Chapter 249: The River Accounts
The Tarbela reservoir was eleven feet below its operating minimum on the morning of the twenty-second of October, 1976, and the man standing on its spillway platform, looking down at the exposed mud flats where there should have been forty feet of water, was not an engineer and was not a politician but was, in the specific calculus of Pakistani agricultural economics, more important than either.
His name was Ghulam Rasool Khokhar. He was sixty-one years old, the senior canal officer for the Punjab Irrigation Department's western circle, responsible for the water allocation schedule that determined whether four million acres of Punjabi farmland received the irrigation flow that the rabi wheat crop required between November and April. He had been a canal officer for thirty-one years. He had managed droughts. He had managed floods. He had managed the specific chronic underinvestment in Pakistan's irrigation infrastructure that every canal officer in the country had been managing since 1947.
He had never managed this.
The notification had come eleven days earlier, on the eleventh of October, through the formal channel that the Indus Waters Treaty's Permanent Indus Commission used for all communications between the two countries' water authorities: a letter from India's Commissioner for Indus Waters, addressed to Pakistan's Commissioner, stating that the Government of India was placing certain operational discretions under the Treaty's Article IV provisions — the provisions governing storage, flow timing, and the use of "natural channel flexibility" that the Treaty permitted to each riparian within defined limits — under review pending Pakistan's response to the formal accusation of state-sponsored terrorism that the Government of India had communicated through diplomatic channels on the second of October.
The letter had not said India was abrogating the Treaty. It had been very careful, in the specific bloodless language of treaty diplomacy, not to say that. What it had said was that India was exercising the maximum operational discretion available to it within the Treaty's existing framework — discretion that India had, for sixteen years since the Treaty's signing in 1960, chosen not to fully exercise, as a matter of what the letter called "good faith accommodation of downstream timing requirements."
That good faith accommodation had ended on the eleventh of October.
What this meant, in practice, on the ground, at the Tarbela reservoir and at every other point along the western rivers where Pakistan's agricultural economy depended on the timing and volume of water releases from Indian-controlled upstream works, was that India had begun managing its own reservoirs — the Chenab system works, the Jhelum diversions, the Indus headworks in the recovered territories of Gilgit-Baltistan — according to India's own optimal timing rather than according to the informal coordination that had governed releases for sixteen years.
The Treaty did not require that coordination. The Treaty specified maximum storage volumes and minimum flow guarantees and a framework for resolving disputes. Within those limits, each country had latitude. India had spent sixteen years using less latitude than the Treaty permitted, releasing water on a schedule that matched Pakistan's planting calendar even when that schedule was suboptimal for India's own hydroelectric generation and India's own emerging agricultural needs in the recovered Sindh territories.
India had stopped doing that on the eleventh of October.
The result, eleven days later, was a reservoir at Tarbela that was eleven feet below operating minimum, a canal system in Punjab province that was running at sixty percent of normal rabi-season flow, and a senior canal officer named Ghulam Rasool Khokhar standing on a spillway platform looking at exposed mud where there should have been water, doing in his head the calculation that every canal officer in Pakistan had been doing for eleven days: how many acres would not receive irrigation this season, how many tons of wheat would not be grown, and what that meant for a country whose wheat reserves were already thin enough that the previous year's harvest shortfall had required emergency imports.
He picked up the radio handset that connected him to the Irrigation Secretary's office in Lahore.
"The flow at Tarbela is sixty-one percent of scheduled," he said. "If this continues for another two weeks, we lose the planting window for approximately one point two million acres in the western circle alone."
The voice on the other end — the Irrigation Secretary himself, a man named Anwar Saeed who had held the post for four years and who had spent the previous eleven days fielding calls of exactly this kind from every canal circle in the province — said: "I know. I am telling Islamabad the same thing every day. Nobody in Islamabad is listening to engineers right now. They are listening to generals and to the Foreign Ministry."
"Tell them," Khokhar said, "that generals and the Foreign Ministry do not grow wheat."
In Islamabad, in the Prime Minister's Secretariat, the conversation that Anwar Saeed's reports were feeding into had been running, in various configurations, since the second of October — since the morning the Indian Foreign Ministry's public statement had named Brigadier Mohammed Imtiaz and Directorate S as the planners of the IC-421 hijacking and had announced the suspension of trade, the closure of the Wagah crossing, and the formal notification to the Indus Waters Treaty secretariat.
The conversation in the first two weeks had been dominated by the political and military establishment's instinct to manage the diplomatic and intelligence dimensions of the crisis — the denial, the counter-accusation, the internal attribution of the operation to unauthorised Directorate S initiative, the careful management of Brigadier Imtiaz's administrative leave in a way that protected the institutional structure without fully exonerating the individual.
The conversation about water had not been part of that management in the first two weeks, because the political and military establishment in Islamabad — General Zia-ul-Haq's faction within the Army, the ISI leadership, the Foreign Ministry — did not think about water as a strategic instrument in the way that they thought about diplomatic recognition or military posture. Water was infrastructure. Water was an engineering question that belonged to the Irrigation Ministry and the provincial governments. It was not, in the mental categories of the men managing the crisis, a front in the same conflict as the diplomatic and intelligence dimensions.
This had been a serious miscalculation, and by the twenty-second of October, the men in the Prime Minister's Secretariat had fully understood that it had been a serious miscalculation, because the political consequences of the Tarbela shortfall had begun arriving in the Secretariat in a form that was impossible to manage through diplomatic channels: telephone calls from provincial chief ministers, telegrams from the Pakistan Muslim League's Punjab organisation warning of the political consequences of a failed rabi crop in the province that was the bedrock of the party's electoral base, and — on the twentieth of October — a delegation.
The delegation was not sent by Islamabad. It had not been authorised by anyone in the federal government. It had organised itself, in the way that delegations of consequence sometimes organise themselves when the people affected by a crisis conclude that the formal channels are not adequately representing their interests, and it had driven from Lahore to Islamabad in four vehicles and had arrived at the Prime Minister's Secretariat on the morning of the twentieth of October requesting — not formally, because they had no formal standing to request anything, but requesting in the way that a sufficiently large and sufficiently agitated group of people requests something — a meeting with the Prime Minister.
The delegation was led by Tariq Bashir Chaudhry, the senior Punjab province irrigation official who had, three weeks earlier, made an unauthorised visit to the Indian High Commission in Islamabad and had asked, on his own initiative, what India would require to resolve the situation. That visit had been reported through channels and had produced, at the time, mild administrative censure — a note in his file, a verbal warning from his superior about freelancing on matters of national diplomacy. Chaudhry had absorbed the censure without complaint and had then spent the following three weeks doing the thing that career irrigation officials do when they understand a crisis better than the politicians managing it, which was continuing to track the data, continuing to calculate the consequences, and continuing to conclude that someone needed to tell Islamabad clearly what was actually at stake.
He had assembled the delegation himself: four other senior irrigation officials from across Punjab and Sindh provinces, two representatives of the Pakistan Muslim League's agricultural constituency organisations, and — the addition that gave the delegation its political weight — Malik Meraj Khalid, a senior Punjab politician with deep roots in the province's agricultural landholding class, who had agreed to lend his political standing to Chaudhry's technical urgency because Khalid had done his own calculations about what a failed rabi crop would mean for the political stability of the province he represented.
They were received, after some hours of uncertainty about whether they would be received at all, by the Prime Minister's principal secretary, who listened to Chaudhry's presentation — delivered with the specific data-driven intensity of an engineer who has spent three weeks watching a catastrophe develop in slow motion and has not been able to get anyone with authority to act on it — and who then did something that the principal secretary would later describe, privately, as the most consequential thing he had done in his career: he interrupted a Cabinet meeting that was in progress to insist that the Prime Minister hear this directly.
The meeting between the delegation and the Prime Minister happened that afternoon, the twentieth of October, in a smaller room than the formal Cabinet chamber, with fewer than a dozen people present.
Chaudhry had brought charts. He had built them himself, over the preceding three weeks, working from the daily flow data that his network of canal officers across Punjab had been reporting and that he had been compiling into a single coherent picture that no single government department had previously assembled, because no single government department had the combination of technical expertise and personal motivation that Chaudhry had developed.
He laid the first chart on the table.
"This is the Tarbela storage level," he said, "from the first of October to today." The line on the chart descended steadily, a slope that any literate person could read without needing the technical vocabulary of reservoir management. "We are eleven feet below operating minimum. At the current rate of decline, we reach the absolute minimum operating level — the level at which the dam cannot release water through controlled outlets and we are dependent entirely on natural inflow, which in October is minimal — within nine days."
He laid the second chart on the table.
"This is the irrigated acreage in the western Punjab circle that will fail to receive adequate water for rabi planting if the current flow restriction continues for thirty days." He pointed to a number circled in red: 2.8 million acres. "This is not a projection of reduced yield. This is a projection of fields that will not be planted at all, because the planting window for wheat closes in mid-November and a field that does not receive its first irrigation by then cannot be planted this season regardless of what water becomes available afterward."
He laid the third chart on the table.
"This is the wheat production consequence." Another number, circled in red: a shortfall, against the previous year's harvest, of approximately 2.1 million tonnes. "Pakistan's strategic wheat reserve currently stands at approximately 1.4 million tonnes. A shortfall of this magnitude, combined with the reserve we currently hold, produces a national wheat deficit for the coming year that requires emergency import of approximately three million tonnes to avoid civilian food shortage." He paused. "At current international wheat prices, that import requires foreign exchange expenditure of approximately four hundred million dollars, which is foreign exchange that this country's reserves cannot currently absorb without consequences for every other category of essential import — fuel, fertiliser, industrial inputs."
He looked at the Prime Minister directly, which was not standard protocol for a provincial irrigation official in a room of this seniority, but Chaudhry had concluded, over three weeks, that standard protocol was a luxury the situation did not afford.
"Prime Minister," he said. "I am not a diplomat. I do not understand the considerations that are governing the decisions about Brigadier Imtiaz and Directorate S and the question of public acknowledgement. I understand water and I understand wheat, and what I understand about water and wheat is that this country is approximately nine days from a crisis that will not be resolved by anything the Foreign Ministry does, because the Foreign Ministry cannot make it rain in the western Punjab in November, and the only entity that can release the water this country needs is the Government of India, and the only thing standing between us and that water is whatever the Government of India requires before they release it."
The room was very quiet.
Malik Meraj Khalid, the politician whose presence had given the delegation its weight, spoke next. He was a man who understood, with the specific instinct of a politician who has spent decades managing the relationship between a government and the landholding class that constitutes its political base, exactly what the consequence of a failed rabi crop would be for the political stability of Punjab province and, by extension, for the political stability of the government in Islamabad that depended on Punjab's quiescence.
"Prime Minister," he said, "I have spent forty years building relationships with the landholding families of this province. I am telling you, as someone who understands those relationships better than anyone in this room, that those families will not absorb a failed wheat season quietly. They will not wait for the diplomatic process to resolve itself on its own timeline. The political cost of this water crisis, if it is allowed to continue for even another two weeks, will exceed the political cost of whatever acknowledgement India is demanding."
He paused.
"I am not telling you what to do about Imtiaz," he said. "I do not have access to that intelligence and I am not qualified to assess it. I am telling you what I know, which is that the rabi crop is more important to the stability of this government than the question of how we frame our response to India's accusation."
The Prime Minister had been silent through both presentations, listening with the specific stillness of a man absorbing information that he had known, in the abstract, but had not understood in the visceral, specific detail that Chaudhry's charts and Khalid's political assessment now provided.
He looked at the Irrigation Secretary, Anwar Saeed, who had accompanied the delegation.
"Is the engineer's projection accurate?" he said.
"It is accurate," Saeed said. "I have my own department's data and it matches Chaudhry sahab's calculations to within margin of error."
The Prime Minister was quiet for a moment.
"What does India require?" he said.
This was the question that Chaudhry had been waiting three weeks to be asked directly by someone with the authority to act on the answer.
"I do not know the full diplomatic requirements," he said. "I made an unauthorised inquiry at the Indian High Commission three weeks ago, before I understood the severity of the water situation, and the response I received was that the requirements were stated in their public statement of the second of October." He paused. "I have read that statement carefully in the time since. The demands are: public acknowledgement of Directorate S's role, dismissal and prosecution of Brigadier Imtiaz and the officers under his command, a commitment to cease intelligence operations on Indian soil, and suspension of Directorate S's India desk operations verified by international monitoring."
"All of which are politically difficult," the Foreign Minister said, speaking for the first time.
"All of which are politically difficult," Chaudhry agreed. "I am not qualified to assess how difficult, or what alternative formulation might be acceptable to both sides. I am qualified to tell you that whatever the diplomatic resolution turns out to be, it needs to happen before the planting window closes, because after the planting window closes, the resolution of the diplomatic crisis does not un-fail the rabi crop."
The Cabinet meeting that followed, later that same day, was the meeting at which the Pakistani government's position on the IC-421 crisis genuinely changed — not because the intelligence assessment of the Directorate S operation had changed, and not because the military establishment's institutional resistance to acknowledging the operation had weakened, but because the specific, concrete, dated reality of an irrigation crisis had introduced a deadline that the diplomatic posturing of the previous three weeks had not had.
The Director General of ISI, present at this meeting in a capacity that was deliberately ambiguous about whether he was advising on the intelligence dimensions or defending the institutional position, argued — as he had argued consistently since the second of October — that a public acknowledgement of Directorate S's involvement would produce institutional damage that extended far beyond the specific personnel involved in the IC-421 operation, that it would establish a precedent for accountability that future Pakistani intelligence operations could not function under, and that the political cost of submission to India's demands would itself be severe, potentially more severe than the economic cost of the water crisis.
The Army Chief, who had been managing his own calculation throughout the crisis about the balance between institutional protection and national stability, said something that several people in the room would later describe as the turning point of the meeting.
"General," he said, addressing the ISI Director General, "I understand the institutional argument. I have made it myself, in private, for three weeks. But I want you to explain to me, specifically, what happens to this country's stability if two point eight million acres of Punjab go unplanted this season. I want you to explain what happens when the families that Malik Sahab described — the landholding families whose sons fill the ranks of this Army — discover that their wheat crop failed because we would not acknowledge an operation that, by every intelligence assessment in this room, did in fact happen exactly as India says it happened." He paused. "I am not asking you to defend the operation's wisdom. I am asking you to weigh the cost of acknowledgement against the cost of this water crisis, honestly, in front of this Cabinet, rather than in the institutional terms you have been using for three weeks."
The Director General of ISI did not have an adequate answer to this question, and his silence, in front of the full Cabinet, was itself a form of answer.
The Prime Minister made the decision that afternoon, though its implementation would take several more days to fully negotiate.
Pakistan would acknowledge.
Not in the unqualified, maximally damaging form that India's public statement had demanded, but in a formulation that the Foreign Minister, working through the night with his senior diplomatic staff, would spend the following four days negotiating through the back channel that had remained open since the second of October — the same channel through which Swaran Singh had briefed the British High Commissioner, now repurposed for direct, careful, deniable-where-necessary communication between the two foreign ministries.
The negotiation that followed, between the twenty-first and the twenty-sixth of October, was conducted primarily through two intermediary channels: the British High Commission's diplomatic good offices, which both governments had quietly requested once it became clear that direct negotiation carried domestic political risks that an intermediated process could absorb, and a direct technical channel between the two countries' Indus Waters Commissioners, who needed to negotiate the specific operational terms of any water resumption regardless of how the broader diplomatic and security questions resolved.
Karan was in Delhi for the first three days of this period, having flown from Lucknow on the twentieth at the Prime Minister's request, because the Indus Waters dimension of the crisis was, in the Cabinet's assessment, sufficiently significant and sufficiently connected to the original strategic logic that Karan had articulated in the Cabinet Committee meeting three weeks earlier that his presence in the negotiating preparation was considered necessary.
He sat in on the preparatory sessions with Swaran Singh and with India's Indus Waters Commissioner, a career water resources engineer named Brijesh Kumar Mathur who had held the post since 1973 and who understood the Treaty's technical architecture with the specific completeness of someone who had spent three years administering it.
"The question," Mathur said, in the preparatory session on the twenty-first, "is what India should ask for, now that Pakistan has signalled willingness to acknowledge."
"Not signalled," Karan said. "Decided. The Punjab delegation's visit to Islamabad and the subsequent Cabinet meeting — our intelligence on the meeting, which Rao's people obtained through the same channels that have been tracking the post-statement reaction — indicates the decision has been made internally. What remains is the negotiation of the specific formulation and the sequencing."
"Which means we have leverage we did not have three weeks ago," Mathur said.
"We have had leverage since the public statement," Karan said. "What has changed is that Pakistan now has an urgent deadline that we do not have. The planting window closes in approximately three weeks. We can wait. They cannot." He paused. "That asymmetry should inform what we ask for."
Swaran Singh, who had been reviewing the original public statement's demands, looked up. "The original demands were about the terrorism acknowledgement and the institutional accountability for Directorate S. The water question was leverage, not an end in itself — we placed the Treaty under review as a pressure instrument, not because we had a substantive grievance about the Treaty's terms."
"That was true three weeks ago," Karan said. "I want to revisit whether it should remain true."
He looked at Mathur.
"Walk me through the Treaty's allocation logic," he said. "Not the legal text. The substantive logic. Why does Pakistan get the western rivers and India get the eastern rivers, in the proportions the Treaty specifies?"
Mathur considered the question with the care of an engineer being asked to explain the foundational assumptions of a system he had spent his career operating within rather than questioning.
"The 1960 Treaty," he said, "divided the six rivers of the Indus system by allocating the three eastern rivers — the Sutlej, the Beas, and the Ravi — to India with essentially unrestricted use, and the three western rivers — the Indus, the Jhelum, and the Chenab — to Pakistan, with India retaining limited rights for irrigation of specific areas, for non-consumptive use such as hydroelectric generation, and for storage within defined limits." He paused. "The logic, at the time of negotiation, reflected the geography of partition. Pakistan's Punjab — what became West Pakistan — depended overwhelmingly on the western rivers for its existing irrigation infrastructure, infrastructure that had been built under British administration to serve a unified Punjab before partition divided the canal command areas between two countries. The Treaty was constructed to preserve Pakistan's access to the water that its existing infrastructure depended on, in exchange for India receiving full sovereignty over the eastern rivers."
"And the geography in 1960," Karan said, "did not include Indian control of Gilgit-Baltistan, POK, or any portion of Sindh."
Mathur looked at him.
"No," he said slowly. "In 1960, the entirety of the western rivers' headwaters and upper courses — the Indus through Gilgit-Baltistan, the Jhelum through POK, significant portions of the Chenab catchment — were under Pakistani administrative control or contested control. The Treaty's allocation logic assumed a geography in which Pakistan controlled essentially the entire western river system from source to mouth, with India holding only the eastern rivers and limited upstream access on the western rivers through what remained of Indian Kashmir."
"That geography no longer exists," Karan said.
The room was quiet.
"Since 1971," Karan said, "India has held Gilgit-Baltistan, Pok, and Sindh as administered territory. The Indus river's entire course from its glacial source through Gilgit-Baltistan, through the Sindh delta to the Arabian Sea — every kilometre of the Indus river itself, as opposed to its tributaries — now flows entirely through Indian-administered territory except for a stretch through Pakistani Punjab and the North-West Frontier between Attock and the Sindh border." He paused. "The Jhelum's entire upper course through Kashmir is Indian-administered. The Chenab's headwaters in the eastern Kashmir region remain as they were, but its course through Kashmir before it reaches Pakistani Punjab is now Indian-administered territory."
He looked at Mathur.
"The 1960 Treaty's fundamental geographic assumption — that Pakistan controls the western river system and India controls the eastern — has been substantially overturned by the territorial outcome of the 1971 war. We have been continuing to operate under an allocation framework that reflects a geography that no longer exists, out of what the original letter called 'good faith accommodation.'" He paused. "I am asking whether that accommodation should continue unchanged, now that we have established, through this crisis, that Pakistan understands the cost of its withdrawal."
Swaran Singh was looking at him with the specific attention of a diplomat who understands that he is hearing the outline of a negotiating position that will be considerably more ambitious than the position he had been preparing to advance.
"You are proposing to renegotiate the Treaty's fundamental allocation," he said. "Not simply resolve the current crisis."
"I am proposing that the current crisis is the occasion to correct an allocation that was always going to need correction once the territorial situation changed permanently," Karan said. "We did not renegotiate after 1971 because the relationship was not at a point where renegotiation was possible without producing exactly the kind of crisis we are now in anyway. We continued operating under the old terms because continuing was less disruptive than forcing the issue." He paused. "Pakistan has now forced the issue, through Directorate S, through IC-421. They created the crisis that makes renegotiation possible. I do not think we should resolve this crisis by returning to the 1960 terms. I think we should resolve it by establishing terms that reflect 1976 geography."
Mathur was making calculations on his notepad. "The engineering reality," he said, "is that India already has, through its administration of Gilgit-Baltistan and pok, substantially more operational control over the western rivers than the 1960 Treaty contemplated. We could, if we chose, exercise that control in ways that go considerably beyond what we have done in the past three weeks. We have chosen, for sixteen years and through the current crisis, to operate within a framework of restraint that the underlying geography no longer requires."
"What would a renegotiated allocation look like," Karan said, "that reflects current geography while remaining defensible as fair? I am not interested in an allocation so punitive that it produces humanitarian catastrophe in Pakistani Punjab and Sindh — that does not serve Indian interests, it produces instability on our border and a refugee and famine crisis that we would have to manage regardless of whose fault it is. I am interested in an allocation that reflects the actual leverage and actual geography we now have, while leaving Pakistan a sustainable agricultural water supply."
Mathur worked through this for several minutes, consulting the hydrological data that his commission maintained.
"The current Treaty allocation gives Pakistan approximately 135 million acre-feet of water annually from the western rivers, with India's permitted use — irrigation of specific areas, limited storage, hydroelectric generation — consuming a relatively small fraction of that total," he said. "A renegotiated allocation that reflected India's current upstream control could reasonably increase India's permitted consumptive use — particularly for the agricultural development of the recovered Sindh territories, which the current Treaty's geography did not anticipate India needing to irrigate, since Sindh was Pakistani territory in 1960 — while still leaving Pakistan with a flow sufficient to sustain its existing irrigated acreage, provided Pakistan accepts more disciplined water-use efficiency standards than it currently practices."
"What increase are we talking about," Karan said.
"A defensible renegotiation," Mathur said, "given the territorial change and given India's demonstrated capacity for operational discretion, could reasonably move the allocation from the current roughly 95-to-5 split in Pakistan's favour for the western rivers to something closer to 80-to-20, with the additional 15 percent directed specifically toward irrigation development in Sindh and Gilgit-Baltistan — territory that is now Indian and that has agricultural and hydroelectric potential that has been undeveloped specifically because of the old Treaty's restrictive irrigation use limits, which were designed to protect Pakistani water rights in a geography where Sindh was Pakistani."
"And the punitive mechanism," Karan said. "The clause that ensures this does not happen again."
This was the element that Swaran Singh had been waiting to discuss, because it was the element with the most complex legal and diplomatic implications.
"A standard treaty modification does not typically include a punitive snap-back clause," Swaran Singh said. "Treaties are not usually structured to allow one party to unilaterally suspend terms in response to future conduct — that undermines the basic stability that makes a treaty valuable in the first place. If we structure this as: India can suspend the agreed water flow at will if Pakistan does X, we have built an instrument that is functionally a sword over Pakistan's economy in perpetuity, which Pakistan will resist accepting and which international legal opinion may view unfavourably if the clause is too broadly drawn."
"I am not asking for an instrument that allows suspension at will for any grievance," Karan said. "I am asking for a specific, narrowly defined mechanism: if Pakistan's state intelligence services are found, through a process with defined evidentiary standards, to have directed or materially supported an act of terrorism against Indian civilians, India retains the right to suspend operational discretion under the Treaty — the same discretion we have exercised in the past three weeks — for a defined period, with the suspension subject to review by an agreed international arbitration mechanism if Pakistan disputes the finding."
He paused.
"This is not unprecedented in international instruments," he said. "Trade agreements routinely include security exception clauses. What I am proposing is the water equivalent: the Treaty's normal operation is guaranteed, full stop, as long as the underlying relationship is not characterised by one party sponsoring terrorism against the other. If that condition is violated, in a way that is verifiable rather than asserted, the guarantee is suspended for a defined remedial period."
Swaran Singh was quiet for a moment, working through the legal architecture.
"This requires careful drafting," he said. "The evidentiary standard for the suspension trigger needs to be specific enough that it cannot be abused for unrelated political purposes, but not so narrow that it fails to address exactly the kind of situation we are currently in. I would propose: suspension is triggered by either a formal finding by India's own investigative process, communicated through the same diplomatic channel used in the current crisis, which then becomes subject to review by a three-member arbitration panel — one member nominated by each country and a third agreed jointly or, failing agreement, nominated by the President of the International Court of Justice — within ninety days. If the panel does not overturn the Indian finding within ninety days, the suspension is confirmed as legitimate under the Treaty's terms."
"And if Pakistan refuses to participate in the arbitration?" Mathur asked.
"Then the suspension stands by default," Swaran Singh said. "A party that will not submit a finding to neutral review forfeits the benefit of the doubt."
Karan was nodding slowly. "This is the structure. I want it built into the renegotiated Treaty's text, not as a side letter or an informal understanding, but as an integral clause that any future Pakistani government — not just the current one — is bound by."
The technical negotiations between the Commissioners began on the twenty-third of October, conducted in Delhi rather than at the Treaty's traditional neutral venues, a deliberate signal that India was negotiating from a position that did not require the symbolic accommodation of meeting halfway.
Pakistan's delegation was led by their own Indus Waters Commissioner, a man named Fazal Muqeem Khan, accompanied by Anwar Saeed from the Irrigation Ministry and, in a late addition that signalled the political seriousness Islamabad now attached to the negotiation, by the Foreign Secretary himself.
The opening session was difficult, in the specific way that negotiations are difficult when one party understands, with complete clarity, that they are negotiating from a position of acute vulnerability and the other party understands, with equal clarity, that this vulnerability is real and is not going to be allowed to pass without consequence.
Fazal Muqeem Khan opened by requesting immediate, unconditional resumption of normal flow, pending the broader diplomatic resolution of the terrorism acknowledgement question.
Mathur's response was direct. "The flow restriction and the terrorism acknowledgement are not, in India's assessment, separate negotiations to be sequenced independently. They are connected, and the connection reflects the underlying reality that the relationship governing this Treaty cannot be insulated from the broader bilateral relationship in which one party has, by its own intelligence service's planning, conducted a terrorist operation against the other's civilians." He paused. "We are prepared to discuss the sequencing of flow resumption against specific, verifiable steps in the broader resolution. We are not prepared to discuss unconditional resumption."
"This is using water as a weapon against civilian populations," Khan said. "The farmers affected by this restriction had no involvement in whatever happened with IC-421."
"The farmers affected by IC-421's intended outcome — twenty-one hours of hostage negotiation broadcast globally to humiliate the Indian government — also had no involvement in planning it," Mathur said. "Seventy-one passengers and seven crew had no involvement. A state that plans terrorism against civilians and then objects to the affected party using every lawful instrument available to compel accountability is not in a position to claim the moral high ground on the question of who is using leverage against whom."
The Pakistani Foreign Secretary, recognising that the opening exchange was not productive, intervened.
"I want to propose," he said, "that we move past the framing dispute and address what is actually achievable. We understand the Indian position that the water question and the broader resolution are connected. We are prepared to discuss what specific, verifiable steps in the broader resolution would correspond to what specific steps in flow resumption. We need clarity on the sequence, because we have a planting deadline that the broader diplomatic resolution's pace cannot accommodate."
This was the opening that allowed the actual negotiation to begin.
Over the following three days, the two delegations worked through a sequenced framework: Pakistan would issue a formal acknowledgement of Directorate S's role in the IC-421 operation, in language that the two foreign ministries had been negotiating in parallel through the British-facilitated channel, within five days of the water negotiation's resolution. Brigadier Imtiaz would be formally relieved of his position and referred for prosecution under Pakistani military law, with the prosecution's progress subject to periodic reporting to the same channel. Directorate S's India desk operations would be suspended, subject to an international monitoring mechanism that both sides would need to negotiate separately but that would, in principle, involve a small team of neutral technical observers — the specific composition still under discussion — with access sufficient to verify compliance without compromising either country's broader intelligence sovereignty.
In exchange, India would resume normal flow timing and volume on a graduated schedule tied to verified progress on each of these steps, with full restoration of the pre-crisis operational accommodation occurring upon completion of the acknowledgement and the Imtiaz prosecution referral.
This resolved the immediate crisis.
The renegotiation of the Treaty's underlying allocation was a separate and more difficult conversation, and it was the conversation that occupied the final two days of the technical negotiation, from the twenty-fourth to the twenty-sixth of October.
Karan was present for the final two days, having returned to Delhi specifically for this phase, because the allocation renegotiation was, in his assessment and in the Prime Minister's, a strategic matter that required his direct involvement rather than delegation entirely to the technical commissioners.
He sat across the table from Fazal Muqeem Khan and the Pakistani Foreign Secretary on the morning of the twenty-fourth, with Mathur and Swaran Singh on his side of the table.
"I want to be direct about what we are proposing," Karan said, "and I want to be direct about why, so that there is no confusion in Islamabad about whether this is opportunism or principle."
He looked at the Pakistani delegation.
"The 1960 Treaty allocated the western rivers to Pakistan and the eastern rivers to India based on a geography in which Pakistan administered the entire western river system from its glacial headwaters to the Arabian Sea, with the exception of a limited area of Indian Kashmir that touched the upper Jhelum and Chenab." He paused. "That geography has not existed since December 1971. India now administers Gilgit-Baltistan, through which the Indus itself flows for its entire upper course. India now administers pok, through which the Jhelum flows for its entire upper course and through which a significant portion of the Chenab's course runs. India now administers Sindh, through which the Indus flows for its entire lower course to the sea."
He let this sit for a moment.
"The only portion of the western river system's main channels that remains under Pakistani administration is a stretch of the Indus through Punjab and the North-West Frontier between Attock and the Sindh border, and the lower reaches of the Chenab and Jhelum in Pakistani Punjab before they join the Indus." He paused. "Pakistan controls a middle segment of a river system whose source and whose mouth are both now Indian territory. The 1960 Treaty's allocation logic — built on a geography of Pakistani control from source to sea — has no remaining geographic foundation."
Fazal Muqeem Khan said, carefully: "The Treaty was not written as a geography-contingent instrument. It was written as a permanent allocation, specifically to remove water from the realm of territorial dispute and place it in the realm of fixed international law, precisely because both parties understood in 1960 that territorial circumstances could change and wanted the water arrangement to be stable regardless."
"That was the design intent," Karan agreed. "It served both parties well for sixteen years, including through a war in which India could have, at several points, used upstream control as a coercive instrument and did not. I want that restraint to be understood and acknowledged — it was not an accident, and it was not weakness. It was a deliberate choice to preserve a stable framework even as the underlying geography shifted dramatically in India's favour." He paused. "That restraint was exercised in the expectation that the relationship it preserved would not include one party's intelligence service planning terrorist operations against the other's civilian aviation. The restraint was, in effect, conditional on good faith, even though the Treaty's text did not say so explicitly. Pakistan has now demonstrated, through Directorate S, that the good faith condition does not reliably hold. We are responding by building the conditionality explicitly into the instrument, rather than relying on unstated assumptions that have now been shown to be unreliable."
He looked at the Pakistani Foreign Secretary.
"I want to address the substantive allocation question directly," he said. "We are not proposing an allocation that destroys Pakistani agriculture. That would be catastrophic for everyone, including India — a famine and refugee crisis on our western border serves no Indian interest. What we are proposing is an allocation that reflects two realities: first, that India now has substantial unmet irrigation needs in Sindh and Gilgit-Baltistan, territory that was not India's to irrigate in 1960 and is India's now; second, that the punitive mechanism we have discussed needs real water behind it to have real deterrent value — a treaty clause that allows suspension of a benefit that does not actually matter to the country receiving it provides no deterrence."
Mathur presented the specific numbers.
The proposal: a renegotiation of the western river allocation from the current effective split — approximately 95 percent of water reaching Pakistani territory with consumptive use, against roughly 5 percent reserved for India's permitted irrigation and limited storage — to a new allocation of 80 percent for Pakistan and 20 percent for India, with the additional 15 percentage points directed specifically toward two purposes: irrigation development in the Sindh agricultural network, where the Bharat Urja Mahasagar petroleum operations and the broader Shergill agricultural network's expansion had created substantial unmet irrigation demand in territory that the 1960 Treaty had never anticipated would be Indian; and hydroelectric and irrigation development in Gilgit-Baltistan, where the existing small hydroelectric installations that Aditya had described in the October financial review represented only a fraction of the territory's developable potential.
The 80-20 split would still leave Pakistan with what Mathur's hydrological modelling assessed as sufficient flow to sustain its existing irrigated acreage at current efficiency levels, provided Pakistan accepted, as a condition of the renegotiated Treaty, a set of water-use efficiency commitments — lined canals to reduce seepage loss, which Pakistan's irrigation system had historically suffered from at rates considerably higher than modern engineering standards required, and a graduated reduction in flood-irrigation practices in favour of more efficient distribution.
"You are asking us to modernise our irrigation infrastructure as a condition of accepting a reduced allocation," Khan said.
"We are observing that Pakistan's irrigation system currently loses approximately thirty to thirty-five percent of its diverted water to seepage and inefficient distribution before it reaches the crop root zone," Mathur said. "A modernisation programme that reduced that loss to twenty percent — which is achievable with infrastructure investment that India is prepared to discuss financing on commercial terms, given our own agricultural infrastructure division's relevant experience — would offset the majority of the volume reduction from the allocation change. The net effect on actual crop-available water could be considerably smaller than the headline allocation percentages suggest, if Pakistan undertakes the modernisation."
This was, Karan understood watching the exchange, the element of the proposal that gave Pakistan's negotiators something to bring home that was not simply loss — a path by which the allocation reduction's practical impact could be substantially mitigated through investment that India was prepared to help finance, which served India's interests as well, because a more efficient Pakistani irrigation system reduced the long-term political volatility of water disputes between the two countries.
The Pakistani Foreign Secretary looked at Khan, and something passed between them that Karan read correctly as the recognition that this was a negotiable proposal rather than an ultimatum, which changed the character of what followed.
"We will need to take this back to Islamabad," the Foreign Secretary said. "The allocation percentage and the efficiency investment proposal both require Cabinet-level authorisation that I do not have standing to grant in this room."
"We understand," Karan said. "I want to be clear about the timeline. The immediate crisis — the flow resumption tied to the terrorism acknowledgement sequence — resolves on the schedule we have already agreed, independent of the broader allocation renegotiation. Pakistani farmers do not need to wait for the allocation question to be resolved before water flows again at normal volumes; that resumption happens as soon as the acknowledgement and Imtiaz referral steps are completed." He paused. "The allocation renegotiation is a separate, longer-term process. I am proposing we target conclusion within six months — by April 1977, before the next rabi planting cycle's irrigation requirements begin in earnest — but I am not proposing that the current crisis remains unresolved while that longer negotiation proceeds."
This separation — decoupling the immediate humanitarian and political crisis from the longer strategic renegotiation — was, Karan understood, the element that made the proposal politically viable in Islamabad. A government that had just survived an internal crisis over whether to acknowledge Directorate S's role could not simultaneously absorb, in the same week, news that the country's water allocation was being permanently reduced. Separating the timelines gave Pakistani domestic politics room to process each element without the cumulative shock overwhelming the government's capacity to manage public reaction.
The formal acknowledgement was issued by the Government of Pakistan on the twenty-eighth of October, six days after the events recounted at the Tarbela reservoir, in language that had been negotiated word by word through the British-facilitated channel: a statement acknowledging that "elements within Pakistan's Inter-Services Intelligence Directorate S, acting under the authority of Brigadier Mohammed Imtiaz, planned and directed the hijacking of Indian Airlines Flight IC-421," that this constituted "an unauthorised and unacceptable act that the Government of Pakistan condemns," and that Brigadier Imtiaz had been "relieved of all duties and referred for prosecution under the Pakistan Army Act for conduct prejudicial to good order and the national interest."
The statement did not use the precise phrase "state-sponsored terrorism" that India's original public statement had used — this had been the single most difficult point of the language negotiation, with Pakistan's negotiators arguing that the phrase implied a level of state authorisation that the "unauthorised initiative" framing was specifically designed to avoid, and India's negotiators eventually accepting a formulation that achieved the substantive acknowledgement of Directorate S's institutional responsibility without requiring Pakistan's government to formally accept the "state-sponsored" characterisation, on the calculation that achieving the acknowledgement of fact mattered more than achieving the specific legal phrase, and that the institutional consequences — Imtiaz's prosecution, the Directorate S suspension — provided the substantive accountability regardless of the precise wording.
Water flow at Tarbela and across the western river system returned to normal scheduled volumes on the thirtieth of October, eight days after Ghulam Rasool Khokhar had stood on the spillway platform looking at exposed mud flats, in time, by Anwar Saeed's calculation, to save approximately eighty percent of the planting window that had been at risk — a partial loss, not a catastrophe, but a partial loss significant enough that every canal officer in Punjab and Sindh provinces would remember October 1976 as the month the water stopped.
In Lucknow, on the evening of the thirtieth of October, Karan sat at his desk with the final negotiated text of the immediate resolution and the framework document for the allocation renegotiation, which would occupy the technical commissions through the winter and into the spring.
Meera came in with the evening's final file, and found him looking at the framework document with the specific stillness that she had learned, over years of working with him, indicated not satisfaction exactly but the particular quality of a man assessing whether a completed thing had been done correctly rather than simply done.
"The Tarbela numbers," she said. "Eighty percent of the planting window preserved. Saeed's office confirmed it this afternoon."
"Eighty percent," Karan said. "Not all of it."
"You wanted all of it preserved," Meera said. It was not quite a question.
"I wanted the crisis resolved as quickly as it could be resolved without surrendering the leverage," Karan said. "Those two things were in tension. We resolved the tension in favour of the leverage, which cost Pakistan twenty percent of a planting window." He paused. "I think that was the correct decision. I am not certain it was costless in the way that decisions feel costless when you are confident they were correct."
Meera looked at him for a moment. "The farmers in Punjab whose fields didn't get planted this season," she said. "They are not the men who planned IC-421."
"No," Karan said. "They are not." He set the document down. "This is the genuine difficulty of using water as an instrument in a state-to-state dispute. The instrument's effects do not respect the boundary between the state that committed the offence and the citizens who had no part in it. I understood that when I proposed the strategy in the Cabinet Committee meeting three weeks ago. I understand it more specifically now, having seen Chaudhry's charts and having calculated what twenty percent of a planting window means in tonnes of wheat and in families whose income depends on that wheat."
"And the alternative?" Meera said.
"The alternative was to resolve the diplomatic crisis through statements and sanctions alone, without ever placing real pressure on the actual decision-makers in Rawalpindi who calculate, the way Imtiaz's planners calculated, that there is no cost proportional to what they are willing to do," Karan said. "Statements and sanctions did not change Pakistan's calculation in any of the previous incidents over the past decade. What changed the calculation this time — what produced an actual Cabinet meeting in which the Army Chief himself argued for acknowledgement — was the water. The water reached the people whose political weight could move the government in a way that diplomatic protest never has." He paused. "That is precisely why it is a difficult instrument to use. It works because it reaches people who are not responsible for the offence. Its effectiveness and its moral cost are the same fact viewed from two directions."
He picked up the allocation framework document.
"The renegotiation," he said, "is the part of this that I think we got right without the same moral complexity. We are not punishing Pakistani farmers with the new allocation. We are correcting an allocation that was built for a geography that no longer exists, and we are pairing the correction with an offer to help finance the efficiency improvements that mitigate most of the practical impact. If Islamabad accepts the modernisation investment, the actual water available to Pakistani crops in five years could be very close to what it is today, even at the reduced headline allocation." He paused. "That is the version of this that I am comfortable having designed. The crisis resolution — the eight days at Tarbela — I am less fully comfortable with, even though I believe it was necessary."
Meera was quiet for a moment.
"The clause," she said. "The punishment mechanism, if this happens again."
"The clause is the part of this that matters most for the next thirty years," Karan said. "Not the allocation percentage, though that matters. The clause says: if Pakistan's state intelligence apparatus does this again — plans another IC-421, sponsors another act of terrorism against Indian civilians — the consequence is not a diplomatic protest that gets filed and forgotten. The consequence is water, real and immediate, and it does not require a new negotiation each time. It is built into the Treaty's structure, triggered by a defined evidentiary process, reviewable by neutral arbitration but defaulting to India's finding if Pakistan will not participate in the review." He set the document down. "That is the actual deterrent. Not the eight days at Tarbela, which Islamabad will absorb and the political memory of which will fade. The clause is permanent. It changes the calculation for every future Brigadier Imtiaz, in every future planning room in Rawalpindi, for as long as this Treaty exists."
"Will it hold?" Meera said. "Treaties get violated. This one nearly did, this month."
"It nearly did because Pakistan judged, in 1976, that the cost of an operation like IC-421 was lower than the benefit," Karan said. "The judgment was wrong, and the eight days at Tarbela and the Cabinet meeting where the Army Chief overruled the ISI Director General are the proof that it was wrong. Whether the clause holds in 1986 or 1996 depends on whether the men making decisions in Rawalpindi at that time remember October 1976 clearly enough." He paused. "I cannot guarantee they will. What I can do is make October 1976 memorable enough, and make the mechanism real enough, that the memory has the best possible chance of mattering."
He looked at the window. Lucknow at night, the specific quiet of a city that did not know, in any detailed way, that a river system that touched the lives of tens of millions of people on both sides of a border had spent eight days at the centre of a negotiation that would shape water allocation for a generation.
"There is a meeting tomorrow," Meera said, "with Sreedharan, about the Gilgit-Baltistan hydroelectric expansion. If the renegotiated allocation includes the additional fifteen percentage points for Sindh and Gilgit-Baltistan development, the infrastructure programme needs to begin planning for it now, not after the Treaty text is finalised in April."
"I'll be there," Karan said.
"And the Sindh irrigation expansion," Meera said. "The agricultural network's field agent programme — the doubling you authorised in the October review. If the additional water allocation materialises, the network's capacity to actually use it productively becomes the constraint."
"Then we plan for both simultaneously," Karan said. "The water and the capacity to use it. There is no value in winning the allocation if the irrigation infrastructure and the agricultural network cannot translate the additional flow into additional production within a reasonable timeframe."
Meera made a note.
"One more thing," Karan said, before she left. "Chaudhry. The Pakistani irrigation officer who organised the delegation to Islamabad."
"What about him?"
"He did something that mattered," Karan said. "Not for us — for his own country, at real professional risk to himself, because he understood the technical reality better than the politicians managing the crisis and he was willing to act on that understanding even when it meant freelancing past his authority." He paused. "I don't know what happens to a man like that in Pakistan's system after a crisis like this resolves. Whether he is quietly commended or quietly punished for having been right in a way that embarrassed people above him."
"Why does it matter to us?" Meera asked.
"Because the renegotiated allocation, the efficiency modernisation investment — all of it depends on Pakistan having engineers like Chaudhry who will tell their government accurate things even when accurate things are unwelcome," Karan said. "If the system on their side punishes people for being right, the next crisis will be managed worse, not better, because the Chaudhrys of Pakistan's bureaucracy will learn to stay quiet." He paused. "There's nothing direct we can do about that. But if there is ever a channel — through the technical commission's ongoing work, through the modernisation programme if Islamabad accepts it — to make clear that India considers competent, honest engineering counterparts on the Pakistani side valuable to the relationship's stability, I want that channel used. Not as charity. As an investment in the people on the other side of this Treaty who make it function correctly."
Meera wrote this down.
"I'll raise it with Mathur," she said. "The technical commission relationship is the right channel for it."
"Good," Karan said.
She left.
Karan sat for a while longer with the two documents — the immediate resolution, signed and already in effect, the water flowing again across the western rivers as of that morning; and the framework for the longer renegotiation, which would occupy the winter months and which would, if it concluded as he expected it to conclude, establish for the first time since 1947 a water relationship between India and Pakistan that reflected the actual geography both countries now occupied, rather than the geography of a partition that had ended militarily five years earlier but had continued, in this one specific instrument, to govern as though nothing had changed.
He thought about Ghulam Rasool Khokhar, a man he had never met and would never meet, standing on a spillway platform looking at mud where water should have been, doing the arithmetic of a season's wheat crop in his head.
He thought about the boy in the seventh bed of the Gorakhpur drug treatment ward, and about Kamla Devi's hands on her son's face, and about the line that ran, in his mind, from that ward to the IC-421 cockpit to the Pentagon conference room three thousand miles away that Rumsfeld had closed two days earlier with an order to assess where Shergill Industries would be in fifteen years, to a reservoir in Pakistani Punjab eleven feet below its operating minimum.
Every part of it was connected. Every part of it was the same work, conducted across different instruments — interrogation rooms and hospital wards and Cabinet meetings and treaty texts and reservoir gauges — toward the same underlying purpose, which was a country that no longer absorbed what was done to it and called the absorption restraint, and that had, in October 1976, demonstrated across five separate fronts simultaneously that absorption was no longer the policy.
He put the documents away.
There was still work to do.
There always was.