Where We Left Off
Chapter Eight concedes three things before the verdict, and one of them removes an event most accounts of this subject place at the centre of the charge. What survives is smaller and much harder to answer.
Part Six took the northern frontier. This part takes the western one, and it is the only place in this series where a promise made in Part Two comes due.
Four findings carry into this part, and the first is a debt.
Part Two entered four forward promises into the record. One of them was the plebiscite India offered the Security Council in January 1948. Part Two said Part Seven owed an account of when that offer stopped being live. Chapter One is that account, and the answer is stranger than either side of the argument expects.
Article 370 was placed under a heading containing the word “temporary”. Part Two also left that owing. Chapter Three traces what happened to it, order by order, and finds that the interesting question is not 2019 at all.
A warning is received only where somebody has the authority to act, an incentive to believe, and a decision it changes. Part Six’s finding, from four warnings about a road. Chapter Seven applies the same test to a politician who said in 1987, in public and in almost exactly these words, what was going to happen.
Article 356 was used against an elected state government for the first time in 1959. Part Three established the precedent and said every later use stood on it. This part contains the use that mattered most, and it came six years earlier in a different form — which is why Chapter Two is where it is.
One thing about the scope. This part stops in 1989. What happened in the Valley from January 1990 onwards is a different story with a different cast, and Chapter Eight explains why a substantial piece of it does not belong in a book about Congress at all.
How to read this book
The Six Boxes
Six coloured boxes run through the series, each doing one job. Here they are with examples from the material ahead.
A Word Box explains a hard word the moment it first appears.
Presidential Order: in the context of Jammu and Kashmir, an order made by the President of India under Article 370 extending a provision of the Indian constitution to the state.
It is not an Act of Parliament. It is not debated, not voted on, and not reported. It is signed, published in the gazette, and takes effect.
Between 1954 and the 1990s there were dozens of them. Together they did more to Article 370 than anything else that has ever happened to it, and almost nobody in India — including people with passionate views about the subject — could name three.
Why it matters here: Chapter Three is about an instrument that changed a constitutional settlement by accumulation, without a single vote.
An In Real Terms box turns a number too big to picture into something with a body.
Sheikh Abdullah was imprisoned for about eleven years, from 1953 to 1964.
Eleven years is a schoolchild’s entire education. It is longer than the time between the first and second world wars. It is longer than Nehru’s remaining life after the arrest.
The man being held was the leader of the largest popular movement in the state, who had supported accession to India when it counted, who had been made head of its administration by the Maharaja at Delhi’s own urging, and who had never been convicted of anything.
A How We Actually Know This box shows the physical evidence and then says what it cannot prove.
The evidence for this part divides very cleanly, which is unusual.
The constitutional record is complete. Every Presidential Order is published in the gazette with a date and a text. Anybody can count them, read them, and see exactly which provisions were extended and when. Nothing in Chapter Three requires interpretation.
The electoral record is contested but well attested. For 1987 there are results, contemporaneous newspaper reports, named polling agents, named candidates, and admissions by participants including the beneficiary.
The diplomatic record is partly closed. What was said at Simla in the room, and what was agreed orally, rests on memoirs and on one declassified American telegram.
What none of it can show: what was in anybody’s mind. This part records decisions and their consequences and repeatedly declines to guess at motive.
An Argument box appears where serious people disagree, each side at its strongest.
Everything in Chapter One depends on this and it is not the simple question either side pretends.
India offered it, at the Security Council, in writing, in January 1948. The UN built a sequence around it in April 1948. As late as August 1953, one week after dismissing Sheikh Abdullah, India and Pakistan jointly agreed to appoint a Plebiscite Administrator within six months. That is not the conduct of a state that thinks the thing is impossible; it is the conduct of a state that keeps agreeing and never delivering. What changed was not feasibility but the expected result.
Resolution 47 required Pakistan to withdraw its irregulars first. It never did, and no plebiscite could be held on territory a foreign army occupied. Beyond that, the practical problem was insoluble: who votes? The state’s population had been displaced in both directions since 1947, the territory was split, and no register existed. A vote conducted under those conditions would have produced a result neither side accepted, which is worse than no vote.
What would settle it: nothing. The counterfactual is unmeasurable and Part One explained why.
Why people care so much: because Pakistan’s entire diplomatic case rests on the offer having been made, and India’s rests on the conditions never having been met, and both of those are accurate statements about the same document.
A Hidden Assumption box digs out a belief sitting underneath an argument that both sides accept without discussing.
Everyone who argues about Kashmir — in Delhi, in Islamabad and in Srinagar — assumes that the dispute is about Kashmir.
The Indian version: it is a territorial claim by a neighbour. The Pakistani version: it is an unresolved question of self-determination. Both treat the territory as the object.
For both states it has functioned primarily as a domestic instrument, and it has done so continuously for seventy-nine years. In Pakistan it justified the army’s central position in the state and the budget that goes with it. In India it has been available to every government as the subject on which any opposition can be called unpatriotic, and Part Nine will show what that is worth to a government in difficulty.
The test of this is what happens when it would be convenient to settle. Chapter Six describes India holding 93,000 prisoners and the strongest hand any subcontinental power has ever held, and coming away without a settlement. Chapter Nine asks whether a settlement was ever actually wanted by anybody with the power to make one.
Why neither state can drop the assumption: a government that admitted the dispute was useful to it at home would have to explain the dead. So both describe a territorial question, and both behave as though the territory were the least important thing about it.
The people for whom it genuinely is about the territory are the ones who live there, and they are the only party to the dispute who have never been asked anything.
And every chapter closes with a Remember This box: the chapter in the plainest words available.
Two debts from Part Two come due here: the plebiscite offered in January 1948, and Article 370's “temporary” heading. Chapters One and Three pay them.
The constitutional record is complete and public — every Presidential Order is in the gazette. Chapter Three requires no interpretation at all, only counting.
For both India and Pakistan, Kashmir has functioned mainly as a domestic instrument. The only party for whom it is genuinely about the territory has never been asked anything.
1The Promise That Decayed
India offered a plebiscite in January 1948 and has never held one. It has also never withdrawn the offer. There is no date on which it ended, because nothing ended it.
1.1 — What was offered
Part Two established the sequence and this chapter picks it up. India took the Kashmir dispute to the Security Council on 1 January 1948, under Article 35 of the Charter, which sits in the half that produces recommendations rather than orders.
In its complaint India said two things. That Pakistan was assisting an armed invasion of Indian territory. And that although the accession was legally complete, India was prepared to hold a plebiscite to confirm the wishes of the people once conditions allowed.
The second was volunteered. Nobody demanded it. It was offered as evidence of good faith and Part Two called it giving away the best card on the first day.
Resolution 47, in April 1948, built a sequence on it: Pakistan withdraws its irregulars, India then reduces its forces, and a plebiscite follows under United Nations administration.
The first step never happened. So on the strict terms of the resolution, India’s obligation never matured, and that is India’s answer to this day. It is a good answer and it is not the whole of the matter.
1.2 — What was still being said in 1953
Here is the fact that makes this chapter necessary.
On 16 August 1953 — one week after the events described in Chapter Two — India and Pakistan held bilateral talks in Delhi and jointly agreed to appoint a Plebiscite Administrator within six months.
Read the date twice. Five and a half years after the offer, with Pakistani forces still in occupation of part of the state and therefore with the precondition still unmet, the Indian government was agreeing bilaterally to a timetable for the machinery of a vote.
No Plebiscite Administrator was appointed. No date was set. Nothing further happened.
1.3 — When did it stop?
Part Six’s opening question for this part was: when exactly did the plebiscite stop being live, and was the change ever announced?
The answer, after going through the record, is that there is no such moment.
Nobody stood up and said the offer was withdrawn. No note was sent to the Security Council. No statement was made to Parliament announcing a change of policy. Instead the position moved by degrees, over about a decade, through a series of positions each of which was defensible on its own.
The precondition was not met, so the obligation had not arisen. Then: the state had a Constituent Assembly, which ratified the accession, and that was an expression of the people’s will. Then: circumstances had changed too much for a vote to be meaningful. Then: the matter was settled and there was nothing to discuss. Then, after 1972: everything was bilateral, and the Security Council resolutions were superseded.
Each of those is arguable. Not one of them is a withdrawal. At no point did any Indian government say: we offered this and we are no longer offering it, and here is why.
Lapsed: a promise that quietly stops applying because a condition was never met or because time passed, without anybody cancelling it.
The difference from a broken promise matters. A broken promise has a moment and somebody who broke it. A lapsed one has neither, which means there is nothing to apologise for, nobody to blame, and no date on which anybody has to explain anything.
Why it matters here: this is the most consequential lapsed promise in Indian history, and Part One warned that a promise which is neither kept nor withdrawn is a third thing that this series would name whenever it found one.
What that leaves behind is not an ordinary grievance, and the difference is worth feeling rather than merely reading.
Imagine being told, at twenty, that a decision about the place you live will be put to you.
Nobody sets a date. Every few years the assurance is repeated in slightly weaker terms. Then it stops being repeated. Nobody ever tells you it has been cancelled, so there is never a day on which you can be angry about a decision, complain about a decision, or accept a decision.
You are seventy-eight now. The thing was never refused. It was never granted. There is no letter to frame and no hearing to attend.
That is not the same experience as being told no, and the difference is the whole of the next box.
Which points at something both sides of this argument take for granted.
Everybody arguing about the plebiscite — Indian, Pakistani and Kashmiri — assumes that it was withdrawn.
The Pakistani version: India reneged. The Indian version: India was released from it because the condition failed. Both are describing an obligation that ended, and they disagree only about whether it ended honourably.
It did not end. There is no document, no date, no statement and no decision. What happened is that a promise stopped being mentioned, and then stopped being expected, and then became something only one party brought up — at which point raising it was itself evidence of hostile intent.
This matters more than it looks, and here is why. An obligation that is formally withdrawn can be argued about, compensated for, replaced with something else, or forgiven. An obligation that decays cannot be any of those things, because there is nothing to negotiate about. It leaves a permanent grievance with no procedure attached to it.
And it produces a specific and very durable political effect on the ground. A population told for years that it would be consulted, and then never told that it would not be, has not been refused. It has been left waiting, and waiting has no end and no appeal. Chapter Seven is about the last occasion on which that population tried a different method.
Why neither state will name the decay: India cannot say the offer was withdrawn without conceding it was once binding. Pakistan cannot say it lapsed without losing its case. So both describe an ending that never occurred, and the promise sits there, seventy-eight years old, technically outstanding.
One qualification before the summary, because this chapter can be read as harsher than it is. Nothing here says India was obliged to hold a plebiscite; Resolution 47's precondition genuinely failed, and that is a complete legal answer. The finding is about the form of the ending, not the fact of it. A state that had said plainly in 1954 that the offer was no longer on the table would have faced a furious argument and would have been in a much stronger position for the next seventy years than one that simply stopped mentioning it.
India volunteered the plebiscite at the Security Council in January 1948. Resolution 47 made it conditional on a Pakistani withdrawal that never happened, which is India’s answer and is a good one.
On 16 August 1953, five years later, both countries jointly agreed to appoint a Plebiscite Administrator within six months. Nothing followed.
There is no date on which the offer ended, because nothing ended it. A promise that decays leaves a grievance with no procedure attached — nothing to negotiate about, and no appeal.
21953: The Man Who Signed
The most popular leader in Kashmir supported accession when it mattered, was made head of its administration at Delhi’s urging, and spent eleven years in prison without being convicted of anything.
2.1 — Who he was
Sheikh Mohammad Abdullah led the National Conference, which had spent the 1930s and 1940s agitating against the Maharaja’s rule. It was Muslim-led, explicitly secular, and hostile to the Muslim League — which is why, in 1947, it was the single most valuable asset India had in the Valley.
Part Two set out why. India’s legal case rested on a signature by a Hindu autocrat over a Muslim-majority population. What made that case politically survivable was that the largest popular movement in the state supported it.
He was made Head of the Emergency Administration on 30 October 1947 by the Maharaja, at Delhi’s urging. He became Prime Minister of the state in March 1948.
2.2 — The Delhi Agreement
By 1952 the relationship was under strain from two directions at once, and both belong on the page.
From Delhi’s side, Abdullah was suspected of drifting towards a demand for something more than autonomy. His speeches were read as leaving the door to independence open, and Nehru — who was personally close to him — wanted a clearer commitment.
From Srinagar’s side, the pressure ran the other way. Article 370 was under attack in Indian politics from those who wanted the state absorbed like every other princely state, and Abdullah was being asked to defend an autonomy that was already being questioned in Delhi.
The result was the Delhi Agreement of 24 July 1952, which set out what the state would keep and what would pass to the Union. It conceded a good deal and it settled the question for about a year.
2.3 — 9 August 1953
Floor test: a vote in a legislature to establish whether a head of government still commands a majority.
It is the simplest procedure in parliamentary government and it exists precisely because “he has lost the confidence of his colleagues” is a claim anybody can make about anybody. The vote converts an assertion into a fact in about an hour, in public, with the result recorded.
Why it matters here: everything defensible about the events of 9 August 1953 would still have been available after a floor test, and everything indefensible about them follows from there not being one.
On 9 August 1953, Sheikh Abdullah was dismissed as Prime Minister of Jammu and Kashmir by the head of state, Karan Singh — the son of the Maharaja whose signature had brought the state into India — on the ground that he had lost the confidence of his cabinet.
He was arrested the same day. Bakshi Ghulam Mohammad, his own deputy since 1948, was installed in his place.
The constitutional objection is not complicated and it is the whole of the charge in this chapter.
He was not allowed to test his majority on the floor of the assembly.
If a head of government has lost the confidence of his legislature, there is a procedure for establishing that, and it is public, quick and unarguable. It was not used. A head of state appointed by the Union declared the confidence gone and appointed a successor, and the person best placed to disprove it was in custody before he could.
Thousands of his supporters were detained. In 1958 a prosecution was begun — the Kashmir Conspiracy Case — alleging conspiracy against the state. The trials ran for years. In 1964 Nehru had the case dropped and Abdullah released.
He was never convicted of anything. He was imprisoned for approximately eleven years and then let go, and the state that held him abandoned the case rather than losing it.
The bare facts here are not contested by anybody, including by those who defend the dismissal. The date, the ground given, the absence of a floor test, the identity of the successor, the length of the detention and the abandonment of the prosecution are all matters of record.
What is contested is Delhi’s role. That Bakshi acted with the support or knowledge of the central government is asserted by most historians of the period and denied by nobody in a form that survives examination. But there is no released Indian document ordering it, and there may never have been one — this is precisely the kind of arrangement that does not require a written instruction.
Grade: the dismissal and detention are Solid. Central authorship is Alleged, and this chapter’s charge does not depend on it, because a Union-appointed head of state acting alone is a sufficient constitutional problem on its own.
The length of the detention is easy to state and hard to feel, so here it is against something familiar.
Put the eleven years against the life of the man holding him.
Abdullah was arrested in August 1953. Nehru died in May 1964, having ordered the case dropped and the prisoner released only months before.
So the detention lasted almost exactly as long as the remainder of the life of the Prime Minister who had described him as a friend, had urged the Maharaja to install him in 1947, and had built India’s political case in Kashmir on him.
The two men had known each other for twenty years before it started. That is what makes 1953 a harder story than either the Indian or the Kashmiri version allows.
Which leaves the question this chapter exists to put.
The defence of 1953 is stronger than critics of it usually allow, and it deserves stating properly.
Abdullah had begun to talk about independence, in a state that had acceded to India, on a frontier where a war had recently been fought and where Pakistan was actively pressing a claim. A head of government of a state within the Union openly exploring secession is not an ordinary political disagreement; it is the one thing a federation cannot accommodate. His own cabinet colleagues had turned against him, which is why the ground given was loss of confidence rather than anything else, and Bakshi was not an outsider imposed by Delhi but the deputy who had served under him for five years.
Grant every word of that. There was still a procedure. A vote on the floor would have removed him lawfully, in a day, with the same outcome and none of the consequences — and if he would have won that vote, then he had not lost his cabinet’s confidence and the stated ground was false. Instead India demonstrated, in the one state whose accession rested on popular consent, that its popular leader could be removed by a Union appointee and held for eleven years without conviction. Every subsequent Indian claim that the Valley’s future would be decided by its people had to be made against that fact.
What would settle it: the numbers in the assembly in August 1953. If he commanded a majority, the stated ground was false and the case collapses entirely. Nobody has assembled the count.
Why people care so much: because Abdullah is either the man who delivered Kashmir to India and was betrayed for it, or the man who nearly took it out again — and Indian and Kashmiri accounts have needed different answers for seventy years.
One further consequence, which runs into everything after it. On 9 August 1955, exactly two years later, Abdullah’s lieutenant Mirza Afzal Beg founded the Plebiscite Front, demanding both a vote on accession and Abdullah’s release. The dismissal did not remove the demand of Chapter One from Kashmiri politics. It gave it an organisation and a martyr.
Abdullah led the movement whose support made India’s legal case politically survivable, and was made head of the administration at Delhi’s urging in October 1947.
On 9 August 1953 he was dismissed by a Union-appointed head of state for losing his cabinet’s confidence, and was not permitted to test it on the floor. He was held about eleven years and never convicted; the case was dropped in 1964.
The objective may have been necessary. Nothing in it required skipping the vote or the eleven years. And two years later the plebiscite demand acquired an organisation and a martyr.
3The Hollowing
Article 370 was not removed in 2019. It was emptied over four decades by orders that were signed rather than voted on, and by the time anybody abolished it there was very little left to abolish.
3.1 — What it originally did
Part Two set out the original arrangement. Article 370 gave Jammu and Kashmir a limited relationship with the Union: Parliament could legislate for the state only on the subjects covered by the Instrument of Accession — defence, external affairs and communications — and anything beyond that required the state government’s concurrence.
The state had its own constitution and its own constituent assembly. Its head of state was called the Sadar-i-Riyasat and was elected by the state legislature. Its head of government was called the Prime Minister.
Those two titles are worth noticing. They are not decoration. A state with a Prime Minister is making a statement about what kind of unit it is.
3.2 — The instrument
Article 370 contained the mechanism of its own dismantling, and this is the part almost nobody knows.
The article allowed the President, by order, to extend provisions of the Indian constitution to the state — with the concurrence of the state government. That mechanism was intended as a way of building the relationship gradually by agreement.
It became the way the relationship was consumed.
The Constitution (Application to Jammu and Kashmir) Order of 1954 was the first big one. It extended a large body of central provisions, and it inserted Article 35A, which protected the state’s definition of who was a permanent resident.
Dozens of further orders followed over the next four decades. Each extended something: a chapter of the constitution, an article, a jurisdiction, a central authority. Each was made under Article 370. Each required the concurrence of the state government of the day — and Chapter Two has just described how state governments in Kashmir came to be constituted.
That is the mechanism in one sentence. The centre secured a compliant state government, and the compliant state government concurred in the extension of central power. Nothing illegal occurred at any point.
Imagine a tenancy where the landlord may only enter the two rooms named in the agreement, and may enter any other room with the tenant’s written permission.
That is a real protection, as long as the tenant is somebody the landlord cannot choose.
Now suppose the landlord acquires the power to decide who the tenant is. Nothing in the agreement changes. Every entry is still by written permission and every permission is genuine. Within a few years the landlord has the run of the house, and if the tenant complains he can be shown his own signature on every single door.
One word in that mechanism carries all the weight, and it is worth stopping on before going further.
Concurrence: agreement. In Article 370 it means the state government’s formal consent to a Presidential Order extending central law.
The safeguard depends entirely on one thing: that the party giving consent is independent of the party asking for it. Consent from somebody you appointed is not a check on your power, it is a signature on your own decision.
Why it matters here: this is the hinge of the whole chapter, and it is why Chapter Two is placed before Chapter Three rather than after it.
The record of what followed is unusually easy to establish, which is worth saying because so much writing on this subject proceeds by assertion.
Every Presidential Order made under Article 370 was published in the Gazette of India, with a date, a number and a full text setting out exactly which provisions of the Indian constitution were being extended.
That means nothing in this chapter is a matter of interpretation. Anybody can obtain the list, count the orders, read what each one did, and see the cumulative effect. It is one of the best-documented constitutional transformations anywhere.
It is also, for that reason, one of the least excusable gaps in Indian public knowledge. The material has been sitting in the gazette for seventy years and the national argument about Article 370 is conducted almost entirely without reference to it.
What it cannot show: what was said when concurrence was sought. The orders record that consent was given, not how, by whom, or under what pressure.
3.3 — The two that mattered most
Two extensions deserve naming individually, because they changed the kind of thing the state was.
1964: Articles 356 and 357 were extended to Jammu and Kashmir. Until then the centre could not impose President’s Rule on the state. From 1964 it could. Part Three’s whole argument about Article 356 — that it descends from a colonial provision written so a Governor could override elected Indian ministries, and that its first political use was in 1959 — arrives in Kashmir here, five years after Kerala.
1965: the titles were changed. The Sadar-i-Riyasat, elected by the state legislature, became a Governor, appointed by the President of India. The state’s Prime Minister became a Chief Minister.
The second of those is often treated as symbolic. It was not. An elected head of state answerable to the state assembly and an appointed head of state answerable to Delhi are different offices with different loyalties, and Chapter Two has already shown what a head of state can do in that place on a single morning.
3.4 — The ones you can look up
Chapter Three’s argument is that this requires counting rather than interpreting, so here are the load-bearing orders. The list is not complete — no short table could be — and the full sequence is in the gazette for anybody who wants it.
| Year | What it did |
|---|---|
| 1950 | The first order. Applied the Union’s powers over the three accession subjects — defence, external affairs and communications — and very little else. This is the arrangement Article 370 was written to describe. |
| 1954 | The large one. Extended a substantial body of central provisions including fundamental rights with modifications, brought the state into the customs union, and inserted Article 35A protecting the state’s definition of a permanent resident. The single most consequential order of the whole sequence, and the one that established the method. |
| 1954 onwards | Dozens of further orders across four decades, each extending something: articles, chapters, central jurisdictions and authorities. Each made under Article 370. Each with the concurrence of the state government of the day. None debated or voted on in either legislature. |
| 1964 | Extended Articles 356 and 357. From this point the centre could impose President’s Rule on Jammu and Kashmir, which it could not do before. |
| 1965 | The Sadar-i-Riyasat, elected by the state legislature, becomes a Governor appointed by the President. The state’s Prime Minister becomes a Chief Minister. |
| 1986 | Extended Article 249, allowing Parliament to legislate on state subjects for Jammu and Kashmir on a Rajya Sabha resolution. Made during a period of Governor’s rule — which is to say, the concurrence required by Article 370 was given by a central appointee on behalf of a state with no elected government. |
That last row deserves a second look, because it is the mechanism of this chapter reaching its logical end point. Article 370 required the state government’s consent. By 1986 the state had no elected government, the Governor was a central appointee, and the consent was given anyway.
3.5 — What was left
By the mid-1970s the substance had largely gone. Most of the Union List applied. The Supreme Court’s jurisdiction applied. The Election Commission’s jurisdiction applied. The centre could impose President’s Rule. The head of state was a central appointee.
The 1974 accord that returned Sheikh Abdullah to office is instructive here. He came back demanding the restoration of the pre-1953 position. He did not get it. What he got was office, and the extensions stayed.
What remained after that was mostly Article 35A — the permanent resident provisions — and the symbolic fact of the article’s existence.
Every order was lawful. That is not in dispute. What is in dispute is whether a lawful process can amount to something that was never agreed.
Article 370 was headed “temporary” in the constitution’s own words, and it contained the extension mechanism deliberately, because the framers expected the relationship to deepen over time. Every order followed the procedure the article laid down and received the concurrence the article required. A state joining a federation more fully over decades is not a scandal; it is the ordinary history of federations, and demanding that a 1950 arrangement be frozen for ever is demanding something no constitution promises.
The provision was a safeguard for one party against the other, and the other party acquired the ability to determine who gave consent on the first party’s behalf — in 1953, in 1984, and by the mechanism of Article 356 after 1964. A consent requirement that the stronger party controls is not a safeguard that has been used; it is one that has been disabled. And the honest test is the one the defenders avoid: if this was integration by agreement, why was not one of these dozens of orders ever put to a vote in either legislature?
What would settle it: the concurrence files for the major orders. Not released.
Why people care so much: because both of India’s political camps have a version of 2019 that requires the previous forty years to be uninteresting.
Which is exactly what the national argument about this article assumes.
Everyone arguing about Article 370 today assumes that it was removed in 2019.
Those who celebrate it treat 2019 as the moment a historic anomaly ended. Those who mourn it treat 2019 as the moment an autonomy was destroyed. Both are describing the same event as decisive.
The autonomy Article 370 originally described had been substantially gone for forty years by then, removed piece by piece by Presidential Orders made under the article itself, with the concurrence of state governments, overwhelmingly under Congress governments at the centre, and without a single parliamentary vote on any of it.
What 2019 removed was the remainder and the name.
Notice what the assumption does for each side. It lets the article’s defenders locate the loss in a single act by a government they oppose, rather than in four decades of erosion by governments many of them supported. And it lets the article’s opponents claim credit for ending something that had already been ended, mostly by their opponents, quietly.
The general point is the one this series keeps meeting and it is not about Kashmir. A protection that can be reduced by an instrument requiring nobody’s vote will be reduced. The Ninth Schedule in Part Three had the same property. So did preventive detention in Part Two. In each case the safeguard survived on paper for decades after it had stopped operating, and the paper is what everybody argues about.
Carry one thing out of this chapter into Part Eight. The instrument described here — a lawful power, exercised repeatedly, each exercise defensible on its own, cumulatively producing something nobody voted for — is not a Kashmir instrument. It is the characteristic move of the Indian state in this series, and the next part is about the decade in which it was used on the constitution itself.
Article 370 contained the mechanism of its own dismantling: the President could extend central provisions by order, with the state government’s concurrence. Dozens of such orders were made between 1954 and the 1990s. None was voted on.
1964 brought Article 356 to Kashmir. 1965 turned an elected head of state into an appointed Governor and a Prime Minister into a Chief Minister.
2019 removed the remainder and the name. A protection that can be reduced by an instrument requiring nobody’s vote will be reduced — and the paper is what everybody argues about.
41965
Pakistan sent men into the Valley expecting a rising. The rising did not happen, and why it did not happen is the most important fact in this chapter.
4.1 — Operation Gibraltar
In August 1965 Pakistan infiltrated several thousand men across the ceasefire line into Indian Kashmir, in civilian dress, in organised groups. The plan was called Operation Gibraltar and its purpose was not conquest. It was to trigger a general uprising in the Valley, which Pakistan believed was waiting to happen.
The reasoning was not stupid, and from Rawalpindi it looked strong. Chapter Two had put the state’s most popular leader in prison. Chapter One’s promise had visibly decayed. India had just been humiliated by China. Nehru had died the previous year. The Plebiscite Front existed and had support.
4.2 — The rising that did not happen
It did not happen.
The infiltrators were not sheltered in the numbers required. In many places local people reported them to the Indian authorities. The movement collapsed within weeks and Pakistan escalated to conventional operations because the plan it was designed to support had failed.
This has to be stated plainly because it is a credit item and it is the pivot of the whole part. In 1965, offered an armed rising against India by a foreign state, the Kashmir Valley declined.
Why is genuinely arguable. Part of it is that the National Conference’s political tradition was hostile to the Muslim League and to Pakistan, and that tradition was still alive. Part of it is that Bakshi’s decade in office had brought money, construction and jobs into the Valley on a scale it had not seen. Part of it is simply that armed men from outside asking you to start a war are not attractive.
Whatever the mix, the fact is the fact, and it is the baseline against which everything in Chapter Seven has to be measured. Something changed between 1965 and 1989, and it was not Pakistan’s willingness to try.
4.3 — The war
With the infiltration failing, Pakistan launched armoured operations in the Chhamb sector in the south, aiming to cut the road link to the Valley.
India’s response was to widen the war. On 6 September 1965 Indian forces crossed the international border towards Lahore — which changed the conflict from a fight over Kashmir into a general war between the two states.
The heaviest fighting was armoured. At Asal Uttar, near Khem Karan, Indian forces destroyed a large part of Pakistan’s armoured strike force in what was then among the largest tank battles since the Second World War.
India also took the Haji Pir Pass, in the mountains north-west of the Valley — the main route by which infiltrators crossed. Remember that.
A ceasefire under United Nations auspices took effect on 23 September. Neither side had achieved its objective. India had, however, demonstrated something that mattered enormously three years after 1962: that its army worked.
Status quo ante bellum: “the situation as it was before the war”. A settlement on these terms returns everything to where it stood when the fighting started.
It sounds neutral and it is not. It is a good outcome for whichever side lost ground and a bad one for whichever side gained it, and it is what an outside mediator almost always proposes, because it is the easiest thing to get both sides to accept.
Why it matters here: this is what Tashkent produced, and it is why India handed back a mountain pass it had taken at cost.
4.4 — Tashkent
The Soviet Union brokered a settlement at Tashkent, signed on 10 January 1966 by Lal Bahadur Shastri and Ayub Khan. Both sides withdrew to the positions held before August 1965.
That meant India returned the Haji Pir Pass.
Shastri died of a heart attack in Tashkent hours after signing, and the shock of that has coloured Indian memory of the agreement ever since.
Which produced the single most argued-about clause in Indian post-war diplomacy.
The return of Haji Pir has been argued about in India for sixty years and both cases are serious.
Haji Pir was not a bargaining chip; it was the door. It is the pass through which infiltration into the Valley physically runs, and India had taken it at real cost. Handing it back restored to Pakistan the exact capability it had just used, and every subsequent infiltration campaign — including the one that mattered after 1989 — used the ground India gave away. Whatever was gained at Tashkent was not worth that.
India had not won the war; it had fought a draw and both economies were under severe strain after a failed monsoon. The mediation was Soviet, and the Soviet Union was the one great power broadly sympathetic to India at a moment when Washington had suspended arms supplies to both sides and India needed the relationship. Refusing a status-quo settlement would have meant continuing a war India was not winning, alone, in a famine year — the same famine year Part Four describes India spending on imported grain.
What would settle it: the Indian negotiating record at Tashkent. Not released.
Why people care so much: because Shastri died there, which converted an argument about a mountain pass into something much harder to discuss calmly, and it has never recovered.
And carry the other half of this chapter forward too, because it is the fact the rest of the part is measured against. Tashkent is an argument about a pass. The rising that did not happen is an argument about a population, and it is much the more important of the two.
Pakistan infiltrated the Valley in August 1965 expecting a rising. The rising did not happen. Local people reported the infiltrators and the plan collapsed within weeks.
India widened the war on 6 September, fought a draw, and took the Haji Pir Pass — the physical route infiltration runs through.
Tashkent returned everything, including Haji Pir. And the baseline stands: in 1965 the Valley declined an armed rising. Something changed by 1989, and it was not Pakistan’s willingness to try.
51971
The most complete victory any Indian government has ever achieved, and the clearest evidence in this series that the state could learn. This chapter is almost entirely a credit item.
5.1 — What happened first
In March 1971 the Pakistani army began an operation against the population of East Pakistan, following an election its own political system could not absorb. What followed was mass killing and a refugee movement of a size that has few parallels.
Something in the order of ten million people crossed into India — into West Bengal, Assam, Tripura and Meghalaya, states which Part Five’s ledger already showed as the least equipped in the country to absorb anybody.
India had, on any reading, a legitimate national interest at stake. This was not a distant humanitarian question. It was a burden being placed on Indian territory by another state’s conduct.
Refugee burden: the cost to a receiving country of people arriving with nothing — food, shelter, sanitation, disease control, and the political strain of housing a large foreign population in a poor region.
It is the reason mass displacement is treated in international law as a matter affecting the receiving state and not only the sending one. A country cannot be expected to absorb an unlimited number of people because a neighbour has made their homes unlivable.
Why it matters here: this was the legal and moral basis of India’s case in 1971, and it was a strong one. India was not intervening in somebody else’s civil war for sentiment; it was responding to a burden being placed on its own territory.
5.2 — The eight months
Here is the part that matters for this series, and it is the reason this chapter exists in a book of charges.
The crisis began in March. India went to war in December. In between, it did the work.
It supported and trained the Bengali resistance, so that when regular operations began there was an ally already inside the territory. It made a sustained diplomatic effort to explain the refugee burden internationally. It signed a treaty with the Soviet Union in August 1971, which covered the risk of great-power intervention. And it waited for the winter, when the Himalayan passes close and China cannot easily intervene from the north.
Compare that with Part Six. In 1961 a policy was adopted on an intelligence assessment, against a professional objection, with no logistics and no contingency, and executed by penny packets at altitudes nobody had acclimatised to. In 1971 the same state identified an objective, built the alliance, secured the flank, chose the season, and moved.
Whatever else this series says about the government of Indira Gandhi — and Parts Eight and Nine say a great deal — the conduct of 1971 was of a completely different order from 1962, and it was nine years later.
5.3 — Thirteen days
Full-scale war began on 3 December 1971 and ended on 16 December with the surrender of the Pakistani forces in the east at Dhaka. Around 93,000 prisoners passed into Indian hands. Bangladesh came into existence.
In the west, India also took ground — something in the region of five thousand square kilometres.
There is no comparable outcome anywhere else in the military history of independent India. A neighbour was divided, a new state was created, and the strategic balance of the subcontinent was permanently altered, in under a fortnight.
Ninety-three thousand prisoners is difficult to picture, so convert it.
It is a full sports stadium. It is more men than most countries have in their entire army. Feeding, housing, guarding and providing medical care for them is a permanent operation requiring camps, rations, doctors and thousands of Indian soldiers who are doing that instead of anything else.
It was also, for the government holding them, the largest bargaining asset any South Asian state has ever possessed — and Chapter Six is about what it was exchanged for.
Unusually for this series, almost none of that has to be taken on anybody’s word.
The 1971 war is unusually well documented from every direction: Indian and Pakistani official histories, the Pakistani government’s own Hamoodur Rahman Commission report on the defeat, the Bangladeshi record, and a very large body of contemporary international reporting on the refugee movement.
Note the contrast with Part Six. Pakistan commissioned an inquiry into its 1971 defeat and a version of it has been public since the early 2000s. India commissioned an inquiry into its 1962 defeat and it remains classified sixty-three years later. That is not a comparison in India’s favour, and it is worth recording in a series about Indian governments.
What this cannot show: the refugee figure. Ten million is the number in general use and it is an estimate of a moving population made under emergency conditions. The true figure is not knowable and Part One’s rule about counting applies.
One more observation before the summary, and it belongs in a book of charges. The government that conducted 1971 is the same government that four years later did what Part Nine describes. Competence and restraint are different things, and this chapter is evidence for the first and none at all for the second. A state that can plan eight months of preparation, secure a great power, choose a season and win in thirteen days is a capable state. What it does with that capability is a separate question, and the answer arrives in June 1975.
Around ten million refugees crossed into the least-equipped states in India. The interest at stake was real and national, not distant.
The crisis began in March and the war began in December. India used the eight months: it trained the resistance, made the diplomatic case, signed a treaty covering the great-power risk, and waited for the passes to close.
Thirteen days, 93,000 prisoners, a new country. Set beside Part Six, this is the clearest evidence in the series that the Indian state could learn — and it took nine years.
6Simla
India arrived holding more than any Indian government had ever held. It left with a principle, an oral assurance and no settlement.
6.1 — What India was holding
On 2 July 1972, Indira Gandhi and Zulfikar Ali Bhutto signed an agreement at Simla.
India held approximately 93,000 prisoners of war. It held around five thousand square kilometres of Pakistani territory in the west. It had just dismembered the other state and created a third one. Pakistan’s army was broken, its eastern wing gone, its leadership discredited and its new head of government a civilian in a country that had never let civilians last.
There has never been a moment before or since when the balance was so completely in India’s favour.
Prisoner of war: a member of an armed force captured by the other side. Under the Geneva Conventions they must be housed, fed and treated humanely, and repatriated without delay after hostilities end.
That last clause is the one that matters at Simla. Holding prisoners indefinitely as a bargaining counter is not straightforwardly available under the law of war, which weakens the “we should have kept them longer” argument more than its advocates usually acknowledge.
It also cuts the other way: the 193 men Pakistan wanted protected from trial were facing allegations of war crimes, and Bangladesh had the strongest claim on that question.
6.2 — What India wanted
India’s objective at Simla was not obscure. It wanted a final settlement of Kashmir, and specifically it wanted the line of control converted into a permanent international boundary — the division as it stood, made legal, closed.
That would have ended the dispute. Both sides would have kept what they held. It is, as it happens, approximately what most serious proposals for a settlement have looked like ever since.
6.3 — What India got
Three things, and the second is much more valuable than Indian critics of the agreement allow.
The ceasefire line became the Line of Control. Both sides undertook to respect it and not to alter it unilaterally. It was not made an international border.
Bilateralism. Both countries committed to settling their differences by peaceful means through bilateral negotiation. This is the sentence that has done the most work for India of anything signed with Pakistan. It is the basis on which India has, for fifty-four years, refused third-party mediation and treated the Security Council resolutions as superseded. Chapter One’s decaying promise finally acquired, at Simla, something like a legal roof.
Nothing on the substance. There was a commitment that the two heads of government would meet again to settle the Kashmir question finally. They did not.
India then returned the prisoners and the territory.
6.4 — The assurance that was not written down
Indian accounts of Simla contain a persistent and important claim: that on the night of 2 July, after the formal talks had deadlocked, Bhutto met Indira Gandhi privately and gave an oral assurance that the Line of Control would in time acquire the characteristics of an international border, but that he could not sign anything to that effect and survive politically at home.
India accepted the assurance. Nothing was written. Pakistan subsequently denied any such understanding, and has done so ever since.
Grade: the existence of the private meeting is Solid. The content of what was said is Alleged, resting on Indian participants’ accounts.
There is one contemporaneous document from outside both delegations. Four days after Simla, Bhutto told a visiting American official that it had been obvious at the conference that India was not interested in resolving the prisoner question on its merits but was using the prisoners and Bangladesh as leverage to settle Kashmir, and that he had not deluded himself for a minute about it.
That is a man describing what he had just successfully resisted.
The agreement itself is a published treaty and everything in section 6.3 can be read in it.
What happened in the room is a different order of evidence. The main Indian and Pakistani accounts were written by participants with reputations at stake, and they disagree. The standard scholarly work on the summit sets out both national perspectives side by side rather than choosing.
The American telegram of 6 July 1972 is the most useful single document, precisely because Bhutto had no reason to flatter anybody in it and the Americans had no stake in the Indian domestic argument. It records his own reading of the negotiation days after it ended.
What none of it can show: whether Bhutto would have signed a border settlement if India had held the prisoners longer. He said afterwards that he could not have survived it. He would say that either way.
All of which produces the largest argument in Indian foreign policy, and one where the conventional Indian answer is weaker than it sounds.
This is the largest single “what if” in Indian foreign policy and both cases are strong.
Ninety-three thousand men and five thousand square kilometres is the maximum leverage any Indian government will ever hold, and it was exchanged for a principle and an unwritten promise. India did not even get the Line of Control made permanent — the one thing that would have closed the dispute — and it returned the prisoners on a timetable rather than against performance. Every death on that frontier since 1972 has occurred because the question was left open at the one moment it could have been shut.
Three things constrain the maximalist reading. The prisoners were not solely India’s to dispose of; Bangladesh had a claim on them, including on the 193 facing possible war crimes trials, and India could not settle their fate alone. Bhutto was the only civilian leader Pakistan had, and a treaty that visibly humiliated him would have brought the army back within months and been repudiated by it — which is exactly what happened to every other agreement Pakistan’s civilians signed. And bilateralism is not a consolation prize: it removed the United Nations from the Kashmir question for fifty years, which was India’s most important diplomatic objective and which India achieved.
What would settle it: the Indian negotiating file. Fifty-four years old and not released — the third such file in this part.
Why people care so much: because 1971 is the one unambiguous triumph in the Indian record, and the argument about Simla is an argument about whether the triumph was allowed to mean anything.
Underneath that disagreement, though, both sides are relying on something about victories that is not obviously true.
Both sides of that argument assume that a military victory can be converted into a political settlement.
The critics say India failed to convert. The defenders say the exchange rate was worse than it looked. Both are treating the victory as a currency and disagreeing about the price.
What India actually held was a wasting asset with a carrying cost. Ninety-three thousand prisoners had to be fed, housed, guarded and kept alive, by an army that had other things to do, in a country that Part Four shows was importing grain. Every month of holding them made India’s position worse, not better, and both governments knew it.
And the deeper problem is what a signature from Bhutto would have been worth. A settlement is only a settlement if the party signing it can deliver its own state. Pakistan’s civilian governments have never durably been able to, which is why every agreement India has signed with a Pakistani civilian has been honoured only as long as that civilian lasted. A signed border in 1972 from a man who was hanged in 1979 would have been repudiated in 1980.
This is the uncomfortable conclusion and it cuts against the whole framing of this chapter. If the assumption is wrong — if military victory over a state whose government cannot bind it produces no convertible currency at all — then Simla was not a squandered opportunity but an accurate reading of what was on offer, and the Indian argument of the last fifty-four years has been an argument about a price that never existed.
I do not think that is entirely right, for the reason in the verdict above: the prisoners could have been released against performance rather than against a calendar, and that required no signature from anybody. But it is right enough that anybody quoting “93,000 prisoners and we got nothing” as a self-evident indictment has not thought about what they imagine India could have bought.
Simla closes the military half of this part. What follows is domestic, and it is where the charge becomes specific rather than strategic — because the decisions in Chapter Seven required no foreign party’s agreement at all.
At Simla India held 93,000 prisoners and around five thousand square kilometres — the strongest hand any Indian government has ever held.
It got the Line of Control, bilateralism — which removed the UN from the question for fifty years and is a real and lasting gain — and nothing on the substance. The assurance about the border was oral and was later denied.
The prisoners were a wasting asset with a carrying cost, and a signature from a man Pakistan’s army later hanged would not have bound Pakistan. The defensible charge is narrower: they were returned against a calendar rather than against performance.
7The Elections
A politician said in public in 1987, in almost these words, what was about to happen. Two years later it happened, and the men who said it were the ones it happened through.
7.1 — 1977, which was clean
The state election of 1977 is widely regarded, including by people highly critical of everything else in this part, as the first genuinely free election Jammu and Kashmir ever had.
It came after the Emergency, which Part Nine deals with, and in the immediate aftermath of a national government’s defeat. Sheikh Abdullah — released in 1964, returned to office in 1975 under an accord with Indira Gandhi — won it.
This matters and it should not be skipped. It establishes that a clean election in Kashmir was possible. Whatever else is argued about turnout, alienation and the impossibility of normal politics in a disputed territory, it was done, in living memory, and it produced a government.
7.2 — 1984: the elected government that was removed
Sheikh Abdullah died in 1982 and was succeeded by his son Farooq Abdullah, who won the state election of 1983 against a Congress campaign.
In July 1984 that government was brought down. A group of National Conference legislators defected, led by Farooq’s own brother-in-law, G.M. Shah. The Governor accepted the change and installed Shah as Chief Minister.
The pattern is the one Chapter Two described thirty-one years earlier, in a milder form: an elected Chief Minister removed through the office of a Governor appointed by Delhi, in a state where the centre’s party had just lost an election.
Shah’s government was itself dismissed in 1986 and the state went under Governor’s rule. Then, in late 1986, an accord was reached between Rajiv Gandhi and Farooq Abdullah: Farooq returned to office, and the National Conference and Congress would contest the next election in alliance.
Hold that last fact. The party that had removed him in 1984 was now his partner, and the two of them together were the government going into 1987.
7.3 — 23 March 1987
The Muslim United Front was a coalition of religious and local parties formed in 1987. It was not a separatist front standing on a boycott; it was contesting the election, which is precisely what every Indian government had spent forty years asking Kashmiris to do instead of anything else.
It drew large crowds. Pre-poll assessments suggested it might take somewhere between fifteen and twenty seats in the Valley.
Turnout was around eighty per cent in the Valley — the highest ever recorded there.
What happened on and after polling day on 23 March 1987 is documented in unusual detail for a disputed election.
Opposition polling agents were ejected from counting stations. Ballot boxes were removed under security escort. Over a hundred agents were arrested in one constituency alone. Pre-stamped ballots were found. In Amira Kadal in Srinagar, the MUF candidate Mohammad Yusuf Shah led the early count and was declared defeated by the National Conference incumbent by 4,289 votes.
The MUF was officially credited with four seats.
Farooq Abdullah — the beneficiary — has himself acknowledged the rigging.
This is better evidenced than most contested elections anywhere, for three reasons.
First, the losing candidates were named public figures who described what happened at the time, in the Indian press, in Indian courts and in Indian political language. Abdul Ghani Lone, a mainstream pro-India politician contesting in Handwara, gave a public account of police ejecting his agents, removing ballot boxes and arresting over a hundred people.
Second, the beneficiary has admitted it. An election is not usually disputed by the man who won.
Third, the counting-station accounts come from people whose names, constituencies and roles are known — including one polling agent whose subsequent career made him among the most closely documented individuals in the region.
What it cannot show: how many seats the MUF would have won. Pre-poll estimates of fifteen to twenty are estimates. What is Solid is the conduct at the booths and the count; the counterfactual result is Alleged, and this chapter does not need it.
The techniques described above have names, and the names matter because they identify who has to be involved.
Booth capturing: taking physical control of a polling station in order to cast or destroy votes. It was common enough in Indian elections of this period to have its own name.
The related techniques are ejecting the opposition’s polling agents, so nobody hostile is watching; removing ballot boxes from the count under escort; and stamping ballots in advance.
Every one of these requires the cooperation or absence of the police, which is why they are almost never done by a party that does not control the state government.
Why it matters here: all four were reported on 23 March 1987 by named people in named constituencies.
7.4 — What the losing candidates did next
This is the section that makes 1987 different from any other rigged election in Indian history, and it can be told almost entirely in biographies.
Mohammad Yusuf Shah, the MUF candidate in Amira Kadal, crossed into Pakistan-administered territory and became Syed Salahuddin, commander of the Hizbul Mujahideen — the largest militant organisation of the insurgency.
Yasin Malik, who had worked as his polling agent, crossed the Line of Control in 1988 for training and became a leader of the Jammu and Kashmir Liberation Front, which began the armed campaign.
Others who crossed in the same period had been in the same campaign.
And Abdul Ghani Lone, the mainstream politician whose agents were arrested at Handwara, said publicly at the time — in words widely reported and never retracted — that if people were not allowed to cast their votes, their anger would go into anti-national feeling instead.
That is a warning, delivered in public, by a pro-India politician, in the year it mattered, naming the mechanism and the consequence.
Part Six established the test: a warning is received only where somebody has the authority to act, an incentive to believe, and a decision it changes. Lone’s warning failed all three, and it failed them in the presence of a national press that reported it.
Eighty per cent turnout is not a number Indians should skim past.
It is higher than any Indian general election has ever recorded. It is higher than the United States manages in a presidential year. It is the figure of a population that has decided, collectively and at some risk, to use the method it was offered.
Then set beside it the figure from 1989 onwards, when turnout in the Valley collapsed to single digits in places and stayed there for years.
Between those two numbers there are twenty-four months and one count.
Which raises the causal question that dominates every discussion of modern Kashmir.
This is the central causal claim about modern Kashmir and it is genuinely contested.
Pakistan had been trying to start an insurgency in Kashmir since 1947 and had run a full infiltration campaign in 1965. The 1980s gave it something it had never had before: a decade of experience arming and training an insurgency in Afghanistan, with weapons, camps and money supplied at scale by outside powers. That capability arrived independently of anything that happened at a Srinagar counting station. Blaming an Indian election for a campaign planned in another country flatters Indian self-importance and lets Pakistan off entirely.
Nobody claims 1987 built the training camps. The claim is about who walked into them. Compare 1965: identical Pakistani intent, an infiltration force already in the Valley, and the population declined and reported them. The difference by 1989 was not on the Pakistani side of the line. It was that a specific cohort of young men had just tried the constitutional method, at the highest turnout ever recorded there, and had it taken from them at the count — and the two most consequential militant leaders of the following decade were literally a candidate and his polling agent from that election.
What would settle it: the recruitment biographies of the first militant cohort — how many were MUF workers in 1987. Partial work exists and points strongly one way; nobody has done it systematically.
Why people care so much: because one answer makes the insurgency a foreign attack on India and the other makes it partly a consequence of an Indian decision, and Indian politics has needed the first answer continuously since 1990.
Notice that both of those positions are counting seats. That is the shared move, and it is the wrong measure.
Everybody arguing about 1987 assumes that the rigging mattered because it stole seats.
The critical version: the MUF was robbed of fifteen or twenty seats and the theft caused the rising. The defensive version: the MUF would not have won that many and the theft was smaller than claimed. Both are arguments about the arithmetic of the result.
The seats were the least of it. What 1987 destroyed was not a set of election victories but the argument that participation works — and it destroyed it in the hands of the specific people who had just made that argument to everybody else.
Consider what a young MUF worker had spent that spring doing. He had told his friends that the way to change things was to organise, campaign and vote. He had put his name on it. Eighty per cent of the Valley turned out, which is a higher figure than most established democracies achieve. And then the count was interfered with in front of him, and he was arrested for objecting.
A man in that position has not merely lost an election. He has been publicly shown to have been wrong about the method, by the state that told him to use it, in front of the people he persuaded.
This is why the seat arithmetic is beside the point, and why the defensive version of the argument — that the MUF would only have won six or eight — is no defence at all. The damage was not proportional to the number of seats taken. It was done by the visibility of the taking.
And it generalises well beyond Kashmir. A rigged election in a settled democracy produces a wrong government. A rigged election in a place where the legitimacy of the state is itself the question produces something else entirely, because the vote was never only about who governs.
One thing this chapter deliberately does not say. It does not say that the men who crossed the Line of Control were justified in what they did afterwards, or that the killing which followed was a proportionate response to an interfered count. It says that a state which asks a population to use elections instead of guns has taken on a specific obligation about elections, and that in March 1987 it did not meet it.
1977 was clean, which proves it was possible. In 1984 an elected government was brought down through engineered defections and a Governor. By 1986 the party that removed it was its election partner.
On 23 March 1987, at the highest turnout ever recorded in the Valley — around eighty per cent — agents were ejected, boxes removed, over a hundred arrested in one constituency, and pre-stamped ballots found. The winner has admitted it.
The MUF candidate at Amira Kadal became the commander of Hizbul Mujahideen. His polling agent became a leader of the JKLF. A mainstream politician said publicly at the time exactly where the anger would go.
8What Does Not Belong Here
Three concessions before the verdict. One of them removes from this book the single event most people associate with the charge.
8.1 — January 1990 was not a Congress government
The images that dominate any discussion of Kashmir’s collapse are from the winter of 1989 and 1990: the beginning of organised militancy, the killings, and the flight of the Kashmiri Pandits from the Valley.
None of that happened under a Congress government.
Congress lost the general election of November 1989. The government from December 1989 was the National Front under V.P. Singh, supported from outside by the Bharatiya Janata Party and by the Left.
In December 1989 the Union Home Minister’s daughter was kidnapped and militants were released in exchange, which is generally identified as the moment the insurgency acquired confidence. That Home Minister was Mufti Mohammad Sayeed, in the National Front government. In January 1990 that government appointed Jagmohan as Governor. The Pandit exodus followed within weeks.
Grade: Solid, and conceded without qualification. A series that attributes those months to Congress is lying, and a great deal of what circulates on this subject does exactly that.
What survives the concession is the same shape as Part Five’s. The hole was dug earlier. The specific cohort that made 1990 possible was produced in 1987, by a state government in alliance with the party then governing at the centre, and the constitutional conditions were built between 1953 and 1965. The National Front inherited a Valley in which the constitutional method had been publicly discredited twenty-one months earlier.
8.2 — Pakistan has agency
The second concession is larger and it is one this series should make explicitly, because a book about Indian governments can slide into writing as though India were the only actor.
Pakistan tried to take Kashmir by force in 1947, by infiltration in 1965, and by supporting an insurgency from 1989 onwards. It did all of that on its own initiative, for its own reasons, and it would have kept trying whatever India did.
Nothing in Chapters One to Seven caused that. India did not create the training camps, the weapons pipelines or the strategic doctrine behind them, and the argument in Chapter Seven explicitly concedes that the means of the insurgency were external.
What this series can say about India is what India controlled, and that is a narrower and more useful subject than the whole history of the dispute.
Agency: the capacity of a person or a state to act for its own reasons, rather than only in response to what somebody else does.
It is a dull word doing important work. An account that explains everything Pakistan did as a reaction to India has denied Pakistan agency, and an account that explains everything Kashmiris did as a reaction to Delhi has denied them agency too — which is a form of condescension that often arrives disguised as sympathy.
Why it matters here: this series is about Indian governments, and the discipline that requires is remembering that other people in the story were also deciding things.
8.3 — Kashmiri politics is not only a reaction to Delhi
The third concession is the one most likely to be resented and it is also true.
The National Conference was itself accused of rigging elections from the 1950s onwards, well before 1987 and sometimes against opponents who were not separatists. Corruption in the state’s administration was a real grievance with local authors. The rise of religious politics in the Valley in the 1980s had causes of its own, including regional and global ones, that had nothing to do with Indian election management.
Treating every development in Kashmiri politics as a response to something Delhi did is its own kind of condescension. It denies the place any politics of its own.
The honest audit, conducted before the verdict rather than after it.
Take the concessions seriously. The collapse happened under a non-Congress government. The means came from Pakistan. The local politics had local authors. What remains is a dismissal in 1953, a series of constitutional orders, a rigged election in 1987 conducted by a state government, and a general charge that India governed a disputed territory the way states govern disputed territories. Every part of that is real and none of it is unique to one party or to Kashmir.
Every concession is about a later or an external cause and none of them touches what India did or when. The plebiscite decayed without ever being withdrawn. The state’s most popular leader was removed without a floor test and held eleven years without conviction. The autonomy was hollowed by dozens of orders that nobody voted on. An elected government was brought down in 1984 through engineered defections. And the 1987 election was conducted by an alliance of the state party and the national governing party. None of that becomes untrue because V.P. Singh was Prime Minister in January 1990.
What would settle it: the recruitment study named in Chapter Seven.
Why people care so much: because the concessions are useful to one side of Indian politics and the instruments to the other, and almost nobody has an interest in stating both.
8.4 — The ledger for this part
Part Five committed every subsequent part to carry the running record of which states were affected by its material. Here is the entry.
| Where | What was done |
|---|---|
| Jammu and Kashmir | A plebiscite offered and never withdrawn. Its head of government dismissed without a floor test in 1953 and held eleven years. Its autonomy reduced by dozens of Presidential Orders between 1954 and the 1990s, none voted on. Article 356 extended in 1964; its elected head of state converted to an appointed Governor in 1965. An elected government brought down in 1984. An election conducted in 1987 in a manner its own beneficiary has admitted. |
| Punjab | The 1965 war was fought across it, and the Haji Pir Pass taken and returned. Part Ten takes the rest. |
| West Bengal, Assam, Tripura, Meghalaya | Absorbed something in the order of ten million refugees in 1971 — the same states Part Five’s ledger identified as least equipped to absorb anybody, doing it for the second time in twenty-four years. |
| Rajasthan and Gujarat | The western theatre in 1965 and 1971. Territory taken in 1971 and returned at Simla. |
January 1990 was not a Congress government. The National Front was in office, Mufti Mohammad Sayeed was Home Minister, and it appointed Jagmohan as Governor. Conceded without qualification.
Pakistan has agency and would have kept trying whatever India did. It supplied the means. And Kashmiri politics has local authors, which treating everything as a reaction to Delhi denies.
What survives: India’s conduct did not create the insurgency, and it produced the conditions in which a foreign attempt that failed in 1965 succeeded in 1989.
9What Would Have Prevented It
The defence at full strength, and a question this part has been circling since Chapter One: was a settlement ever actually wanted by anybody who could deliver one?
9.1 — Four arguments for the defence
Here are the strongest arguments a serious defender of the Indian record would make. Two of them are very strong.
One. The territory was disputed and the other party was armed. Every measure criticised in this part was taken in a place where a foreign state was actively pressing a territorial claim by force. Governments do not run disputed frontiers the way they run settled interiors, anywhere in the world, and holding India to a standard nobody meets is not analysis.
Two. It held. This is the strong one. Kashmir is still in India. Given a hostile neighbour with a permanent claim, three wars, an insurgency lasting decades, and a genuine crisis of legitimacy in the Valley, the state did not lose the territory and did not lose the rest of the country either. Compare the record of other multi-ethnic post-colonial states facing a secessionist movement backed by a neighbour, and the comparison is not close.
Three. Democracy kept being attempted. Elections were held in 1977 and were clean. The state had a legislature, courts, a press and mainstream parties throughout — none of which is true of the territory on the other side of the line, where no comparable election has ever been held at all.
Four. Simla removed the United Nations. Chapter Six conceded it and it is worth repeating here, because it is the single most durable diplomatic achievement in this part. The dispute has been kept bilateral for fifty-four years, and every Indian government since has depended on a sentence Indira Gandhi obtained in 1972.
The second of those defences is the one this part cannot answer and does not try to. Whatever was done between 1953 and 1987, the union held, and there is a version of the counterfactual in which it did not.
9.2 — The settlements that were available
Part One requires an alternative that somebody proposed at the time. There were three.
1953. A floor test instead of a dismissal. It would have taken a day. If Abdullah had lost it, the outcome would have been identical and lawful. If he had won it, the stated ground for removing him was false — which is the point.
1972. The prisoners released against performance rather than against a calendar. Chapter Six’s verdict: this required no signature from anybody and was not attempted.
1987. A clean count. The MUF would have held some seats in a state assembly, which is what a state assembly is for. This is the cheapest alternative in the entire series and it was available on a single day in March.
None of the three required a change of policy on Kashmir, a concession to Pakistan, or any risk to the union. All three were procedural. That is what makes them charges rather than second-guessing.
9.3 — What I would have to retract
Three things would break this part.
If the assembly numbers of August 1953 showed that Abdullah had genuinely lost his majority, Chapter Two weakens to a procedural complaint about a correct outcome.
If a study of the first militant cohort showed that few of them had any connection to the 1987 campaign, Chapter Seven’s causal claim collapses and the external explanation takes the whole weight.
If the Indian negotiating file from Simla showed that staged release against performance was proposed and refused by Pakistan, Chapter Six’s narrow charge disappears.
The first and third depend on files the Indian government holds. The second is ordinary academic work that anybody could do.
The three alternatives in section 9.2 are unusual in this series because none of them is a counterfactual about outcomes. Each is a claim about a procedure that existed and was not used, and procedures leave records.
A floor test was available in 1953 because the assembly existed and was sitting. Staged release was available in 1972 because prisoner repatriation was necessarily a phased operation taking months, which means the phasing was a variable somebody set. A clean count was available in 1987 because counts are conducted by identifiable officials under rules.
That is why Chapter Nine can name alternatives with more confidence than Part Four could about the economy. Part One’s test asks whether the alternative was available at the time; for a procedure, availability is a matter of record rather than of modelling.
What it cannot show: whether taking any of the three would have changed the outcome. A cleaner 1987 might still have produced an insurgency. The claim is that the alternative existed, not that it was a guarantee.
Which leads to the thing this whole part has been circling, and it is not a question about Kashmir.
Everyone who writes about Kashmir — Indian, Pakistani and Kashmiri — assumes that the alienation of the Valley was caused by things that happened in the Valley.
The critical version lists local events: 1953, 1984, 1987. The defensive version lists other local events: Pakistani infiltration, religious mobilisation, local corruption. Both are looking for the cause inside the same borders where the effect appeared, which is exactly the error Part Five’s front matter identified about regional poverty.
Look at the instruments instead of the events. A head of government removed by a centrally appointed head of state without a floor test. A constitutional protection reduced by executive orders requiring no vote. Article 356. An elected government brought down by engineered defections. An election managed by the party governing at the centre.
Not one of those is a Kashmir instrument. Every single one appears elsewhere in this series, applied to places with no dispute, no insurgency and no foreign claim. Part Three found Article 356 used against Kerala in 1959. Part Five found a Prime Minister ending a live Supreme Court case over Punjab’s rivers in 1981. Part Three found the Ninth Schedule growing from thirteen entries to two hundred and fifty by exactly the mechanism Chapter Three describes here.
Kashmir did not get special instruments. It got the ordinary instruments of the Indian state, earlier and harder than anywhere else, in the one place where the population’s consent was the whole basis of the claim.
Which reframes the question this part has been asking. “Why did Kashmir alienate” has an answer that is not about Kashmir. The right question is why the same instruments produced an insurgency there and merely produced resentment in Kerala and Punjab — and the answer is not that Kashmiris are different. It is that in Kerala and Punjab the state’s right to be there was never the thing at issue, so the instruments damaged a government’s standing. In Kashmir they damaged the claim itself, every time, because the claim rested on consent and each use was a demonstration that consent was not required.
That is the case against this part. The second defence — that the union held — is the one I cannot answer, and I want to be clear that I am not answering it. There is a version of the last seventy-nine years in which a state less willing to use these instruments loses more than a reputation. What this part claims is narrower: that the three alternatives in section 9.2 cost nothing, risked nothing, and were not taken.
Four defences: the territory was disputed and the neighbour was armed; the union held; democracy kept being attempted and 1977 was clean; and Simla removed the UN for fifty years. The second is the one this part cannot answer.
Three alternatives, all procedural and all cheap: a floor test in 1953, release against performance in 1972, a clean count in 1987. None required a policy change or any risk to the union.
Kashmir did not get special instruments. It got the ordinary ones, earlier and harder, in the one place where consent was the whole basis of the claim.
10An Honest List of What We Do Not Know
Two of this part’s central claims rest on counts nobody has made, and both could be made by anybody who wanted to.
10.1 — Genuinely unknown
Six things in this part cannot be established. Two of them are closed files, two are unmade counts, and two are genuinely hard.
Whether Sheikh Abdullah held a majority in August 1953. He was dismissed for losing his cabinet’s confidence and was not allowed to test it.
Why it is unknown: nobody has assembled the numbers. This is an unmade count, not a closed file — the composition of the assembly and the known alignments are public, and a serious researcher could settle it. It would decide Chapter Two.
How many of the first militant cohort came out of the 1987 campaign. Chapter Seven’s causal claim rests on the connection between the MUF’s workers and the men who crossed the Line of Control in 1988 and 1989.
Why it is unknown: also an unmade count. The biographies of the leadership are well known and point strongly one way. The rank and file have never been studied systematically. This is the single most useful piece of work anybody could do on modern Kashmir and it requires no government’s permission.
What was said between Indira Gandhi and Bhutto on the night of 2 July 1972. Indian accounts describe an oral assurance about the Line of Control. Pakistan denies it.
Why it is unknown: two people, no record, both dead. The nearest independent evidence is an American telegram recording Bhutto’s own account four days later, and he had reasons of his own for what he said in it.
Whether staged release of the prisoners was proposed at Simla. Chapter Nine names it as the available alternative.
Why it is unknown: the Indian negotiating file is closed, fifty-four years on. The third such file in this part.
What the plebiscite would have produced. A counterfactual, and Part One explained why those cannot be measured. The additional problem here is that the electorate itself is undefined: the state’s population moved in both directions after 1947 and the territory is divided.
Whether the MUF would have won fifteen seats or six. Pre-poll estimates are estimates.
Why it is unknown: the count was interfered with, which destroys the evidence of what the count would have been. Chapter Seven’s hidden assumption argues this does not matter, because the damage was done by the visibility of the interference and not by its size.
10.2 — Solid
The plebiscite chronology. India’s letter of 1 January 1948, Resolution 47 of April 1948 and its sequence, and the joint agreement of 16 August 1953 to appoint a Plebiscite Administrator within six months. All record. And the absence of any subsequent withdrawal, which is established by there being no document to cite.
The dismissal of 1953. The date, the ground given, the absence of a floor test, Bakshi’s installation, the detention of roughly eleven years, the Kashmir Conspiracy Case begun in 1958 and dropped in 1964, and the fact of no conviction.
Every Presidential Order. Published in the gazette with a date and a text. The 1954 Order and Article 35A. The extension of Articles 356 and 357 in 1964. The change of nomenclature in 1965. Nothing in Chapter Three requires interpretation; it requires counting.
The 1965 war. Operation Gibraltar, the failure of the rising, the crossing of the international border on 6 September, Asal Uttar, the capture of the Haji Pir Pass, the ceasefire of 23 September, and the Tashkent Declaration of 10 January 1966 restoring the pre-war positions.
The 1971 war. The refugee movement, the Indo-Soviet treaty of August 1971, the thirteen days from 3 to 16 December, the surrender at Dhaka, roughly 93,000 prisoners and around five thousand square kilometres of territory in the west.
The Simla Agreement. Signed 2 July 1972. Published treaty. The Line of Control, the bilateralism clause, the absence of any substantive Kashmir settlement, and the return of the prisoners and the territory.
The 1984 removal and the 1987 election. The defections and the installation of G.M. Shah. The 1986 accord and the National Conference–Congress alliance. Polling on 23 March 1987, turnout around eighty per cent, four seats to the MUF, the documented conduct at booths and counts, and Farooq Abdullah’s own acknowledgement.
The subsequent careers. The MUF candidate at Amira Kadal and his polling agent are among the most thoroughly documented individuals in the region.
That January 1990 was under a National Front government. Election results and the composition of a Union cabinet.
10.3 — The two counts nobody has made
Most of what this series cannot establish is locked in government files. This part is unusual, because its two most consequential gaps are not.
The assembly arithmetic of August 1953 would decide whether the stated ground for the most consequential dismissal in the state’s history was true. The material is public.
The recruitment biographies of the first militant cohort would decide the causal claim in Chapter Seven, on which most of this part’s weight rests. The material exists in court records, in journalism, in the accounts of the men themselves, and in the memory of a generation that is still alive.
Neither requires declassification. Neither requires access to anything. Both have been available to anybody for decades and neither has been done, and the reason is not difficulty.
It is that the answer to the first would embarrass one national account and the answer to the second would embarrass the other. A question nobody wants answered stays open indefinitely, whatever the state of the archives — and that is a different mechanism from the one in Part Six, and worth naming separately.
Genuinely unknown: whether Abdullah held a majority, how many first-cohort militants came from the 1987 campaign, what was said on the night of 2 July 1972, whether staged release was proposed, what a plebiscite would have produced, and how many seats the MUF would have won.
Solid: the plebiscite chronology and the absence of any withdrawal, the 1953 dismissal, every Presidential Order, the three wars, the Simla text, the 1984 removal, the conduct of 23 March 1987, and that January 1990 was a National Front government.
The two most important gaps are not closed files. They are counts nobody has made, using material anybody could reach — and each would embarrass a different national account.
Sources & further reading — Part 7
Timeline, 1947–1990
Dates only. The arguments are in the chapters.
| Date | What happened |
|---|---|
| 30 Oct 1947 | Sheikh Abdullah made Head of the Emergency Administration by the Maharaja, at Delhi’s urging. |
| 1 Jan 1948 | India takes the dispute to the Security Council and volunteers a plebiscite to confirm the accession. |
| 21 Apr 1948 | Resolution 47: Pakistan withdraws first, then India reduces, then a plebiscite. The first step never happens. |
| Jan 1949 | Ceasefire. The line stands where the armies stopped. |
| 24 Jul 1952 | The Delhi Agreement between Nehru and Abdullah sets out what the state keeps and what passes to the Union. |
| 9 Aug 1953 | Abdullah dismissed for losing his cabinet’s confidence, without a floor test, and arrested. Bakshi Ghulam Mohammad installed. |
| 16 Aug 1953 | India and Pakistan jointly agree to appoint a Plebiscite Administrator within six months. Nothing follows. |
| 1954 | The Constitution (Application to Jammu and Kashmir) Order extends a large body of central provisions and inserts Article 35A. Dozens more follow over four decades. |
| 1951 | A Constituent Assembly is elected in Jammu and Kashmir to frame the state’s own constitution. |
| Jan 1957 | The state’s constitution comes into force, declaring Jammu and Kashmir an integral part of India. India treats this as the people’s ratification of the accession; Pakistan and the Security Council do not accept it as a substitute for a plebiscite. |
| 9 Aug 1955 | The Plebiscite Front is founded, demanding a vote on accession and Abdullah’s release. |
| 1958 | The Kashmir Conspiracy Case is begun against Abdullah. |
| 1964 | Articles 356 and 357 are extended to Jammu and Kashmir. The centre can now impose President’s Rule. The conspiracy case is dropped and Abdullah is released after about eleven years. |
| 1965 | The Sadar-i-Riyasat, elected by the state legislature, becomes a Governor appointed by the President. The state’s Prime Minister becomes a Chief Minister. |
| Aug 1965 | Operation Gibraltar. Pakistani infiltrators enter the Valley expecting a rising. It does not happen; many are reported by local people. |
| 6 Sep 1965 | India crosses the international border towards Lahore. Heavy armoured fighting follows at Asal Uttar. India takes the Haji Pir Pass. |
| 23 Sep 1965 | Ceasefire under United Nations auspices. |
| 10 Jan 1966 | The Tashkent Declaration restores pre-war positions. Haji Pir is returned. Shastri dies hours later. |
| Mar 1971 | The crisis in East Pakistan begins. Around ten million refugees cross into India over the following months. |
| Aug 1971 | The Indo-Soviet treaty is signed, covering the risk of great-power intervention. |
| 3–16 Dec 1971 | The war. Surrender at Dhaka. About 93,000 prisoners. Bangladesh comes into existence. |
| 2 Jul 1972 | The Simla Agreement. Line of Control, bilateralism, no substantive settlement. Prisoners and territory returned. |
| 1972–74 | Talks between Delhi and Sheikh Abdullah, conducted through intermediaries, on the terms of his return to politics. |
| 1975 | Sheikh Abdullah returns to office under an accord with Indira Gandhi. The pre-1953 position is not restored. |
| 1977 | A state election widely regarded as the first genuinely free one in the state’s history. |
| 1982–83 | Sheikh Abdullah dies. Farooq Abdullah succeeds him and wins the 1983 election against Congress. |
| Jul 1984 | Farooq Abdullah’s government is brought down by defections and the Governor installs G.M. Shah. |
| 1986 | Shah’s government is dismissed. An accord between Rajiv Gandhi and Farooq Abdullah restores Farooq and creates a National Conference–Congress alliance. |
| 23 Mar 1987 | The state election. Turnout around eighty per cent. Agents ejected, boxes removed, over a hundred arrested in one constituency, pre-stamped ballots found. The MUF is credited with four seats. |
| 1988 | MUF figures including a defeated candidate and his polling agent cross the Line of Control for training. |
| Nov–Dec 1989 | Congress loses the general election. The National Front takes office. The Home Minister’s daughter is kidnapped and militants released in exchange. |
| Jan 1990 | The National Front government appoints Jagmohan as Governor. The insurgency and the Pandit exodus follow. Not a Congress government. |
Part Eight starts here. It returns to Delhi in 1966, with a new Prime Minister nobody expected to last, and follows the nine years in which she took the party, the courts and the constitution apart and put herself where they had been.
Glossary
Every hard word used in this part, in plain English.
| Word | What it means |
|---|---|
| Article 35A | The provision, inserted by Presidential Order in 1954, protecting the state’s own definition of who counted as a permanent resident. |
| Article 356 | The provision allowing the centre to dismiss an elected state government. Extended to Jammu and Kashmir in 1964. |
| Agency | The capacity to act for one’s own reasons rather than only in reaction to somebody else. Denying it to Pakistan or to Kashmiris is a common error in accounts written from Delhi. |
| Bilateralism | The principle, established at Simla in 1972, that India and Pakistan settle their disputes between themselves without outside mediation. India’s most durable diplomatic gain in this part. |
| Booth capturing | Taking physical control of a polling station to cast or destroy votes. Requires the cooperation or absence of the police, which is why it is almost never done by a party out of power in the state. |
| Concurrence | The state government’s formal consent to a Presidential Order under Article 370. A safeguard only if the consenting party is independent of the party asking. |
| Floor test | A vote in a legislature establishing whether a head of government still commands a majority. Public, quick and unarguable. Not used in August 1953. |
| Instrument of Accession | The document by which a princely ruler joined India, transferring three subjects only: defence, external affairs and communications. |
| Lapsed | A promise that stops applying because a condition was never met or time passed, without anybody cancelling it. It has no date and nobody to blame. |
| Line of Control | The military line dividing Jammu and Kashmir, created from the 1971 ceasefire line at Simla. Not an international border, and both sides undertook not to alter it unilaterally. |
| Muslim United Front | A coalition of religious and local parties formed in 1987 which contested the state election rather than boycotting it. Officially credited with four seats. |
| Prisoner of war | A member of an armed force captured by the other side. Under the Geneva Conventions they must be treated humanely and repatriated without delay once hostilities end. |
| Plebiscite | A direct vote of a population on a single question, such as which country to join. Offered by India in January 1948 and never held. |
| Presidential Order | An order extending a provision of the Indian constitution to Jammu and Kashmir under Article 370. Signed and gazetted, not debated or voted on. |
| Sadar-i-Riyasat | The state’s head of state, elected by its legislature, until the title was replaced by an appointed Governor in 1965. |
| Refugee burden | The cost to a receiving country of people arriving with nothing. The legal and moral basis of India’s case in 1971. |
| Status quo ante bellum | “The situation as it was before the war”. A settlement returning everything to where it stood when fighting began — good for whoever lost ground, bad for whoever gained it. |
What Comes Next
Part Eight — Indira Rising, 1966 to 1975
Part Eight returns to Delhi and takes the nine years in which the party, the courts and the constitution were rearranged around one person. It asks:
- She was chosen in 1966 because a group of party bosses thought she would be manageable. What did they think they were getting, and at what point did they discover otherwise?
- The party split in 1969. What was the split actually about, who won what, and what happened to the organisation that had chosen every leader since 1885?
- Bank nationalisation and the abolition of the privy purses were both enormously popular. Part Two owes a verdict on the second — was withdrawing a constitutional guarantee given as the price of accession justified?
- Judges were superseded in 1973 for the first time in Indian history. Who was passed over, on what reasoning, and what did the court do afterwards?
- Garibi Hatao won an election on a promise about poverty. Measured against the government’s own targets, what arrived?
- By June 1975 what remained of the checks Part Three found disappearing in 1954 — and what happened when a judge in Allahabad used the last one?
Questions, not topics. If any of them is answered badly, you will be able to tell.