided the units tell the story of that inheritance plainly. The Part A States were the former governors’ provinces, nine of them: Assam, Bihar, Bombay, Madhya Pradesh, Madras, Orissa, Punjab, Uttar Pradesh and West Bengal. The Part B States were the former princely states or unions of princely states, administered by a Rajpramukh, and they included Hyderabad, Jammu and Kashmir, Madhya Bharat, Mysore, the Patiala and East Punjab States Union, Rajasthan, Saurashtra and Travancore-Cochin. The Part C States were the former chief commissioners’ provinces and smaller units administered more closely by the Centre, among them Ajmer, Bhopal, Coorg, Delhi, Himachal Pradesh, Manipur, Tripura and Vindhya Pradesh. Part D was a single territory, the Andaman and Nicobar Islands, administered by a Lieutenant Governor. The classification was a ranking by origin and by the degree of central control, not a scheme of government designed for a federation, and nearly everybody who worked with it regarded it as provisional.

Against that inheritance stood a promise the national movement had made to itself long before it held power. At its Nagpur session in 1920 the Indian National Congress reorganised its own provincial committees on linguistic lines, creating a structure of Congress provinces that followed language rather than the administrative map, and it repeated the commitment to redistribution of provinces on a linguistic basis in later resolutions and in the Nehru Report of 1928. The logic was democratic and practical: politics conducted in the language of the governed would draw ordinary people into public life, and administration and education in a shared language would work better than administration through a tongue the administered did not speak. The commitment was reaffirmed in the Congress election manifesto of 1945 and 1946. What changed after independence was not the attractiveness of the idea but its perceived danger. A country that had just been partitioned on one line of identity looked at the proposal to redraw itself on another line of identity and hesitated, and the hesitation was shared by the very leaders, Jawaharlal Nehru above all, who had made the promise in opposition.

The experiments before independence

The question had been tried, in smaller compass, before the republic inherited it. The British, for all that their provinces were accumulations, had twice drawn a boundary with language in view, and both experiments were cited before every commission that followed. In 1912 the partition of Bengal was annulled and Bihar and Orissa were separated from the Bengal Presidency on grounds that mixed administration with the distinctness of the Oriya and Hindi speaking populations. In 1936 Orissa became a province in its own right, the first Indian unit formed avowedly on a linguistic basis, after a movement led by Madhusudan Das that had pressed the claim of the Oriya speaking tracts scattered among Bengal, Madras and the Central Provinces for a generation. Sindh was separated from Bombay in the same year, on grounds in which language and community were entangled. The precedents cut in both directions in the argument of the nineteen fifties. The advocates of linguistic states pointed to Orissa and asked why a principle good enough for one province in 1936 was dangerous for the whole country in 1956. The opponents answered that Orissa had been a small and homogeneous tract whose separation injured nobody, and that the scale of the general demand, touching every province and every language at once, was a different thing in kind. Both were right about the precedent, which is why it settled nothing.

The Congress’s own internal map was the other precedent, and it was harder for the party’s leadership to escape. When the Congress reorganised itself at Nagpur in 1920 it divided its organisation into provincial committees drawn by language, creating in time more than twenty Congress provinces where the Raj had a third as many administrations, and it conducted much of its provincial business in the provincial language. For thirty years the national movement had therefore governed itself, in its internal life, on the very principle its government declined to apply to the state. The JVP Committee’s discomfort, and the careful wording of its escape clause about insistent public sentiment, are unintelligible without that history: the men of 1949 were not confronting a novelty but retreating from a commitment, and they knew that the movements of the nineteen fifties would quote the commitment back at them, as in fact the movements did, in memorandum after memorandum before the Commission.

Two refusals: the Dar Commission and the JVP Committee

The first formal examination of the question after independence produced a refusal. On the seventeenth of June 1948 Rajendra Prasad, as President of the Constituent Assembly, set up the Linguistic Provinces Commission to examine whether provinces should be formed on a linguistic basis. Its chairman was S. K. Dar, a retired judge of the Allahabad High Court, and its other members were Jagat Narain Lal and Panna Lall. The Commission reported on the tenth of December 1948, and its conclusion was unqualified: the formation of provinces on exclusively or even mainly linguistic considerations was not in the larger interests of the Indian nation. It recommended instead that the provinces whose reorganisation could not wait, Madras, Bombay and the Central Provinces and Berar, be redrawn primarily on the basis of geographical contiguity, financial self-sufficiency and ease of administration. Language, in its view, was a divisive force in a country that needed consolidation, and to build it into the structure of the state would be to institutionalise the division.

The refusal did not stand alone, and it did not go unchallenged from inside the constitutional process itself. B. R. Ambedkar submitted a memorandum to the Dar Commission in October 1948 supporting the formation of linguistic provinces, proposing a Marathi-speaking Maharashtra with Bombay as its capital, and answering the unity objection with a device of his own: the official language of every province, he suggested, should be the official language of the central government, so that linguistic states would conduct their politics in their own tongue while remaining bound into a single administrative language. The contrary argument was put at its strongest by K. M. Munshi, who warned that the political ambition of a linguistic group could only be satisfied by the exclusion and discrimination of other linguistic groups within the area, and that no safeguards and no fundamental rights could protect a minority against the subtle psychological exclusion that linguism implied. Both arguments were made before the Commission reported, and the Commission sided, in substance, with Munshi.

The Congress then examined the question through its own senior leadership. At its Jaipur session in December 1948 the party appointed a committee of three, Jawaharlal Nehru, Vallabhbhai Patel and the Congress president Pattabhi Sitaramayya, known from their initials as the JVP Committee. Its report of the first of April 1949 reached substantially the same conclusion as the Dar Commission: the time was not suitable for the formation of new provinces, and the priority of the state had to be security, unity and economic development. Yet the JVP report contained a sentence that the Dar report did not, and it is the sentence on which the whole later history turns. If public sentiment was insistent and overwhelming, the committee said, then they, as democrats, had to submit to it, though subject to certain conditions in regard to the good of India as a whole. The door was formally closed and deliberately left unlocked. The leadership had refused the principle and conceded the exception, and within four years the exception had been invoked in the one place where sentiment was strong enough to meet its terms.

It is worth pausing on the character of the two refusals, because they were not the work of men hostile to democracy or indifferent to language. Dar was a judge of distinction and the JVP Committee comprised the three most authoritative figures in the government and the party, two of whom, Nehru and Sitaramayya, were themselves men of the linguistic provinces tradition, Sitaramayya indeed a lifelong advocate of the Andhra cause. Their refusal was a judgment about sequence: that a Union still fighting for food, for the integration of its territory and for the survival of its constitutional experiment should not simultaneously undertake the redrawing of every internal boundary, and that the claims, however just, would keep. The movements’ answer, which events sustained, was that claims of that kind do not keep; they compound. The whole tragedy of the passage, if the word is not too large for an exercise that ended in legislation rather than in ruin, is that both judgments were defensible and both were honestly held, and that the settlement of 1956 had to be built, in the end, out of the materials of the defeated argument as much as out of those of the victorious one.

The years of postponement, 1949 to 1952

The interval between the JVP report and the death of Sreeramulu is sometimes written as though the question slept, and the record says otherwise. The government of the day was occupied with the rehabilitation of refugees, with food scarcity, with the war and the ceasefire in Kashmir, with the integration of Hyderabad by police action in 1948 and with the preparation of the first general election, and it treated the reorganisation question as one that a state carrying those burdens could honourably postpone. The movements treated the same years as an organising opportunity. The Andhra demand acquired a province wide organisation, and in 1951 Swami Sitaram undertook a fast for the Andhra cause that ran for five weeks before it was ended on the appeal of Vinoba Bhave, an episode that demonstrated both the readiness of the technique of the fast and the reluctance, at that stage, of its leaders to press it to the death. The first general election of 1951 and 1952 then changed the arithmetic of refusal: the Congress was returned in Madras, but the strength of the regional sentiment in the Telugu districts was visible in the results, and the men who had signed the JVP report’s escape clause could no longer be certain that the sentiment it described as insistent and overwhelming had not already arrived. When Sreeramulu began his fast in October 1952 he was therefore not creating a crisis from nothing. He was calling in, at the point of maximum effect, a concession the leadership had drafted for itself three years before.

Andhra breaks the dam

The Telugu-speaking districts of Madras Presidency, and then of Madras State, had demanded a separate Andhra province since before independence, and the demand had been the most organised of all the linguistic claims. After the two refusals it did not subside. In October 1952 Potti Sreeramulu, a Gandhian worker of long standing in the cause, began a fast for a separate Andhra state. He died on the fifteenth of December 1952 after fifty six days of fasting, and his death set off rioting across the Telugu districts in which government property was attacked and the authority of the state visibly collapsed. Within days the Prime Minister announced in Parliament that a separate Andhra State would be formed. The state came into being on the first of October 1953, carved out of the Telugu-speaking districts of Madras with its capital at Kurnool, and it was the first state in independent India created on a linguistic basis.

The significance of Andhra was not the boundary itself but the demonstration it provided. The JVP formula had said that insistent and overwhelming sentiment would have to be submitted to, and Andhra showed what insistence looked like in practice and what it could obtain. Every other linguistic movement in the country drew the conclusion. The demand for Samyukta Maharashtra, a single Marathi-speaking state including Bombay city, acquired an organised movement and a martyrology of its own. The Mahagujarat movement pressed the corresponding Gujarati claim. The Akali Dal under Master Tara Singh pressed the demand for a Punjabi Suba, a Punjabi-speaking state, which was entangled from the beginning with Sikh religious identity in a way the other claims were not. The Malayalam-speaking districts looked to a united Kerala. The Kannada-speaking districts, scattered across four administrations, demanded unification. A general reorganisation had become unavoidable, not because the government’s mind had changed on the principle but because conceding claims one at a time, each under the pressure of agitation, was visibly the worst method of deciding them. The mechanism available for conceding a claim was ordinary legislation by Parliament under Article 3 of the Constitution, with an accompanying amendment for the constitutional changes a general settlement required; that mechanism, and the question whether a state must consent to its own redrawing, is the subject of the complete guide to state reorganisation in India, and what matters for this article is the use made of it in 1956.

The escalation was general, and it was not confined to the south. In Punjab the Akali Dal opened a morcha in 1955 in defiance of the provincial government’s prohibition on the raising of the Punjabi Suba slogan, and its volunteers courted arrest in numbers that made the prohibition unenforceable and the demand impossible to treat as a fringe agitation. In the west the Samyukta Maharashtra movement organised the Marathi districts across three states into a single campaign with a single demand, and the Mahagujarat movement answered it in kind. In the Kannada country the unification movement, which had worked by conference and resolution for a generation, acquired the urgency of the neighbouring examples. The Commission conducted its inquiry in the middle of this escalation, and the movements understood its appointment exactly as the JVP formula had taught them to understand it: the state had conceded that insistent and overwhelming sentiment would prevail, a commission had been appointed to measure the insistence, and the rational course for a movement was therefore to be as insistent, and as visibly overwhelming, as it could make itself. The pressure that produced the settlement was not an interruption of the Commission’s work. It was, by the logic of 1949, a part of the evidence.

Sreeramulu’s fast did not occur in a vacuum, and the government’s capitulation cannot be understood apart from the condition of Madras politics at the moment of his death. The state government of the day, under C. Rajagopalachari, had opposed the separation, and the demand had become entangled with the question of Madras city itself, which the Andhra claim included and the Tamil country refused to yield; the eventual settlement of 1953 gave Andhra neither the city nor any share in it, and the capital question was answered instead at Kurnool. The riots that followed the faster’s death were directed in part at that refusal, and the announcement Nehru made within days, that a separate Andhra State would be formed from the Telugu districts, was drafted to end a breakdown of order as much as to satisfy a claim of principle. The episode fixed the pattern the Commission was appointed to break: a demand refused for years on stated grounds, an agitation that raised the cost of refusal above the cost of concession, and a concession granted in a form narrower than the demand and justified in language that conceded nothing about the principle involved.

The pilot state: Andhra between 1953 and 1956

The three years of Andhra State’s separate existence functioned, for the rest of the country, as a pilot experiment whose results were read in opposite senses. The state had been formed in haste, with a temporary capital at Kurnool, an inherited share of the services and assets of Madras still being disentangled, and a politics organised around the expectation that it was the first instalment of a Telugu union that would one day include Telangana. Its advocates pointed to an administration and a legislature working in Telugu, to the energy of a government that felt itself the expression of a people’s claim, and to the absence of the catastrophe the opponents of linguistic states had predicted; the sky had not fallen in Kurnool. Its critics pointed to the state’s finances, strained by the costs of a new capital and a new apparatus, and to the lesson they drew from its creation, that the concession had been extracted by a fast and by riots rather than granted on the merits, and had therefore taught every other movement the technique that worked. The Commission’s report, written while the experiment was two years old, treated Andhra as a fact and not as a precedent to be either extended or disowned, which was the only treatment available to it. What the pilot settled, beyond argument, was the narrow point the Dar Commission had contested: a state could be formed on language, severed from a composite province, and governed, without the Union suffering any visible strain. The argument after 1953 was never again about whether such a thing was possible. It was about where else, and on what conditions, and at whose expense.

A commission asked to start again

The government’s answer to the pressure was to return the question to a commission, this time one charged with examining the whole map rather than the single question whether language should govern it. The States Reorganisation Commission was appointed at the end of December 1953. Its terms required it to examine the question of reorganisation objectively and dispassionately, having regard to the welfare of each constituent unit and of the nation as a whole, and to make recommendations for the readjustment of state boundaries and the consequential arrangements. Those terms were themselves a decision: the Commission was not asked whether linguistic states were desirable but what the map should be, and the criteria by which it should judge were left, in the first instance, to the Commission itself.

Who sat on the States Reorganisation Commission?

Three men sat on the Commission. Its chairman was Fazl Ali, a former Chief Justice of India. With him sat K. M. Panikkar, historian, administrator and diplomat, and H. N. Kunzru, a veteran parliamentarian of liberal politics. The government appointed the Commission in December 1953 and it reported in September 1955.

The composition mattered to the kind of answer the Commission gave. Fazl Ali brought a judge’s instinct for testing a claim against stated criteria and for refusing a claim that the criteria did not support, however loudly it was put. Panikkar brought a historian’s sense of the regions of India as old cultural and political formations that predated both the British provinces and the language movements, and a diplomat’s attention to the security of frontiers. Kunzru brought the liberal constitutionalist’s suspicion of majorities, linguistic as much as religious, and a corresponding attention to the minorities any boundary would create. The Commission toured the country, received memoranda from governments, parties, movements and individuals, and heard the rival claims made in person, in an atmosphere its report does not disguise: it recorded its dismay at the intensity of the feeling it encountered and at the readiness of neighbouring language groups to treat a shared district as a prize of war. It reported on the thirtieth of September 1955.

What the States Reorganisation Act 1956 was built on: the four considerations

The Commission’s answer to its terms began with a refusal. It declined to accept the theory of one language, one state, and it declined with reasons that went to the structure of the country rather than to the tactics of the moment. India has more languages than any map could give a state to; several languages are spoken across areas so wide that a single state for each would create units of unmanageable size; several regions are mixed in proportions that no boundary can unmix; and the Hindi-speaking area alone, treated as one language unit, would have produced a single state holding a share of the Union’s population large enough to distort the federation itself. A doctrine that could not be applied to the largest case could not be the doctrine. Language was admitted as a consideration of the first importance, because a state whose administration, courts, schools and politics run in the language of the governed works differently in kind from one that does not; but it was admitted as one factor in a balance, and the Commission named the balance.

Did the Commission accept language as the basis for redrawing the map?

It accepted language as one consideration and refused it as the only one. The Commission said expressly that the theory of one language, one state could not be accepted, and it placed the unity and security of India above language in its own ordering of the four considerations it applied.

The four considerations, in the order the Commission stated them, were these. The first was the preservation and strengthening of the unity and security of India, which the Commission treated as the governing test against which every proposal had to be checked. The second was linguistic and cultural homogeneity. The third was financial, economic and administrative considerations, which covered the viability of the proposed unit, its resources, its size in relation to the work of government, and the cost and disruption of change. The fourth was planning and the promotion of the welfare of the people in each state as well as of the nation as a whole, which tied the map to the development plans then being framed and asked whether a boundary helped or hindered the work of raising standards of life. The ordering was the argument. A commission that had listed language first and unity as a qualification would have written a different report and produced a different map; this Commission listed unity first and language second, and treated the third and fourth tests as capable of defeating a linguistic claim that passed the second.

Each test did real work in the recommendations, and the worked examples are the quickest way to see that the criteria were not decorative. Unity and security bore hardest on the frontier and on Punjab, where the Commission refused to divide a border state on a line that ran close to a religious division so soon after Partition. Linguistic homogeneity produced the clearest gains of the settlement, in Kerala, in the enlarged Mysore and in the consolidation of the Hindi-speaking and other regions whose claims were not contested. Financial and administrative viability defeated claims for units the Commission thought too small or too poor to carry the burdens of statehood, and supported the case for a separate Vidarbha, whose resources and administrative history the Commission found sufficient. The planning test supported larger and more coherent units for the great river valley and development projects of the period, and told against fragmenting regions whose development depended on shared works. The artifact below compresses that working into one view.

What the Commission was asked and what it answered

Criterion the Commission applied	A boundary decided by reference to it	A linguistic claim it declined on that ground	What happened to that claim later
Preservation and strengthening of the unity and security of India	Punjab kept as a single bilingual state on the northern frontier	The demand for a Punjabi-speaking state, refused as premature and divisive in a border region	Punjab was reorganised in 1966 into Punjab and Haryana, with Chandigarh as a Union territory
Linguistic and cultural homogeneity	Kerala formed by joining Travancore-Cochin to the Malabar district of Madras	The claim that Bombay city belonged to one language group alone, refused by keeping the city in a bilingual state	Bombay State was divided in 1960 into Maharashtra, with the city, and Gujarat
Financial, economic and administrative considerations	Vidarbha recommended as a separate state on the strength of its viability and administrative history	The claim of small units such as Coorg to separate statehood, refused and resolved by merger into Mysore	Vidarbha was placed in Bombay in 1956 and passed to Maharashtra in 1960, where the claim for separation continued
Planning and promotion of the welfare of the people	Madhya Pradesh consolidated as one large unit for coherent development	The claim that Telangana should merge at once into a single Telugu state, deferred in favour of a safeguarded separate existence	Telangana was merged into Andhra Pradesh in 1956 under safeguards, and became a separate state in 2014

The table also exposes the limits of the exercise, which the Commission itself acknowledged. No arrangement of the boundaries could produce units that were homogeneous, because the populations were intermixed at every scale at which a boundary could be drawn. The real question before the Commission was therefore never whether linguistic minorities would exist inside the new states; it was which minorities would be created, how large they would be, and what safeguards would stand between them and the majority whose language the state adopted. That is why the report’s recommendations on safeguards, on the protection of linguistic minorities in education and administration, and on the treatment of disputed border districts carry as much weight in the document as the boundary lines themselves, and why the constitutional settlement that followed took those safeguards up.

The criteria in their intellectual setting

The four criteria did not descend from theory; each of them was the deposit of a particular experience of the decade. The unity criterion was the deposit of Partition, which every member of the Commission had lived through as a serving official or public man, and which had demonstrated that a line drawn on a map of India could become a line of population movement, massacre and permanent military confrontation within a single season. When the Commission put unity and security first it was not uttering a piety; it was recording a judgment about the age, that the Indian Union was a recent and reversible achievement, and that internal boundaries were to be judged first by whether they strengthened or weakened the thing that held the country together. The homogeneity criterion was the deposit of the national movement’s own experience, of mass politics conducted in the vernacular, of the evident failure of the composite provinces to give their peoples a politics in their own tongue, and of the administrative truth that a district officer, a school and a court that share the language of the countryside govern better than those that do not. The viability criterion was the deposit of the princely integration, in which hundreds of units too small to be states had had to be merged, and of the fiscal weakness of several Part C states which the Centre carried on its own budget. The planning criterion was the deposit of the moment itself: the Commission reported in the year the Second Five Year Plan was being framed, when the great river valley projects, the steel towns and the community development programme made it natural to ask of any proposed boundary what it did to the working of a development plan whose units of operation were the states. A reader who knows where each test came from can predict, for any disputed district of the nineteen fifties, how the Commission would argue, and the predictability was deliberate: the Commission wanted its reasoning to be checkable against its premises, so that even those who lost by it could see the ground of their loss.

One consequence of that design deserves emphasis, because it distinguishes the Indian exercise from the boundary making that preceded it in 1947. The Radcliffe commissions that divided Bengal and Punjab had worked in five weeks, in secret, without stated criteria that the public could test, and their awards were published after the transfer of power they were meant to precede. The States Reorganisation Commission worked for twenty one months, in public, on stated criteria, publishing a report that reasoned its way district by district and left the losing side in possession of the argument it had lost to. The settlement of 1956 was contested, sometimes violently, but it was not mysterious, and the legitimacy it acquired, the sense that the questions had at least been heard by a body that answered them with reasons, is part of why the resulting map, for all its deferred questions, did not have to be defended by force.

The evidence of language: the census problem

Underneath every argument the Commission heard lay a difficulty of evidence that its report handles with unusual candour: nobody could say with precision how many people spoke what. The census of 1951 had enumerated mother tongues, but its classifications grouped a large number of distinct mother tongues as dialects or variants of Hindi, so that the Hindi total on which the unity of the northern belt was argued rested in part on a decision of the census office rather than on the speech of the people counted; the speakers of Maithili, of the Rajasthani tongues and of a dozen other returns had their own view of the matter, and some of those claims, Maithili’s above all, have continued into the politics of recognition ever since. In the mixed districts the figures were not merely imprecise but interested: a mother tongue return was the cheapest vote a villager could cast in the boundary contest, and movements on both sides of every disputed line instructed their followers accordingly. The Commission’s response was to refuse to treat any single return as decisive and to read the figures together with the language of the courts, the schools and the revenue records of the district, a method that made its district level reasoning slower and better. The episode carries a permanent lesson for the subject: language statistics in a country where language is a political claim are themselves political acts, and any future reorganisation that treats the census as a neutral instrument will repeat the errors the Commission took such care to avoid.

The claims as their movements stated them

The Commission did not receive the linguistic claims as abstractions, and this article does not do the movements the disservice of stating their cases weakly. The claim for Samyukta Maharashtra, pressed by the Samyukta Maharashtra Samiti under leaders such as S. M. Joshi and Acharya Atre, was that the Marathi-speaking people, divided among Bombay State, Hyderabad and Madhya Pradesh, constituted one of the great language nations of India and were entitled to a single state whose capital could only be Bombay, a city whose working population, mills, docks and suburbs were in their majority Marathi speaking and whose separation from Maharashtra would leave the new state a body without its head. The corresponding Gujarati claim, pressed by the Mahagujarat movement under Indulal Yagnik, was that Gujarati speakers were entitled to a state of their own, that the commercial and financial life of Bombay had been built substantially by Gujarati capital and enterprise, and that to hand the city to a Maharashtra in which Gujaratis would be a small minority was to expose that enterprise to a politics that had already shown its temper in the riots of January 1956. Each movement regarded the other’s claim to the city as an attempt at annexation, and each could point to census figures for the city and its region that supported its own arithmetic, because the city’s population was mixed in proportions that allowed both cases to be made in good faith.

The Punjabi claim was stated differently, because it could not be separated from religion. The Akali Dal under Master Tara Singh demanded a Punjabi Suba, a state in which Punjabi, written in the Gurmukhi script, would be the language of administration, and in which Sikhs, though not a majority of the whole Punjab, would form a majority or near majority of the reduced unit. The opposing case, made by the Arya Samaj and by Hindu organisations in the state, and reflected in the language returns of the census, in which large numbers of Punjabi speaking Hindus returned Hindi as their mother tongue, was that the demand was communal in substance whatever its linguistic form, and that a state drawn to produce a Sikh majority was a religious state by another name. The province had already tried two devices short of division: the Sachar Formula of 1949, which divided the state into Punjabi and Hindi zones for the purposes of school instruction, and the regional formula of 1956, which the Akali Dal accepted at the time and under which the demand was for some years withdrawn. The Commission’s refusal of the Suba has to be read against that background: it was refusing a claim whose linguistic form it could verify and whose religious substance it believed it could also see.

The Telugu case contained its own internal opposition, which is why the Commission’s Telangana proposal took the form it did. The Vishalandhra movement demanded the union of all Telugu speakers in one state, and its case was the textbook case for linguistic reorganisation: one people, one language, divided between Madras and Hyderabad by the accidents of conquest. The Telangana answer, made by the region’s own leaders, was that the two Telugu regions were not one society: Telangana had lived under the Nizam’s administration, with its own land tenures, its own services recruited under domicile rules that protected local employment, and a level of development markedly below the coastal districts; a merger without safeguards would deliver its services, its colleges and its development funds to the more advanced region. The Gentlemen’s Agreement of February 1956, concluded between the leadership of the two regions, including the Chief Minister of Hyderabad, Burgula Ramakrishna Rao, and the Chief Minister of Andhra, was the attempt to answer that case in writing, with safeguards on employment, education, expenditure and regional oversight, and the subsequent history of the united state is in large part the history of whether those written answers were honoured.

The remaining claims were less contested and are easily forgotten beside the great controversies. The Kannada unification movement, older than independence and given literary form by Aluru Venkata Rao a generation earlier, demanded the gathering of the Kannada districts out of Bombay, Madras, Hyderabad and Mysore into a single state, and obtained substantially the whole of its claim. The Aikya Kerala movement demanded the union of Travancore-Cochin with Malabar and obtained it. The Oriya claim had been satisfied in part a generation before, Orissa having been formed as a province in 1936, the earliest Indian unit drawn with language in view. These successes are part of the record the critics and defenders of the settlement both use: they show the Commission granting linguistic claims where its other criteria did not resist, which is why the report cannot be described either as a linguistic settlement or as a refusal of one.

The proposals, unit by unit

Applying its four tests, the Commission proposed the abolition of the fourfold classification of Part A, Part B, Part C and Part D States and the reconstitution of India as sixteen states and three centrally administered territories. The sixteen states it proposed were Andhra, Assam, Bihar, Bombay, Hyderabad, Jammu and Kashmir, Kerala, Madhya Pradesh, Madras, Mysore, Orissa, Punjab, Rajasthan, Uttar Pradesh, Vidarbha and West Bengal. The three territories were Delhi, the Andaman and Nicobar Islands, and the Laccadive, Minicoy and Amindivi Islands. It recommended that the office of Rajpramukh be abolished and that the special agreements with the former princely states, survivals of accession, be brought to an end. The map it drew was therefore both less and more than the movements had demanded: less, because the two most contested linguistic claims were not granted; more, because it proposed two states, Vidarbha and Hyderabad, that the movements for linguistic unity opposed.

Bombay was the largest of the contested cases, and the Commission’s treatment of it shows the criteria in conflict and in balance. The Marathi claim to a state including Bombay city and the Gujarati claim to a state of their own were both linguistically well founded. The Commission still proposed a bilingual Bombay state, taking in the Marathi and Gujarati regions together with Saurashtra and Kutch, on the ground that the economic integration of the city with both its hinterlands, and the mixed character of the city itself, made a division at that stage a worse outcome than a shared state with safeguards. The reasoning was administrative and economic, the third and fourth tests overriding the second. It was also the recommendation least likely to survive contact with politics, because it required two organised movements to accept, as a settlement, the continuation of the very arrangement both existed to end.

Punjab was the second contested case, and there the first test governed. The demand for a Punjabi Suba had been pressed since before independence, and the Commission examined it and declined it, concluding that the proposed division lacked the general support a reorganisation required and that to carve a Punjabi-speaking state out of a bilingual border province, along a line that corresponded closely though not exactly with the line between Sikhs and Hindus, would endanger the unity the Commission was charged to protect. It proposed instead a single enlarged Punjab, absorbing the Patiala and East Punjab States Union, in which both Punjabi and Hindi would have their place. The refusal was the clearest instance in the report of the ordering of the criteria doing decisive work: the linguistic case for a Punjabi state was not fanciful, and it was refused anyway.

Two further proposals show the criteria working against linguistic unity rather than for it, and both matter for what followed. For the Telugu country the Commission proposed that the state of Andhra, formed in 1953, and the Telangana region of Hyderabad should not be merged at once. Hyderabad should continue as a separate state, with the possibility of union with Andhra after the general elections due in 1961 if the legislature of Hyderabad resolved in favour of it by a two thirds majority. Telangana’s leaders feared absorption into a larger unit in which their region, poorer and governed under a different land system, would be a permanent minority, and the Commission treated those fears as entitled to weight under the welfare test. For central India it proposed the opposite of merger: a separate state of Vidarbha, the Marathi-speaking eastern region of Madhya Pradesh with its capital at Nagpur, whose distinct administrative history under the old Central Provinces and whose resources the Commission found sufficient for statehood. Both proposals rested on the proposition Ambedkar had defended in his own writings on the question, that the correct maxim was one state, one language, and not one language, one state: a single language might properly be spoken in several states, and the multiplication of units within a language was no defect if the units were viable and their peoples willing.

Elsewhere the map was consolidation. Kerala was to be formed from Travancore-Cochin with the Malabar district and the Kasaragod area of Madras, giving the Malayalam-speaking people a single state. Mysore was to be enlarged into a united Karnataka in fact though not yet in name, gathering the Kannada-speaking districts of Bombay, Madras and Hyderabad and the small state of Coorg. Madhya Pradesh was to be consolidated from the old state of that name with Madhya Bharat, Bhopal and Vindhya Pradesh, less Vidarbha. Rajasthan was to absorb Ajmer. Saurashtra and Kutch were to go into Bombay. Assam, Bihar, Orissa, Uttar Pradesh and West Bengal were to continue with boundary adjustments, including the transfer of the Manbhum area question between Bihar and Bengal, which a separate measure later settled by moving the Purulia area to West Bengal. The Commission also proposed safeguards to accompany the whole scheme: facilities for education in the mother tongue, machinery for the protection of linguistic minorities, and arrangements for the disputed border areas whose assignment one side or the other would resent.

The eastern and northern units received shorter treatment in the report, and the brevity is itself informative. For the east the Commission recommended one significant transfer, the Purulia area of the Manbhum district from Bihar to West Bengal, on the ground of its Bengali speaking character and its contiguity, a recommendation Parliament carried out by a separate Act of the same year rather than in the principal Act. For Assam it proposed no division, leaving the hill districts within the state in arrangements whose inadequacy the following decade exposed. For Jammu and Kashmir it proposed nothing, treating the state’s position as governed by considerations outside its remit. For Delhi it recommended continuance under central administration, declining to make the capital the prize of the surrounding claims. The Commission’s silences were consistent with its criteria: it reorganised where claims had been made, tested and sustained, and it did not go looking for symmetry. A tidier commission would have rationalised the whole map. This one answered the questions put to it, and the unevenness of the result, meticulous in the west and south, cursory in the east and north, is the shape of the politics it was answering.

The largest language and the states that were not redrawn

One feature of the settlement is so familiar that it escapes notice: the Commission left the Hindi-speaking belt divided among four large states, Uttar Pradesh, Bihar, Madhya Pradesh and Rajasthan, and refused every proposal to gather them into one unit or to break them into smaller ones. A map genuinely drawn by the rule of one language, one state would have produced a Hindi state holding approaching two fifths of the population of the Union, a unit whose weight would have changed the character of the federation and whose government would have stood to the Centre in a relation no other state could match. The Commission’s refusal is the single clearest proof that language was not its rule. It treated the existing Hindi states as satisfactory units under its third and fourth tests, viable in size, coherent in administration, and serviceable for planning, and it saw no gain in disturbing them. Ambedkar, who favoured linguistic states more warmly than the Commission did, reached the opposite conclusion on this point and proposed in his writings on the question that Uttar Pradesh should in time be divided into three states, on the ground that its size alone made it a danger to the balance of the federation. The disagreement is instructive because it shows that the deepest question in the whole exercise was never language at all. It was size: how large a unit a federation can contain, and how small a unit can carry the burdens of statehood, and on that question the Commission’s answers, conservative in both directions, governed the map for half a century.

The smaller claims the Commission refused

The great controversies have absorbed the memory of the exercise, but the Commission also disposed of a file of smaller claims, and its reasoning on them shows the criteria operating where no great movement stood behind the file. The Jharkhand claim was the most substantial of them. The Jharkhand Party of Jaipal Singh, which had won the tribal districts of south Bihar decisively in the first general election, demanded a state for the tribal country of Chota Nagpur and the Santhal Parganas, extending into the adjoining districts of Orissa, Madhya Pradesh and West Bengal. The Commission examined the claim and refused it, on grounds that repay attention because they were not the grounds of the linguistic argument at all: the proposed state, it found, had no single language, its tribal population was a minority of the whole, and the case rested on a distinctness of the tribal population that a boundary drawn for other purposes could not capture. The refusal closed the question for the life of that generation of the movement, and its reopening, half a century later, was conducted on the different ground of region and development rather than of tribe or tongue.

Beside it stood the claims of regions that were regions and nothing more. Bundelkhand and Baghelkhand, in the country between Uttar Pradesh and Madhya Pradesh, put claims founded on history and neglect; they were refused under the third test, as units whose viability the Commission could not establish. Saurashtra and Kutch, each of which had entered the Union as a unit of its own, Saurashtra as a Part B state with its own ministry and legislature, were refused continuance and sent into Bombay, on the administrative ground that small coastal enclaves could not carry the apparatus of statehood within a reorganised west. The hill districts of Uttar Pradesh, whose separation was demanded in petitions the Commission received, were left within their parent state, a deferral that lasted until the creation of Uttarakhand in 2000. Taken together the refusals describe the Commission’s sense of its mandate as accurately as its grants. It was not holding a general auction of statehood in which every organised claim was entitled to a hearing on equal terms. It was answering the claims that the politics of the decade had made unavoidable, testing them against its four considerations, and refusing, with reasons, those that failed the tests or had not yet acquired the political weight that the JVP formula, nine years before, had made the true qualification for a hearing.

The districts no line could settle

Between the grand questions of Bombay and Punjab lay a hundred smaller questions of districts, taluks and towns, and the Commission’s treatment of them is where its method can be watched at the closest range. A boundary commission working by language has to decide, for each disputed area, which language predominates, and the census returns on which that decision rests were themselves contested, because in the mixed districts the return of a mother tongue had become a political act. In Punjab the returns were distorted by the communal contest over whether Punjabi or Hindi would be recorded. In the border country between the Marathi and Kannada districts, in Belgaum above all, the rival movements organised the recording of language as they organised everything else. The Commission knew the figures were loaded and proceeded by weighing them against the geography of the district, the language of its administration and schools, the wishes expressed by its representatives, and the inconvenience of enclaves and severed communications. Its decisions satisfied the test it had set itself, a balance of considerations, and frequently satisfied neither claimant.

Belgaum is the standing example. The district, with its city, was claimed for the Kannada state on the ground of its district level figures and its administrative history, and for Bombay on the ground of the Marathi character of the city and of many of its villages. The Commission assigned Belgaum to Mysore, and the assignment was enacted in 1956 and survived the division of Bombay in 1960, so that the claim passed to Maharashtra against Karnataka and remained alive for decades, consuming a further commission of inquiry under Justice Mahajan in the nineteen sixties, whose recommendations Maharashtra declined to accept. Kasaragod, at the northern end of the Malabar coast, was assigned the other way, to Kerala, against the claims of the Kannada and Tulu country to its north. The Abu Road area, at the meeting point of Gujarat and Rajasthan, was assigned to Rajasthan. The Kolar district, claimed in the name of its Tamil and Telugu speaking miners, remained in Mysore. None of these decisions was irrational, and none of them closed its question, because a decision by balance can always be reargued by weight. The Commission understood this and said, in substance, that finality in such matters is not obtained by the quality of the reasoning but by the passage of time and the growth of interests around the line as drawn. The observation is the most quoted sentence in the report among those who administer boundaries, and the least quoted among those who campaign about them.

From report to Act: the year of reversal

The government accepted the report in broad outline and then, in the handling of its most contested recommendation, produced the episode that best illustrates the difference between a commission’s settlement and a legislature’s. The Commission had proposed a bilingual Bombay. The government’s first decisions, announced in the months after the report, proposed instead to divide the west into three units: a Marathi state, a Gujarati state, and the city of Bombay constituted as a separate centrally administered unit. The formula satisfied neither movement. In Bombay city it provoked rioting in January 1956 in which people were killed and the authority of the state government was openly defied, and across Maharashtra it converted a movement for a state into a movement against a perceived theft of the city. The government reversed itself. The Bill that was eventually enacted proposed a single bilingual Bombay larger than the one the Commission had drawn, enlarged by Marathwada from Hyderabad and by Vidarbha from Madhya Pradesh, so that the Commission’s separate Vidarbha disappeared into the very state whose division it had been meant to precede, and the question of Bombay was postponed rather than settled.

On Telangana the government moved in the opposite direction, accepting sooner what the Commission had proposed to defer. The leaders of Andhra and of Telangana met in Delhi in February 1956 and concluded the agreement known as the Gentlemen’s Agreement, under which Telangana would merge into a single Telugu state at once, with safeguards for the region written into the settlement: protections for employment and education, a share in development expenditure, and a regional committee of the state legislature to watch over the region’s interests. On the strength of that agreement the enacted map merged Telangana into Andhra Pradesh in 1956, discarding the Commission’s device of a separate Hyderabad state with a later option of union. The two reversals are mirror images. On Bombay the legislature was less willing than the Commission to hold two language groups in one state; on Telangana it was less willing to keep one language group in two states. In both cases the legislature’s instinct ran toward the immediate political settlement and away from the Commission’s staged and safeguarded solutions, and in both cases the question returned.

The Bill itself, the States Reorganisation Bill, was piloted through Parliament by the Home Minister, Govind Ballabh Pant, in the summer of 1956, and the debate on it was among the longest and most closely followed of the decade, because the members were deciding the boundaries of the states from which many of them came. The gap between the Bill as introduced and the Act as passed was narrower than the gap between the report and the Bill: the government had made its concessions before introduction, and Parliament’s work was largely on the details of boundaries, the allocation of services and property, and the safeguards. The Act received assent on the thirty first of August 1956. It is Act 37 of 1956, and it came into force, with the constitutional amendment that accompanied it, on the first of November 1956.

The passage: views of the assemblies and the debate in Parliament

The procedure by which the settlement was enacted deserves a paragraph of its own, because it illustrates the constitutional character of what was done. Under the proviso to Article 3, a Bill that affects the boundaries of a state may not be introduced in Parliament except on the recommendation of the President, and it must first be referred by the President to the legislature of the affected state for expressing its views within the period allowed. The states’ views are sought; they do not bind. The States Reorganisation Bill, affecting as it did every state in the Union, was referred to the state legislatures, and through the first half of 1956 the assemblies of India debated the redrawing of their own boundaries, several of them in an atmosphere of demonstrations outside the chamber. The debates in the Bombay legislature, where the members for the Marathi and Gujarati districts confronted one another over the fate of the city, and in the Hyderabad legislature, where the merger of Telangana was argued on the safeguards of the Gentlemen’s Agreement, were the point at which the settlement ceased to be a commission’s scheme and became a set of political commitments made in public by named men.

In Parliament the Bill was debated through the summer of 1956, and the amendments moved to it trace the outline of every question the government had decided against the Commission or against the movements. Amendments to constitute Vidarbha a separate state were moved and defeated. Amendments to preserve Hyderabad as a separate state, or to postpone the merger of Telangana until the electorate of the region had been consulted again, were moved and defeated. Amendments on the status of Bombay city, the question on which the government had already reversed itself before introduction, were argued at length. The Home Minister’s defence of the Bill, maintained through the debate, was that the House was being asked to enact a settlement and not a selection: that the recommendations of the Commission and the modifications the government had made to them formed a balance in which each part depended on the others, and that to amend the Bill in favour of any single claim was to reopen every claim. It is the argument every government makes when it carries a package, and in 1956 it was also true, because the concessions and refusals had been calibrated against one another and the calibration is visible in the report itself. The accompanying amendment of the Constitution, the Seventh Amendment Bill, travelled the harder road of Article 368, requiring the special majority in each House and, because it altered the representation of states in Parliament among other matters within the entrenched list, ratification by the legislatures of not less than half the states, which was obtained in time for the appointed day. The two instruments, the ordinary Act and the amendment, received assent and came into force together, and an exercise that had begun with a fast in Andhra ended as the largest single alteration of the Indian Union ever effected by legislation.

The Act and the map of 1 November 1956

The States Reorganisation Act, 1956 did the mechanical work of the settlement. It defined the territories of the new states and territories, transferred territory between existing units, merged whole states into their neighbours, created the new category of Union territory in place of the old Part C and Part D classes, adapted the laws in force in the transferred areas so that a change of state did not produce a legal vacuum, allocated the services, property and liabilities of the predecessor units, and provided for the representation of the new units in Parliament and in the state legislatures until the next readjustment of seats. It also established, in its own text, the five zonal councils, for the northern, central, eastern, western and southern zones, as standing machinery through which the states of each zone and the Centre could consult on matters of common interest, from economic and social planning to border disputes and the protection of linguistic minorities. The councils were the Act’s answer to a fear its authors took seriously: that states drawn more closely around language would cooperate less readily, and that the Union needed permanent forums in which the cooperation the map had made harder could be practised.

How many states and union territories were there on 1 November 1956?

Fourteen states and six union territories. The fourteen were Andhra Pradesh, Assam, Bihar, Bombay, Jammu and Kashmir, Kerala, Madhya Pradesh, Madras, Mysore, Orissa, Punjab, Rajasthan, Uttar Pradesh and West Bengal. The six territories were Delhi, Himachal Pradesh, Manipur, Tripura, the Andaman and Nicobar Islands and the Laccadive group.

The way each state was assembled shows how much of the Act was consolidation of the princely and provincial inheritance rather than fresh creation. Andhra Pradesh was the old Andhra State of 1953 enlarged by Telangana, the Telugu-speaking region of the dissolved Hyderabad State, so that the first linguistic state became a larger and different thing three years after its creation. Kerala was the old Travancore-Cochin joined to the Malabar district of Madras and the Kasaragod taluk of South Canara, while five southern taluks of Travancore passed in the other direction to Madras, a transfer whose grievance fed Tamil politics in the Kanyakumari area for a generation. Mysore was enlarged on every side, by Coorg, by the Kannada districts of Bombay including Belgaum and Dharwad, by Bellary and South Canara from Madras, and by the Kannada districts of Hyderabad, becoming the united Kannada state the unification movement had demanded since before independence. Madhya Pradesh was assembled from four predecessors, the old Madhya Pradesh with Madhya Bharat, Bhopal and Vindhya Pradesh, but without Vidarbha, which had been taken from it and given to Bombay. Rajasthan absorbed Ajmer. Bombay absorbed Saurashtra and Kutch and gained Marathwada from Hyderabad and Vidarbha from Madhya Pradesh, becoming the largest and most internally divided state in the Union. Punjab absorbed the Patiala and East Punjab States Union and remained bilingual. Assam, Bihar, Orissa, Uttar Pradesh, West Bengal and Jammu and Kashmir continued, with adjustments at their edges. Of the Part C states, Delhi, Himachal Pradesh, Manipur and Tripura became Union territories rather than being merged out of existence, and the Laccadive, Minicoy and Amindivi Islands were detached from Madras to form the sixth territory.

What was left out is as much a part of the Act as what it enacted. There was no state of Maharashtra and no state of Gujarat; the Marathi and Gujarati claims had been deferred together inside bilingual Bombay. There was no Punjabi-speaking state and no Haryana; the Punjab claim had been refused outright. There was no Vidarbha, though the Commission had proposed one. There was no separate Telangana, though the Commission had proposed to preserve it. The north-eastern hills remained inside Assam. Goa remained a Portuguese possession, and the tribal and regional claims that would surface later, in Jharkhand, in the north-east, in Uttarakhand and elsewhere, had not yet been organised into movements the state had to answer. The settlement of 1956 was, by its own internal logic, a settlement of the claims that had matured by 1955, and it carried no promise that the map was closed.

Reading the Act: its architecture

A reader who opens the Act itself, rather than the accounts of it, finds a document of a distinctive kind. Its early parts work through the territories state by state and territory by territory, defining each new unit by reference to the districts and divisions of the old, in the dry cartographic prose of schedules and appointed days. Its middle parts deal with representation, apportioning the seats of the new units in the two Houses of Parliament and in the legislative assemblies and councils until the next readjustment, and with the delimitation of the constituencies from which the election of 1957 would be fought. Its later parts turn to the machinery: the High Courts, the allocation of the services, the apportionment of property, rights and liabilities, the authorisation of expenditure, the continuance and adaptation of laws, and the removal of difficulties by executive order where the text had not foreseen a case. The zonal councils stand in their own part, the only portion of the Act addressed not to the transfer of power but to its future exercise. What the Act does not contain is equally characteristic. It recites no principle. It declares no policy of linguistic states. Its long title says only that it is an Act to provide for the reorganisation of the states of India and for matters connected therewith, and the whole of the most consequential redrawing of the Indian map is executed in the grammar of definition and transfer, as though the Act were a conveyance rather than a constitution. The choice was deliberate and it was wise: a statute that had proclaimed a principle would have been quoted against every later refusal, and the authors of 1956, who had refused the principle throughout, were not going to enact it in a preamble.

The government of the largest of the new units illustrated the settlement’s character in a single appointment. The bilingual Bombay created on the appointed day was entrusted to Yashwantrao Chavan, a Marathi leader who had supported the united Maharashtra claim and who undertook to govern, for three and a half years, the composite state his own movement existed to divide, administering Marathi and Gujarati districts from the same secretariat and defending the bilingual arrangement in public while its obituary was being written by both movements. That he governed it competently, and then became the first Chief Minister of Maharashtra when the division came, is a small fact with a large moral for the interpretation of the Act: the settlement’s units were not abstractions to the politicians who had to run them, and the interval of bilingual Bombay, usually written as a failure, was also a demonstration that the composite state could be administered by goodwill for as long as goodwill was supplied. What failed in 1960 was not the administration of the bilingual state but the politics of it, and the distinction is the Commission’s posthumous defence: it had recommended the bilingual state as an administrative and economic proposition, and as an administrative and economic proposition it worked, in the one place where its political foundations were weakest.

The appointed day

The transfer of 1 November 1956 was executed without an interregnum, and the mechanics of it are part of the Act’s design. Governors were appointed for the new states, in several cases the same person holding charge of more than one under the new proviso to Article 153, and the ministries of the predecessor units continued, adapted to their new charges, so that no state woke on the appointed day without a government. The legislative assemblies of the new states were constituted in the first instance from the members of the assemblies of the predecessors, sitting for their existing constituencies redistributed among the successor units, and the general election of 1957, the second of the republic, was the first fought on the new map, on constituencies delimited afresh for the new states. The courts sat without interruption under their adapted jurisdictions; the laws of the predecessor states continued in the transferred territories until altered, under the Act’s adaptation provisions; and the public services were allocated by committees whose work, unglamorous and litigated, determined where tens of thousands of officers would spend their careers. A reorganisation remembered as an affair of movements, commissions and statutes was, at the level of the state itself, an affair of lists: lists of territories, of laws, of assets, of posts and of persons, and the quality of the list making is a large part of why the transfer of 1956, unlike the transfer of 1947 with which its authors could not help comparing it, was an administrative event rather than a human catastrophe.

The Seventh Amendment, the constitutional twin

An ordinary Act of Parliament could redraw the territories of the states under Article 3, but it could not by itself rewrite the constitutional provisions that described those states, and so the reorganisation travelled with a constitutional amendment, the Constitution (Seventh Amendment) Act, 1956, which came into force on the same day. The amendment substituted Article 1 and the First Schedule so that the Constitution thereafter spoke of states and Union territories without the fourfold classification that had recorded the accidents of accession; a state was a state, whether it had been a governor’s province, a princely state or a union of princely states, and the office of Rajpramukh disappeared with the class that had required it. It substituted the Fourth Schedule, reallocating the seats in the Council of States among the new units, and amended the provisions governing the composition of both Houses of Parliament, the representation provisions that the new map had made obsolete. Those allocations stood until the next general readjustment of representation, a process whose own statutory history, from the first Delimitation Commission to the long freeze on interstate seat shares, is traced in the history of the Delimitation Acts.

Three of the amendment’s changes were structural rather than consequential, and they are the ones that repay attention, because each of them generalised a device the reorganisation had needed. The amendment added a proviso to Article 153 making it possible to appoint the same person as Governor of two or more states, an economy and a convenience that later practice used well beyond the transitional circumstances of 1956. It substituted Articles 230 and 231 so that Parliament may extend the jurisdiction of a High Court to a Union territory, or exclude it, and may establish a common High Court for two or more states, or for states together with a Union territory; the common High Court for Punjab and Haryana, sitting at Chandigarh, is the standing monument of that provision. And it recast the distribution of legislative business for the new units, amending the Seventh Schedule whose three lists of subjects, Union, State and Concurrent, are mapped in the comparison of the Union, State and Concurrent Lists.

The amendment also reached the question the reorganisation had made urgent and which the Commission’s safeguards had anticipated: the position of the linguistic minorities the new map created in every state. It inserted Article 350A, directing every state to endeavour to provide adequate facilities for instruction in the mother tongue at the primary stage to children belonging to linguistic minority groups, and Article 350B, creating a Special Officer for linguistic minorities, appointed by the President, charged with investigating the safeguards provided for those minorities and reporting to the President, the reports to be laid before Parliament and sent to the governments concerned. The two articles are the constitutional trace of the Commission’s central admission, that the settlement would manufacture minorities as surely as it satisfied majorities, and that a settlement which did not provide for the manufactured minorities would be incomplete on its own terms.

The first elections on the new map

The general election of 1957, held within months of the appointed day, was the settlement’s first examination, and its results were read as a verdict on the whole exercise. The Congress retained office in most of the new states, which the government took as evidence that the reorganisation had satisfied the demands it was meant to satisfy without releasing forces it could not contain. In Kerala, the state whose creation had been among the least contested of the Commission’s proposals, the election produced the result that has been quoted ever since in the argument about linguistic states: a Communist ministry under E. M. S. Namboodiripad, formed through the ballot box in a state whose coherence as a Malayalam speaking unit made a statewide politics, and a statewide mandate, possible in a way the old division between Travancore-Cochin and Malabar had frustrated. Defenders of the settlement cite Kerala 1957 as the proof that a linguistic state deepens democracy by giving it a single public to address. Critics cite the same event, and the dismissal of that ministry by the Centre in 1959, as the proof that the new states would become arenas of a politics the Union would repeatedly be tempted to override. Both readings were available within three years of the appointed day, and the durability of the settlement is in part the durability of that double reading.

The regions the settlement passed by

A general settlement is also a record of the claims that had not matured in time to be considered, and the map of 1956 contains several silences that later decades filled. The hills of the north-east remained within Assam: the Naga Hills, whose people were in armed and political movement for a separate existence through the very years of the Commission’s inquiry, the Khasi and Garo hills, the Lushai Hills and the frontier tracts administered as the North East Frontier Agency. The Commission’s scheme was a scheme for the plains provinces and the princely unions; it was not designed for the tribal hill country, whose administration had always stood apart, and the questions there were answered later and piecemeal, by the creation of Nagaland in 1963 and by the reorganisation of the region in 1971. Goa remained outside the Union altogether, a Portuguese possession on the western coast whose liberation in 1961 and constitution as a Union territory postponed its statehood for a further quarter of a century. The French establishments on the coast, transferred in fact in 1954, remained in a transitional condition the settlement did not touch. Jammu and Kashmir continued under its own constitutional arrangements, listed among the fourteen states but reorganised in nothing. These silences matter to the article’s central claim: the settlement of 1956 was not a theory of the Indian map carried into effect. It was an answer to the questions that had become unanswerable by delay, framed by men who knew that other questions were forming and who left the machinery of Article 3 exactly as wide as they had found it, for whoever would have to answer them.

What the Seventh Amendment did not do is as instructive as what it did. It did not entrench the new map: the First Schedule it substituted could be, and repeatedly has been, rewritten by the ordinary Article 3 process, because Article 4 declares that a law made under Article 3 is not to be deemed an amendment of the Constitution for the purposes of Article 368. It did not make the consent of a state a condition of its own reorganisation, though the reorganisation it accompanied had consulted the states more extensively than the bare text required. And it did not attempt to define the states of the Union by any principle, linguistic or otherwise: the schedule it substituted is a list, not a theory, and the Constitution after 1956 describes the Indian map exactly as the Constitution before 1956 had done, as whatever Parliament has enacted it to be. The settlement’s constitutional philosophy, if the phrase is permitted, is contained in that refusal. The founders of 1956 had the opportunity, in an amendment carried for the purpose, to write language into the definition of a state. They wrote a list instead.

The completions: 1960, 1966, 1971 and 1987

The questions deferred in 1956 returned in the order of their political maturity, and each was answered by the same machinery of an ordinary reorganisation Act under Article 3. Bombay was first. The bilingual state pleased neither of its great language groups, the movements for Samyukta Maharashtra and for Mahagujarat continued to organise and to win elections on the question, and the Bombay Reorganisation Act, 1960 divided the state on the first of May 1960 into Maharashtra, which took the city of Bombay, and Gujarat. The division conceded, four years after the Commission had proposed the opposite, that the economic argument for a shared state could not hold against two democratic movements each commanding its own electorate. Punjab followed. The Punjabi Suba demand, refused by the Commission in 1955, acquired a new organisation and a new urgency, and after a boundary commission under Justice J. C. Shah had examined the tehsils in dispute, the Punjab Reorganisation Act, 1966 took effect on the first of November 1966, creating a Punjabi-speaking Punjab, a Hindi-speaking Haryana, a Union territory of Chandigarh to serve as the common capital, and transferring the hill areas of the old Punjab to Himachal Pradesh, which became a full state in 1971 under its own Act of 1970.

The north-east was reorganised on a different logic. The North-Eastern Areas (Reorganisation) Act, 1971, effective in January 1972, raised Manipur and Tripura from Union territories to states, carved the state of Meghalaya out of Assam, and constituted Mizoram and Arunachal Pradesh as Union territories, answers in the main to tribal and hill identities that the language test of 1956 had not been designed to measure. Goa completed the linguistic series in a sense, becoming a state on the thirtieth of May 1987 under the Goa, Daman and Diu Reorganisation Act, 1987, after an opinion poll in 1967 had rejected its merger into Maharashtra and the Konkani language movement had established the claim on which statehood rested. With Goa, the map the Commission had contemplated was, in its broad linguistic outline, finished, thirty one years after the Act that was supposed to settle it.

What followed belongs to a different history, in which the ground of claim shifted from language to region, development and tribe. The three states created in November 2000, Chhattisgarh, Jharkhand and Uttarakhand, examined in the account of the reorganisation Acts of 2000, were carved out by consenting legislatures and without the agitational politics that had made 1956 necessary. Telangana’s creation in 2014 under the Andhra Pradesh Reorganisation Act, 2014 closed the oldest file of all, enacting sixty years later the separation the Commission had proposed to preserve in 1955, after the safeguards of the Gentlemen’s Agreement had been judged by the region to have failed. And the conversion of the state of Jammu and Kashmir into two Union territories in 2019 demonstrated that the Article 3 power the settlement used remains as wide as it was, running in both directions. None of these later Acts displaced the settlement of 1956; they worked within the constitutional world it made, in which the internal boundaries of India are a standing subject of ordinary legislation rather than a fixed feature of the founding document.

The unfinished file: Vidarbha and Telangana after 1956

Two of the Commission’s recommendations refused in 1956 had long afterlives, and they measure the cost of the legislature’s preference for the immediate settlement over the staged one. Vidarbha had been promised equity before it was promised a state: the Nagpur Pact of 1953, concluded among the Congress leaders of the Marathi regions, had recorded undertakings of fair treatment in development, employment and education for Vidarbha and Marathwada as the price of their joining a united Maharashtra, and the Seventh Amendment itself wrote a part of that bargain into the Constitution, giving the Governor of Maharashtra a special responsibility, under Article 371, for the establishment of separate development boards for Vidarbha and Marathwada, and for their counterparts in Saurashtra and Kutch in Gujarat, charged with the equitable allocation of development funds. The boards themselves were constituted only decades later, and the Vidarbha claim, deprived of the statehood the Commission had recommended, settled into a permanent politics of comparative grievance, in which the region’s spokesmen cite the Commission’s recommendation of 1955 as the standing authority for what was refused them in 1956 and again in 1960.

Telangana’s afterlife was sharper. The safeguards of the Gentlemen’s Agreement, a regional committee of the state legislature, protections for local candidates in public employment under the domicile rules inherited from Hyderabad, and undertakings on the application of the region’s revenues, depended for their force on the goodwill of a state government in which the region was a minority, and within a little more than a decade the region’s complaint was that the safeguards had been honoured in form and defeated in administration. The agitation of 1969 for a separate Telangana, in which students and public employees were the moving force, was the first great internal crisis of a state the settlement had created, and it was answered at the time with formulas and plans rather than separation. The question then outlasted every government that declined it, and its eventual resolution in 2014, by the separation the Commission had proposed to preserve in 1955, is the strongest single piece of evidence for the proposition that the Commission’s staged and safeguarded solutions were not the timid options they were taken for. The legislature of 1956 had preferred the clean merger; the clean merger lasted fifty eight years and ended in the division the Commission had tried to spare it.

What changed inside the states

The settlement is usually discussed as a redrawing of lines, but its deepest effects ran inside the units it created, because a state in the Indian federation is not a label on a map but a working apparatus of government. Each new state acquired a legislature and a ministry responsible to it, a cadre of the all India and state services allocated between the successor units, a share of the property, contracts and liabilities of its predecessors, and a High Court, the Act itself making provision for the courts of the new states, constituting a High Court for Kerala and a High Court for the enlarged Madhya Pradesh and settling the jurisdiction of the courts of the composite states it left behind. The allocation of services was the least visible and most resented part of the machinery: officers who had served a predecessor state found themselves assigned to a successor whose language some of them did not speak and whose politics they had not chosen, and the integration of the services of the princely states with those of the provinces, begun at accession, had to be completed inside units that were themselves new. The protections the settlement wrote for such officers, and the disputes the allocations generated, occupied the administrative tribunals for years.

The change of official language followed the change of boundary and was its practical meaning. Under Article 345 of the Constitution each state legislature may adopt the language or languages in use in the state as its official language, and the new states moved, at differing speeds, to conduct their administration, their subordinate courts and their public examinations in the language of the majority. Education followed administration. The three language questions that dominate Indian public life, the medium of instruction in schools, the language of the courts, and the language of recruitment to state services, were all transformed by the settlement without appearing in its text, because a decision about where the boundary ran was in substance a decision about which language a child would be examined in and in which tongue a litigant would be heard. That is the sense in which the reorganisation, though executed by ordinary legislation about territory, was the most consequential cultural legislation of the republic’s first decade.

The division of property was the settlement’s least philosophical and most laborious part. Every predecessor unit had owned buildings, railways sidings, irrigation works, printing presses, rest houses, debts and funds, and every asset and liability had to be apportioned between the successors, by agreement where agreement was possible and by order of the President where it was not. The principles the Act applied, population as the general measure, location as the measure for immovable property, and the situs of a scheme as the measure for the works that embodied it, were less important than the fact of their application at scale: the Union rewired its own federal finances in a single season without a default, a repudiation or a breakdown of any public service, an achievement of the finance and home ministries’ officials that the political history of the settlement rarely pauses to record. The same cannot be said of the rivers. The new boundaries divided the basins of the Krishna and the Godavari among successor states with competing claims to their waters, and Parliament, legislating in the same year under Article 262, enacted the Inter-State Water Disputes Act, 1956 to provide tribunals for the disputes the map was bound to generate. The connection is direct: a settlement that redrew states around language necessarily redrew them across rivers, and the water disputes of the following half century, between Andhra Pradesh and Karnataka, between Maharashtra and its neighbours, are among its longest running consequences.

The federation’s finances had to be rewired with the map, and the rewiring fell to the constitutional machinery of the Finance Commission. The award of the First Finance Commission had been made for the states of the old map; the Second Finance Commission, appointed in 1956 under the chairmanship of K. Santhanam, was the first to distribute the divisible taxes and the grants in aid of the revenues among the states the settlement had made, and its work illustrates a truth the boundary debate habitually ignores: the viability criterion, on which the Commission had granted and refused claims, was not an abstraction, because a state created in November had to be financed in April, and the difference between a viable unit and an unviable one was the difference between a government that could pay its teachers and one that began its life as a supplicant of the Centre. The Part C states the settlement abolished had been the standing example of the unviable form, administered units carried on the Union’s budget; the settlement’s preference for larger and fiscally complete states, its refusal of the small claims catalogued earlier in this article, and its consolidation of the princely fragments into units with real tax bases, were of a piece with the finance of the federation as the Finance Commission had to practise it. The map and the money were the same question stated twice.

The minorities the map made

Every state the Act created contained a minority whose language was not the language of the state, and some of the minorities were very large. Bombay in 1956 held its two great language groups as mutual minorities in a single polity. The enlarged Mysore held Marathi speakers in the Belgaum area whose claim to join Bombay, and later Maharashtra, outlasted the settlement by decades and produced a boundary dispute between two states of the Union that commissions of inquiry were still examining a generation later. Madras held Telugu speakers along its northern edge. The border districts everywhere, where one language shaded into another across a countryside of bilingual villages, could not be assigned without assigning substantial populations against their preference, and the Commission’s files on districts such as Belgaum, Kasaragod and the taluks between Bihar and Bengal record the impossibility in detail.

The settlement’s answer to the minorities it created had three parts. The first was constitutional and general: the fundamental right of any section of citizens to conserve its language and script, and the right of linguistic minorities to establish and administer educational institutions of their choice, had stood in the Constitution since 1950, and Article 347 empowered the President to direct that a language spoken by a substantial proportion of a state’s population be officially recognised for specified purposes. The second part was added by the Seventh Amendment itself, in Articles 350A and 350B, the undertakings on mother tongue instruction at the primary stage and the Special Officer for linguistic minorities. The third part was administrative and local: the zonal councils, bilingual arrangements in districts of mixed population, and the practice, uneven but real, of permitting the use of the minority language in dealings with the state in districts where the minority was concentrated. The record of these safeguards in operation was mixed, and the Special Officer’s reports became a standing catalogue of their shortfalls, but their existence in the settlement matters to its interpretation: the authors of 1956 did not believe that the linguistic state was self justifying, and they wrote the protection of the minorities it created into the same instruments that created it.

A later generation of scholarship added a comparative frame that both readings have to accommodate. The political scientist Alfred Stepan classified federations as coming together, where previously sovereign units pool their sovereignty, or holding together, where a formerly unitary state devolves power to hold a multinational polity within a single citizenship, and he treated India, with its linguistic reorganisation, as the exemplary holding together case: a federation that survived by conceding, in constitutional form, the very demands that a unitary logic would have suppressed. On that view the settlement’s significance is not national but comparative. Spain and Belgium, in their different ways, faced the choice India faced in 1956; the Indian answer, to redraw the internal map around language while refusing to concede either secession or a hierarchy of units, is the answer that has lasted longest with the least violence, and the four criteria, unfashionable as they sound beside the language of self determination, are the technique by which the answer was reached.

What the passage record predicted

Template and hindsight aside, the manner of the Act’s passage contained a set of predictions about how it would work, and most of them were borne out. A settlement reached by conceding under pressure what had been refused on principle predicted further pressure, and got it: the two deferred questions were reopened within the decade, and the technique of the fast, the morcha and the organised movement remained the recognised currency of statehood claims. A settlement whose safeguards for regions and minorities were written as undertakings rather than as enforceable rights predicted that those safeguards would be tested by administration rather than by litigation, and they were, in Telangana within thirteen years and in Vidarbha for two generations. A settlement that left the Article 3 power unqualified, enacting the whole redrawing by ordinary legislation over the recorded objections of some of the units redrawn, predicted that the power would be used again whenever a majority at the Centre found a map inconvenient, and it has been, in directions the authors of 1956 did not contemplate and under procedures they would have recognised at once. And a settlement reasoned in public against stated criteria predicted that its own report would become the standing brief for every refused claim, which is the use to which the movements of Vidarbha and of a dozen other regions have put it ever since. The passage record of a statute is usually treated as antiquarian. In this case it is the best available guide to the statute’s future, because the conditions the passage created are the conditions the Act has lived under.

Two readings of the settlement

The favourable reading of the settlement is that it saved the Union from its most dangerous internal question by conceding it in time. On this reading, associated above all with Granville Austin’s account of the Constitution in its first decades, the linguistic demand was a democratic demand of great depth, and a state that had refused it indefinitely would have faced the progressive alienation of its largest language groups and the conversion of every provincial grievance into a grievance against the Union itself. By conceding the substance while refusing the doctrine, the settlement drew the language movements inside the constitutional order, gave them governments to run and budgets to defend, and converted movements of protest into administrations with an interest in the Union’s stability. The durability of the result is the reading’s strongest evidence: the map settled in 1956, and completed in 1960 and 1966, was altered only at the margins for the rest of the century, and the states it created became the normal units of Indian politics, the arenas in which national parties compete and through which the Union is governed.

The critical reading begins where the Commission itself began to worry. K. M. Munshi had warned the Dar Commission that linguistic politics satisfied one group by excluding others, and the critics of the settlement argue that the warning was borne out inside the new states: that linguistic majorities treated the state as the property of their language, that minorities in education, employment and public honours paid the price, and that regions whose claims were subsumed inside larger linguistic states, Vidarbha inside Maharashtra, Telangana inside Andhra Pradesh, the hill districts inside Assam, inherited a permanent politics of neglect in which the regional grievance replaced the linguistic one without diminishing. Paul Brass, studying the language movements of north India, added a colder observation: that language loyalties in politics are not simply found but made, elaborated by elites for whom a language movement is an instrument of mobilisation, so that a settlement which rewards mobilisation teaches the next generation of politicians the value of the method. On this reading the settlement did not end the politics of internal boundaries; it licensed it, and the succession of statehood demands that has followed, some granted and some pending for decades, is the working of the licence.

Both readings can point to the same facts, because the settlement genuinely did two things at once. It accommodated democratic demands that no Indian government could have resisted for long without ceasing to be democratic, and it did so by a technique, concession under pressure after formal refusal, that invited the repetition of pressure. The honest assessment has to hold both. Where the favourable reading is plainly right is on the counterfactual: the states that were refused in 1948 and 1949 and conceded in 1953 and 1956 did not become less loyal to the Union for having been conceded, and the two questions left over in 1956 were settled by legislation in 1960 and 1966 without damage to the federation. Where the critical reading is plainly right is on the minorities and the subsumed regions: the settlement’s safeguards were weaker than its boundaries, and the grievances it stored up, in Belgaum, in Vidarbha, in Telangana, were the direct products of its own compromises, visible in its documents to anyone who read them.

The settlement as precedent

The reorganisation of 1956 became the template against which every later exercise of the power has been measured, and the comparison is not flattering to all of its successors. The exercises that followed its method most closely, a commission or committee of inquiry, published reasoning, an Act debated at length, appointed days and detailed provisions for services, assets and courts, were the ones that settled their questions with the least residue of violence: the division of Bombay in 1960 followed further agitation but was executed as administration, and the creations of 2000 were effected by a procedure so consensual that the assemblies of the states divided voted for their own division. The exercises that departed from the method, deciding first and reasoning afterwards, inherited the disputes the method was designed to exhaust. The precedent also fixed the standing questions of Indian reorganisation, which are the questions of 1956 restated: whether the views of the affected state, formally non binding since the settlement and confirmed as such by the courts, ought in constitutional morality to amount to more; whether an inquiry on stated criteria should precede any redrawing as a matter of convention; and whether the protection of the regions and minorities a new boundary subsumes should be written as rights before the boundary is drawn rather than as undertakings after it. The men of 1956 answered the first two questions one way in their texts and another way in their practice, and left the third as the unfinished business of the settlement. A reader of the Act who keeps those three questions in view holds the key both to its history and to every reorganisation debate that has followed it.

The contrast with the most recent exercises of the same power is the clearest measure of what the method was worth. The creations of the year 2000 were preceded by resolutions of the very assemblies that were divided, so that consent, which the Constitution does not require, was obtained in fact before the law was made, and the appointed days passed without a shot or a shutdown. The reorganisation of Jammu and Kashmir in 2019 was effected while the state was under President’s rule, with Parliament expressing the views the state legislature would have expressed, a course the text of Article 3 permits and the practice of 1956 had not contemplated, because in 1956 the states concerned were functioning, were consulted, and were answered. Nothing in the later history shows the machinery of 1956 to have been legally necessary. Everything in it suggests that the machinery was politically necessary, and that the authors of the settlement, who had lived through a transfer of power conducted without it, understood the difference in their bones.

The verdict the record supports

The States Reorganisation Act 1956 was not the adoption of a linguistic principle, and its history cannot be understood if it is read as one. It was the point at which three things met: a national movement’s old promise, a post Partition state’s fear of every line of division, and a set of regional movements strong enough to make the fear more expensive than the concession. The Commission that designed the settlement understood all three and wrote its understanding into the order of its criteria, unity first and language second, with viability and welfare held in reserve as the tests by which a linguistic claim could still be refused. Parliament, in enacting the settlement, softened each of the Commission’s hardest applications of its own tests, deferring Bombay, accelerating Telangana, abandoning Vidarbha, and the questions so softened returned, in 1960, in 1966 and in 2014, to be answered on terms closer to the claims than to the criteria. The Act’s longest lasting achievement may be its least discussed: the abolition of the fourfold classification of states, which ended the constitutional ranking of Indian units by their origin and made every state the equal of every other in form, whatever their differences in size and wealth.

For the reader who wants to work through the primary record, the Commission’s report and the parliamentary debates repay the effort in a way summaries cannot, and readers assembling the report, the debates and their own notes into a working file will find that a document viewer built for long files, such as the office file viewer on ReportMedic, makes the comparison of the report’s recommendations with the Act’s sections considerably easier. This article is a work of legislative history offered as general information about a public record; it is not advice on any question of law, and anyone concerned with a live question touching state boundaries or federal arrangements should go to the bare text of the Act and the amendment rather than to any summary, including this one.

The Act itself remains on the statute book. Most of its territorial provisions are spent, overtaken by the later Acts that divided Bombay, divided Punjab, reorganised the north-east and created the states of the present map, but its provisions for the zonal councils continue to operate, and the councils still meet, the standing institutional survival of the settlement’s theory that the states of a zone must be given a table at which to sit together. The report that preceded it has had the stranger afterlife. It is cited, seventy years of argument later, by the advocates of states the Commission never heard of, because it remains the fullest statement the Indian state has produced of how a boundary claim should be tested: not by the passion behind it, not by the language alone, but by unity, by homogeneity, by viability and by welfare, in that order. The settlement of 1956 will be superseded piece by piece as the map continues to change, as every Indian map has. The criteria have not been superseded, because nobody has improved on them, and the order in which they stand is the order in which the questions of any future reorganisation will in fact be decided, whether the deciders acknowledge the source or not.

Frequently Asked Questions

Q: What did the States Reorganisation Act 1956 change on the map?

It replaced the twenty seven units of the old map with fourteen states and six Union territories on 1 November 1956, and abolished the classification of states into Part A, Part B, Part C and Part D. Whole states disappeared into their neighbours: Hyderabad was dissolved among Andhra Pradesh, Bombay and Mysore, Saurashtra and Kutch went into Bombay, the Patiala and East Punjab States Union into Punjab, Coorg into Mysore, Ajmer into Rajasthan, and Madhya Bharat, Bhopal and Vindhya Pradesh into Madhya Pradesh. It created Kerala, enlarged Mysore and Andhra Pradesh, kept Bombay and Punjab bilingual, and made Delhi, Himachal Pradesh, Manipur, Tripura and the island territories Union territories.

Q: Why was language used as the basis for reorganisation?

Language was used because no other test was available for the claims that had become politically unavoidable, and because it had genuine administrative force: a state whose schools, courts and offices work in the language of the governed is easier to run and easier to hold to account. But it was never adopted as the basis in the sense of a rule. The States Reorganisation Commission treated linguistic homogeneity as the second of four considerations, behind the unity and security of India and ahead of financial viability and planning, and it refused claims that failed the other tests. The map looks linguistic because the irresistible claims were linguistic, not because language governed wherever it was resisted successfully, as Punjab and Bombay showed in 1956.

Q: What did the Dar Commission and JVP Committee recommend?

Both advised against linguistic reorganisation at that stage. The Linguistic Provinces Commission under S. K. Dar, reporting in December 1948, held that forming provinces on exclusively or even mainly linguistic considerations was not in the larger interests of the Indian nation, and proposed geographical contiguity, financial self sufficiency and ease of administration instead. The JVP Committee of Jawaharlal Nehru, Vallabhbhai Patel and Pattabhi Sitaramayya, reporting in April 1949, agreed that the time was unsuitable, but added that if public sentiment proved insistent and overwhelming, democrats would have to submit to it, subject to the good of India as a whole.

Q: Which state was created first on linguistic lines?

Andhra State, created on 1 October 1953 from the Telugu speaking districts of Madras, with its capital at Kurnool. It followed the death of Potti Sreeramulu in December 1952 after a fast of fifty six days for the Andhra cause, and the rioting that forced the government’s hand. Andhra State was not the Andhra Pradesh of the later map: in 1956 it was enlarged by the Telangana region of Hyderabad to form the united Telugu state. The province of Orissa, formed in 1936, had been drawn with language in view a generation earlier, but Andhra was the first state created on linguistic lines in independent India.

Q: What did the Seventh Amendment change?

The Constitution (Seventh Amendment) Act, 1956, in force with the reorganisation on 1 November 1956, substituted Article 1 and the First Schedule to list states and Union territories, and abolished the Part A, Part B, Part C and Part D classification and the office of Rajpramukh. It substituted the Fourth Schedule’s allocation of seats in the Council of States and amended the provisions on representation in Parliament. It added a proviso to Article 153 permitting the appointment of the same person as Governor of two or more states, substituted Articles 230 and 231 to permit the extension of a High Court’s jurisdiction to a Union territory and the establishment of a common High Court for two or more states, and inserted Articles 350A and 350B for the protection of linguistic minorities.

Q: Which states were left out of the 1956 reorganisation?

The claims left unsatisfied were the separate states of Maharashtra and Gujarat, deferred inside the bilingual Bombay the Act created; a Punjabi speaking state, and with it Haryana, refused outright on the Commission’s advice; a separate Vidarbha, which the Commission had recommended and the government declined; and a separate Telangana, which the Commission had proposed to preserve with a later option of merger. The tribal hill country of the north east also remained inside Assam. Maharashtra and Gujarat came in 1960, Punjab and Haryana in 1966, the north eastern states from 1972, and Telangana in 2014. Vidarbha has not been created.

Q: What criteria did the States Reorganisation Commission apply?

Four, in the order the Commission stated them: the preservation and strengthening of the unity and security of India; linguistic and cultural homogeneity; financial, economic and administrative considerations, including the viability of the proposed unit; and planning and the promotion of the welfare of the people in each state as well as of the nation as a whole. The order was the substance: unity ranked above language, and viability and welfare could defeat a linguistic claim that the second test supported. The Commission expressly rejected the theory of one language, one state, and refused to treat language as the sole or overriding test of any boundary.

Q: Which later Acts completed the linguistic map?

The Bombay Reorganisation Act, 1960 divided bilingual Bombay into Maharashtra and Gujarat. The Punjab Reorganisation Act, 1966 created a Punjabi speaking Punjab, the state of Haryana and the Union territory of Chandigarh, and moved the hill areas of the old Punjab to Himachal Pradesh, which became a state in 1971. The North-Eastern Areas (Reorganisation) Act, 1971 raised Manipur and Tripura to statehood and carved Meghalaya out of Assam. The Goa, Daman and Diu Reorganisation Act, 1987 made Goa a state. The Andhra Pradesh Reorganisation Act, 2014, creating Telangana, closed the oldest file of 1956, though its ground by then was regional as much as linguistic.

Q: What did the States Reorganisation Commission propose for Bombay, and what did Parliament enact instead?

The Commission proposed a bilingual Bombay state of the Marathi and Gujarati regions, with Saurashtra and Kutch, on the ground that the city’s economy bound it to both hinterlands and that its population was too mixed for a clean division. The government’s first decision instead proposed three units: Maharashtra, Gujarat and Bombay city as a centrally administered unit. Rioting in the city in January 1956 forced a reversal, and Parliament enacted a bilingual Bombay larger than the Commission’s, enlarged by Marathwada from Hyderabad and by Vidarbha from Madhya Pradesh. The city question was thus postponed, not settled, and the state was divided in 1960.

Q: Why was Vidarbha not made a separate state in 1956?

Because the government’s handling of Bombay overtook it. The Commission had recommended a separate Vidarbha, finding its resources and its administrative history as the eastern division of the old Central Provinces sufficient for statehood. But when the government abandoned its three unit formula for Bombay and settled on a single enlarged bilingual state, Vidarbha was added to Bombay to give the Marathi side of that state its full weight, and the separate state disappeared into the compromise. The Nagpur Pact of 1953 had already promised Vidarbha equitable treatment inside a united Maharashtra, and the claim for statehood has rested on the Commission’s recommendation ever since.

Q: What happened to Hyderabad State under the 1956 settlement?

Hyderabad State, the largest of the Part B states, was dissolved. Its Telugu speaking region, Telangana, was merged into Andhra to form Andhra Pradesh, under the safeguards of the Gentlemen’s Agreement of February 1956, and Hyderabad city became the capital of the united state. Its Marathi speaking region, Marathwada, went to Bombay, and in time to Maharashtra. Its Kannada speaking districts went to Mysore. The Commission had proposed that Hyderabad continue as a separate state, with an option to unite with Andhra after the general elections of 1961 if its legislature resolved in favour by a two thirds majority; the enacted settlement merged it at once.

Q: Why did the Commission refuse a Punjabi-speaking state in 1956?

The Commission concluded that the demand lacked the general support a reorganisation required and that, in a border province so soon after Partition, a division whose line corresponded closely with the division between Sikhs and Hindus would endanger the unity it was charged to protect. It also weighed the language returns of the census, in which many Punjabi speaking Hindus had recorded Hindi as their mother tongue, as evidence that the linguistic case itself was contested on the ground. It proposed instead an enlarged bilingual Punjab, absorbing the Patiala and East Punjab States Union. The demand returned in organised form and was conceded by the Punjab Reorganisation Act, 1966.

Q: Did the States Reorganisation Commission accept the idea of one language, one state?

No. It rejected the theory expressly, on the ground that several languages were spoken over areas too large for single states, that several regions were too mixed for any line to unmix, and that a single state for the Hindi speaking area alone would have held a share of the Union’s population large enough to distort the federation. It accepted the different proposition, associated with B. R. Ambedkar, of one state, one language: a state should as far as possible work in one language, but one language might properly have several states, as Hindi already had four.

Q: How many states and union territories did the Commission recommend, and how did the enacted map differ?

The Commission recommended sixteen states and three centrally administered territories, the territories being Delhi, the Andaman and Nicobar Islands and the Laccadive, Minicoy and Amindivi Islands. Parliament enacted fourteen states and six Union territories. The differences were that Vidarbha and Hyderabad, both recommended as states, were absorbed into Bombay and Andhra Pradesh respectively, while Himachal Pradesh, Manipur and Tripura were retained as Union territories rather than merged into their neighbours. The enacted map was therefore smaller in states and larger in territories than the design, a shift produced by the political modifications of 1956 rather than by any failure of the Commission’s tests.

Q: What did the Act do to the princely unions such as PEPSU, Saurashtra and Travancore-Cochin?

The Act dissolved them into their neighbours. The Patiala and East Punjab States Union went into Punjab; Saurashtra and Kutch went into Bombay; Madhya Bharat, Bhopal and Vindhya Pradesh went into the enlarged Madhya Pradesh; Travancore-Cochin, joined with the Malabar district and Kasaragod from Madras, became Kerala; and Hyderabad was divided among three successors. The Rajpramukh who had headed each Part B state disappeared with the classification the Seventh Amendment abolished. No princely union survived the settlement as a unit, and the map of 1956 is the point at which the two inheritances of 1947, the British provinces and the princely states, finally became a single order of states of equal form.

Q: What are zonal councils and why did the 1956 Act create them?

Zonal councils are statutory advisory bodies, one for each of the five zones, northern, central, eastern, western and southern, established by the States Reorganisation Act itself. Each brings together the chief ministers of the states of the zone with the Union Home Minister in the chair, to discuss matters of common interest: economic and social planning, border disputes, linguistic minorities, inter state transport and the working of the settlement itself. The Act created them because a map drawn more closely around language was expected to make inter state cooperation harder, and the councils were the machinery supplied for the cooperation the map had made more necessary. They continue to meet.

Q: What safeguards did the 1956 settlement create for linguistic minorities?

Three layers of them. The Constitution already protected the right of any section of citizens to conserve its own language and script, and the right of linguistic minorities to establish and administer educational institutions of their choice, and Article 347 allowed the President to direct official recognition of a language spoken by a substantial proportion of a state’s people. The Seventh Amendment added Article 350A, directing states to endeavour to provide mother tongue instruction at the primary stage, and Article 350B, creating a Special Officer for linguistic minorities charged with investigating the safeguards and reporting to the President. Around these stood the settlement’s own machinery, the zonal councils and bilingual arrangements in districts of mixed population.

Q: What was the Gentlemen’s Agreement, and why did Telangana’s merger need it?

It was the agreement concluded at Delhi in February 1956 between the leaders of Andhra and of Telangana, including the two chief ministers, Burgula Ramakrishna Rao of Hyderabad and Bezawada Gopala Reddy of Andhra, on the terms of Telangana’s merger into a united Telugu state. It promised safeguards on public employment and education for the region’s people, on the application of its revenues to its own development, and a regional committee of the state legislature to watch over the region’s interests. Telangana’s merger needed it because the Commission had proposed to keep the region separate with a later option of union, and its leaders would not surrender that protection without written terms in its place.

Q: What did the States Reorganisation Act do to the office of Rajpramukh?

The Rajpramukh was the constitutional head of a Part B state, ordinarily the former ruler of the princely state at its core, as at Hyderabad, Mysore, Travancore-Cochin and the other unions. The settlement abolished the class of Part B states and with it the office: the Seventh Amendment removed the classification from the Constitution and the Act redistributed the territories, so that from 1 November 1956 every state was headed by a Governor appointed by the President. The abolition ended the last constitutional survival of the princely order inside the structure of the Union, completing in the states what accession and integration had begun in 1947.

Q: Was the map drawn in 1956 intended to be permanent?

No, and the instruments say so. Article 3 leaves the internal boundaries of the Union subject to ordinary legislation, and Article 4 provides that a law made under it is not to be treated as a constitutional amendment, a design the settlement used and did not attempt to narrow. The Commission treated its map as a settlement of the claims that had matured by 1955, not as a final geography, and its authors contemplated that lines which failed would be revisited. The map was in fact altered in 1960, in 1966, in 1971 and repeatedly afterwards, each time by the machinery the settlement of 1956 had used.