On the evening of 22 July 2008, the fourteenth Lok Sabha divided on a motion of confidence that the government of Manmohan Singh had moved in its own name, and the 2008 nuclear deal ceased to be a negotiation conducted in Washington and Vienna and became, in full public view, a domestic crisis conducted on the floor of the House. The motion carried by 275 votes to 256, with ten members abstaining, after a two-day debate in which three members of the Bharatiya Janata Party had carried bundles of currency into the chamber and placed them before the Speaker as proof, they said, that votes for the government were being bought. The claim defended in this article is that the crisis, and not the strategic realignment that the agreement with the United States also represented, is the reading that explains the most about what the deal did to Indian politics. A foreign policy instrument that began as a joint statement signed in Washington on 18 July 2005 ended by destroying the coalition arrangement that had governed India since 2004, expelling a Speaker from his own party, putting a prime minister’s continuation in office to a counted division, and leaving a scandal whose investigations ran for five years without settling the question they were opened to answer. The trajectory has four components, and each component is a domestic event before it is a diplomatic one: a statement, a statute, a withdrawal and a vote.

The 2008 Nuclear Deal and the Corridor That Led to It

The corridor to July 2005 was built inside the politics of the preceding decade, and the deal’s domestic meaning cannot be recovered without it. The Pokhran II tests of 11 and 13 May 1998, conducted by the government of Atal Bihari Vajpayee, had closed one era of Indian nuclear policy and opened a negotiation that the tests themselves made necessary, a sequence traced in the account of the Vajpayee-era architectural-political-moment framework. Between June 1998 and September 2000, Jaswant Singh for India and Strobe Talbott for the United States met fourteen times in seven countries, a dialogue Talbott reconstructed in Engaging India, published in 2004, in which the American side pressed for restraint instruments and the Indian side pressed for acceptance of a nuclear reality that sanctions had failed to reverse. The dialogue produced no agreement, and Talbott’s own printed verdict conceded that Jaswant Singh had come closer to his objective. What the dialogue did produce was a shared vocabulary and a cadre of officials on both sides who carried the file forward when the political conditions changed, so that the opening of 2005 was an inheritance and not an improvisation.

Three dated developments converted that inheritance into an opening. On 18 March 2005, in Washington, the two governments announced a wider strategic partnership in the Next Steps framework that the Vajpayee and Clinton governments had launched in January 2004. In March 2005 the United States also moved on the regional file that Indian negotiators watched most closely, confirming the sale of F-16 aircraft to Pakistan while holding out to India a broader defence and technology relationship, a pairing that New Delhi read as evidence that Washington intended a durable tilt rather than a transaction. Inside India, the verdict of May 2004 had installed a coalition whose foreign policy room was narrower than its predecessor’s, a constraint the account of the post-2004 UPA mandate sets out in full: the United Progressive Alliance governed with the outside backing of the Left parties, whose combined strength of fifty-nine members in the Lok Sabha was larger than the government’s own margin over the opposition, and whose programme documents reserved their sharpest language for American power. The Common Minimum Programme adopted in May 2004 committed the coalition to an independent foreign policy and to the pursuit of closer ties with the United States in the same breath, and the tension between those two sentences sat inside the arrangement from its first day, waiting for an issue large enough to tear along it.

The strategic reading of the deal, given its fullest scholarly form by Sumit Ganguly and Rahul Mukherji in India Since 1980, published in 2011, treats the agreement as the capstone of a post-1991 reorientation in which India traded the defensive posture of the Cold War decades for integration with the American-led order. The policy-trajectory reading, carried in the essays of Harsh V. Pant across the years after 2008, follows the negotiation file by file: statement, statute, agreement text, safeguards, waiver. Both readings are true to their evidence, and neither is false. The argument here is narrower and, on the record, stronger: neither reading explains why a trade and technology settlement consumed a government, because the explanation sits in the third register, the domestic one, where the deal was fought as a question about who governed India and on whose sufferance. That register is the subject of the framework this pillar applies across its ten articles, stated in full in the account of the foreign-policy-as-domestic-politics framework, and July 2008 is its hardest test case, because no other episode in the pillar joins an external negotiation to an internal collapse with so complete a paper trail.

What did India offer to place under safeguards in the March 2006 separation plan?

India offered a phased separation under which fourteen of its twenty-two thermal power reactors would be placed under International Atomic Energy Agency safeguards in sequence between 2006 and 2014, while eight reactors, the fast breeder programme and the military facilities stayed outside, and future reactors would be classified by India alone as civilian or military.

The separation plan belonged to the interval between the statement and the statute, and its domestic reception previewed the crisis. When Manmohan Singh laid the plan before Parliament in March 2006, the scientific leadership whose consent the plan required spoke in public with unusual sharpness: Anil Kakodkar, chairman of the Atomic Energy Commission, had already placed on record his objection to any separation that touched the fast breeder programme, and former chairmen of the commission wrote and spoke against terms they judged to mortgage the strategic programme. The Prime Minister answered in Parliament on 7 March 2006 and again in August 2006 with a set of assurances that hardened into the government’s negotiating floor: no capping of the strategic programme, no acceptance of safeguards on the breeder reactors, reciprocity in every step, and full civil nuclear cooperation including fuel supply assurances in exchange. Those assurances, given to the House to hold his own coalition, were later read back to him by every side in the trust vote debate, and the distance between the assurances and the American statute became the ground on which the crisis was fought.

The Arrangement of 2004: Power Without Office

To understand why a reactor agreement could destroy a government, the arrangement it destroyed has to be described in its own terms, because the crisis of 2008 was a crisis of the arrangement before it was a crisis of the deal. The verdict of May 2004 gave the Congress 145 seats in the Lok Sabha, seven fewer than the Bharatiya Janata Party had won in defeat in some earlier cycles and far short of a majority, and it gave the four Left parties fifty-nine seats, the largest Left bloc in the history of the republic. The coalition that took office under Manmohan Singh therefore rested on a double dependence. The United Progressive Alliance’s own constituents supplied the ministers, and the Left supplied the margin, from outside the government, without portfolios, without collective responsibility, and with a published programme, the Common Minimum Programme of May 2004, as the arrangement’s only written constitution. The mechanics of the arrangement are traced in the account of the coalition era’s long arc, and its foreign policy clauses matter here: the programme promised closer relations with the United States and, in adjacent sentences, an independent foreign policy and a refusal to join any design that subordinated Indian decisions to another power, and both promises were quoted back, verbatim, by both sides in July 2008.

Outside backing on this scale was a new instrument, and its properties were discovered only when it failed. A coalition partner inside a cabinet can be bound by collective responsibility and compensated with office; a backer outside can be bound by nothing except the programme and its own calculation of the cost of an election. The Left’s leaders stated their theory of the arrangement openly and repeatedly between 2004 and 2008: their fifty-nine members gave them a veto on policy that touched their stated positions, exercised through a coordination committee that met across the four years and through the threat, always implicit and twice made explicit, of withdrawal. The Congress leadership’s theory, stated with equal openness by its floor managers, was the opposite: consultation was owed, a veto was not, and a government that surrendered its foreign policy to a partner without office would have surrendered the substance of governing. Neither theory was ever put to the House in a form that could settle it, because the arrangement worked, for four years, on every question except the one that finally broke it. It survived the disinvestment quarrels of 2004 and 2005, in which the Left’s objections modified and delayed the sale of stakes in public enterprises, and it survived the pension and banking bills that the Left stalled in committee, because on those questions the government could accept delay as the price of the margin. A foreign agreement with dates attached, negotiated against an American legislative calendar and an agency timetable in Vienna, was the first question on which delay and surrender were the same act, and it is not an accident of the story that the arrangement broke on the first question it could not postpone.

The personnel of the arrangement sharpened the structural flaw. Manmohan Singh was a Prime Minister without a personal following in the House, chosen by the Congress president, Sonia Gandhi, after the verdict of 2004, and his authority inside the coalition rested on the party’s backing and on a reputation for integrity that every camp conceded. Prakash Karat, who led the largest Left party from April 2005, was a general secretary of a different temper from his predecessor Harkishan Singh Surjeet, whose long practice had been to keep the Left’s leverage liquid and its threats unspoken; Karat’s practice, stated in his own writings on the period, was to convert leverage into lines, drawn in documents, in advance, so that the government could never claim to have misread them. Between a Prime Minister whose standing depended on never being seen to bargain away a stated conviction and a party secretary whose standing depended on never being seen to bargain away a stated line, the coordination machinery ground through 2005 and 2006 on the nuclear file with diminishing room to move. The deal did not create the rigidity; it supplied the occasion on which two rigidities, each honourable in its own terms, met on a question with a deadline. That is the first proposition the crisis reading adds to the record, and it is a proposition about institutions, not about reactors.

Component One: The Joint Statement of 18 July 2005

The first component is a single document of a few pages, and its brevity is part of its politics. On 18 July 2005, in Washington, Prime Minister Manmohan Singh and President George W. Bush issued a joint statement that announced a transformed relationship and, inside it, a bargain on civil nuclear energy whose terms were stated as reciprocal undertakings. The United States undertook to work toward full civil nuclear cooperation with India, to seek the adjustment of American laws and policies to that end, and to work with allies to adjust the guidelines of the Nuclear Suppliers Group so that trade closed to India since the nineteen seventies could reopen. India undertook, in the statement’s sequence, to assume the responsibilities and practices of the advanced nuclear states: to identify and separate its civilian and military facilities in a phased manner, to file a declaration and accept agency safeguards on the civilian side, to maintain a unilateral moratorium on testing, to work toward a fissile material treaty, and to keep its export controls aligned with the regimes it had not signed. The statement named no dates for the reciprocal steps and fixed no order in which they would fall due, and that openness, serviceable in July 2005, became the contested space of the next three years, because each side’s domestic critics filled the silence with their own sequence and accused their own government of surrendering the difference.

The domestic announcement was managed as carefully as the text. Manmohan Singh reported to Parliament on 29 July 2005, within days of his return, and placed the government’s reading of the bargain on the parliamentary record before any other capital’s reading could harden: reciprocity, he told the House, was the governing principle, and India would not accept obligations that the statement had not named. The Left parties, whose votes in the House kept his government in office, answered within the same season with a critique that never altered its essentials between 2005 and 2008. In the Left’s documents, the statement was not a bargain but a hinge: it joined India to an American strategic design for Asia, priced in the currency of an independent foreign policy that the coalition’s own Common Minimum Programme had promised to defend. Prakash Karat, who had taken over as general secretary of the Communist Party of India (Marxist) in April 2005, made the nuclear file the test of that promise, and the party’s notes to the government across 2005 and 2006 returned to the same three counts: the separation would expose the strategic programme to pressure, the safeguards would be in perpetuity, and the alignment would be paid for in votes cast against Iran at the agency in Vienna. Each count was documented, dated and argued, and the government’s replies, carried mostly by the minister of external affairs, Pranab Mukherjee, conceded the facts in dispute less often than they disputed the weight the Left placed on them.

Beneath the parliamentary exchange ran a second domestic contest that the strategic reading habitually underplays: the contest over credit and ownership inside the Indian state. The Bharatiya Janata Party, in opposition since May 2004, attacked the statement from a position its own record had built, since the architecture of engagement with Washington, from the Talbott dialogue to the Next Steps announcement of January 2004, had been raised by a government the party had led. Yashwant Sinha and Arun Shourie, speaking for the opposition, charged that the Singh government was accepting terms, on separation and safeguards, that the Vajpayee government had refused, while surrendering the bargaining asset of a completed programme for promises that required an American statute nobody had yet drafted. The charge stung because it was half true: the 2005 statement did commit India to a separation the earlier dialogue had discussed without concluding. The government’s answer, that the statement’s reciprocity language protected every Indian interest the opposition named, was also half true, and the two half truths defined the argumentative field on which the Hyde Act landed eighteen months later. A bargain whose meaning both sides could contest at home had been signed abroad, and the crisis of 2008 was, in that precise sense, drafted into the settlement of 2005.

The statement’s first test came eight months later, on the ground where the bargain had to be made physical. President Bush visited New Delhi in March 2006, and on 2 March 2006 the two governments announced the separation plan that the statement had promised, the division of India’s nuclear complex into a civilian list that would accept safeguards and a military list that would not, with the classification of future reactors reserved to India. The announcement was staged as the visit’s centrepiece, and its domestic management repeated the pattern of July 2005: the Prime Minister carried the plan to Parliament within the week, the atomic energy leadership’s reservations were voiced in public by serving and retired chairmen of the commission, and the Left’s notes treated the visit itself, the reception accorded to a president whose Iraq war the Left had marched against in 2003, as evidence of the alignment the statement had begun. The visit also fixed the calendar that governed the rest of the trajectory, since the administration in Washington required the separation announcement in hand before it could carry the Hyde legislation to the Congress, and the Indian government had therefore spent its first instalment of the bargain, the plan, before receiving any instalment in return. The asymmetry of that sequence, plan first and statute later, assurance first and waiver last, was the Left’s most durable argument, and the government’s acceptance of the sequence, defended as the only order in which the pieces could move, is the point at which the trajectory’s managers chose momentum over insurance, a choice the crisis of 2008 collected with interest.

Component Two: The Hyde Act and the Price Written in American Law

The second component is the point at which the bargain acquired a text that Indian politics could not control and could not stop reading. Because the Atomic Energy Act of 1954 barred the cooperation the statement promised, the Bush administration required a statute, and on 18 December 2006 President Bush signed the Henry J. Hyde United States-India Peaceful Atomic Energy Cooperation Act of 2006, Public Law 109-401, which authorised the President to waive the barring provisions for India subject to a schedule of determinations, reports and conditions. The Act’s architecture mattered more in New Delhi than in Washington. Its statement of policy carried language on securing India’s support for American efforts regarding Iran’s nuclear programme, on India’s participation in the Proliferation Security Initiative, and on discouraging the transfer of enrichment and reprocessing items to states that lacked them; its reporting provisions required the President to keep the Congress informed on India’s compliance, its fissile production and its votes and conduct in the agency; and its waiver authority was hedged by a determination, repeated in the process, that cooperation served American non-proliferation ends. At the signing, President Bush added a statement treating portions of the Act, including its foreign policy passages, as advisory rather than binding on the executive, a gloss that Indian officials cited for years afterwards and that Indian critics dismissed as a courtesy no Indian court or parliament could enforce.

The Act’s own passage through the American legislature had already taught New Delhi to read statutes as signals. The House of Representatives had passed the bill in July 2006 by 359 votes to 68, the Senate in November 2006 by 85 votes to 12, and the conference text that settled the two versions carried amendments, on Iran, on reporting and on the treatment of a future Indian test, that the administration had resisted and accepted in sequence. Indian officials tracked the tallies and the amendments the way earlier generations had tracked treaty ratifications, and the lesson the foreign office drew in its briefings to the political leadership was double-edged: the margins showed a bipartisan American commitment to the opening that no change of administration was likely to reverse, and the amendments showed that the commitment would be priced afresh at every stage, in a legislature whose conditions India could influence only by lobbying and never by right. The Hyde Act debate in the Lok Sabha was thus the mirror of a debate in Washington that had already happened, and the members who cited its clauses by number were, without quite saying so, conducting India’s share of a legislative process in which India had no seat. The point is the framework’s point in its sharpest form: the foreign policy had become domestic politics in two countries at once, and the crisis detonated in the one where the government’s majority, and not merely its policy, was the stake.

The Left’s reading of the statute was set down in writing and never revised. In the notes the Left parties submitted through the coordination machinery and in the statements of Karat and of A. B. Bardhan of the Communist Party of India, the Hyde Act was held to convert the reciprocal undertakings of July 2005 into a unilateral American discipline: India would be reported on, certified and judged, while the American obligations remained matters of presidential intention. The Iran passages drew the sharpest fire, because the government’s critics could point to a sequence that seemed to prove the discipline was already operating. On 24 September 2005, and again on 4 February 2006, India had voted at the agency’s board in Vienna to find Iran in non-compliance and then to report the file to the Security Council, votes the government defended on the merits of the Iranian record and the Left denounced as the first instalment of the alignment’s price. The votes preceded the Hyde Act, and the government’s defenders said so, correctly; the Left’s answer, that the votes showed the direction of travel the statute would merely codify, was a reading of trajectory rather than of text, and it was on trajectory that the argument was joined for the following eighteen months.

Why did the Left parties call the Hyde Act a curb on an independent foreign policy?

The Left parties held that the Act made American cooperation conditional on reports and determinations about India’s conduct, named India’s alignment on Iran as an object of American policy, and so converted a commercial opening into a standing lever over Indian decisions, which their documents described as a surrender of the independent foreign policy the coalition programme had promised.

Behind the statute stood an older Indian memory that the debate kept invoking and that explains the depth of the suspicion the Act aroused. The Tarapur plant, built with American cooperation in the nineteen sixties, had been promised fuel under a 1963 agreement, and after the test of 1974 and the American Non-Proliferation Act of 1978, the United States had moved to cut that supply, leaving India to litigate the promise for years and to draw the lesson, repeated by the atomic energy leadership in evidence and in print for three decades, that American fuel assurances carried a congressional expiry date. When the Hyde Act’s critics in the Lok Sabha demanded fuel supply guarantees that no future Congress could disturb, they were arguing from Tarapur, and when the government’s negotiators answered that the 123 text’s reserve and disruption clauses were drafted against exactly that history, they were conceding the history’s force. The statute debate was therefore never only about the clauses of 2006; it was about whether a promise broken in 1978 could be made good by a text in 2007, and the distrust that question carried into the chamber was earned long before Manmohan Singh or George W. Bush held office.

The government answered the statute with a negotiation. Through the first half of 2007, Indian and American negotiators worked on the bilateral cooperation text, the 123 Agreement, whose number came from the section of the American Atomic Energy Act under which such agreements are concluded, and the talks stuck for months on the same three rocks the parliamentary debate had named: India’s right to reprocess American-origin fuel, the consequences of a future Indian test for fuel supply and for the return of supplied items, and the character of the fuel supply assurances the United States would give. The eventual text, concluded on 27 July 2007 and released on 3 August 2007, gave India consent to reprocess under safeguards in a dedicated facility, recorded American support for an Indian strategic fuel reserve and for agency-arranged supply in the event of disruption, and stated the parties’ understandings on testing in language each capital could present at home as a victory. In Washington the text was received as the outer limit of concession. In New Delhi it was received by the Left, within days, as a document that could not cure the statute above it, and the coordination committee through which the coalition had managed its quarrels since 2004 was converted, in September 2007, into a committee on the deal itself, chaired by Pranab Mukherjee and charged with examining the Hyde Act’s implications and the Left’s objections. The committee met through the autumn and recorded its findings in November 2007 without reconciling the two readings, and the file it could not settle passed, by the logic of the arrangement, to the arithmetic of the House.

The deeper point about the Hyde Act, for the domestic-crisis reading, is that a foreign statute became a domestic constitutional actor. No clause of the Act bound India in law, as the government repeated in every forum; the Act bound the American President, and bound him to report on India. Yet the Indian argument of 2006 to 2008 was conducted as though the Act were an Indian text under amendment, with clauses cited by number in the Lok Sabha, assurances demanded against its operation, and a coalition’s survival staked on its interpretation. That inversion is the framework’s signature in this pillar: the pattern first described for the Nehru years in the study of Nehru’s foreign policy and its domestic uses recurs here in statutory form, where the external instrument does not merely influence domestic politics but supplies the very text over which domestic politics fights. The comparison with an earlier treaty moment is instructive and is drawn out in the account of the architectural-substantive-domestic-signal framework: the Indo-Soviet Treaty of August 1971 sent a signal inward as well as outward, but its text was India’s own, negotiated and owned; the Hyde Act was another country’s text, unamendable in New Delhi, and the helplessness of that position, felt by government and opposition alike, is what raised the temperature of every subsequent step.

The Year of Committees: The Argument of 2007

Between the statute and the withdrawal sits a year in which the argument was conducted, on both sides, with a completeness that later accounts, hurrying to the vote, tend to skip, and the year matters because it fixed the positions from which neither side could retreat in July 2008. The release of the 123 text on 3 August 2007 opened the phase. The government presented the text as the answer to every count in the Left’s indictment: reprocessing consent had been secured, fuel supply assurances recorded, the testing question handled in language that preserved India’s freedom of action while giving the American President the findings his statute required. Pranab Mukherjee, speaking in the Lok Sabha during the debate on the agreement in November 2007, restated the assurances of 2006 and added the government’s structural answer to the Hyde Act objection, that the bilateral agreement and the safeguards text were the only instruments that would bind India, that the Act bound only the American executive, and that in the event of any inconsistency the agreement’s terms, negotiated and signed by India, were the terms India would honour. The opposition benches replied from two directions that never merged. The Bharatiya Janata Party, speaking through L. K. Advani and its former external affairs minister Yashwant Sinha, objected that the text surrendered the option of future testing in substance while pretending to preserve it in form, and that a party which had been denied the same settlement in office would not ratify its rival’s version of it. The Left objected to the frame itself: the question, in Karat’s formulation across the party’s statements that autumn, was not whether the 123 text was skilfully drafted but whether India should enter a regime of American certification at all, and no drafting could answer a question of that order.

The committee device of September 2007, the panel chaired by Mukherjee with the Left’s representatives inside it, was the arrangement’s last invention, and its proceedings deserve the weight the participants gave them. The committee was asked to examine the Hyde Act’s bearing on the agreement and on India’s foreign and security policy, and to record findings that both sides could carry back to their organisations. It met across September, October and November 2007, heard the negotiators and the atomic energy leadership, and produced findings that the government read as a clearance and the Left read as a record of its objections. Sitaram Yechury, the Communist Party of India (Marxist) leader who sat in the committee and carried the party’s case in the Rajya Sabha, stated the distinction in terms that decided the sequel: the committee had been promised that the government would not operationalise the agreement until its findings were considered, the findings had not accepted the government’s case, and operationalisation without acceptance would end the arrangement. The government’s counter-reading, stated by Mukherjee, was that consideration had occurred, that the committee had been a forum and not a court, and that the final judgement on a negotiation rested with the cabinet. In December 2007 the debate moved to the floor, where a short-duration discussion in both Houses rehearsed every argument of the autumn without a vote, and the session closed with the positions published, the lines drawn and the trigger, the approach to the agency in Vienna, identified by both sides as the act that would decide the arrangement’s life. The first half of 2008 was therefore not a drift toward crisis but a loaded pause, in which the government completed the safeguards text with the agency’s secretariat, the text was circulated to the agency’s board in July, and the Left’s central committee, meeting in June, converted its threat into an instruction with the date left blank.

One further element of 2007 belongs to the record because it explains the ferocity of the following July: the argument over the deal became, inside both camps, an argument over the credibility of assurances as such. The Prime Minister’s assurances to Parliament of March and August 2006 had been given to hold the House, and the Left’s documents of 2007 quoted them clause by clause against the Hyde Act’s clauses, demanding to know which text the House had been asked to trust. The government’s defenders answered that assurances govern a government’s conduct and statutes govern another state’s, and that a parliament which could not hold the distinction could not conduct a foreign policy at all. The dispute sounds technical and was not: it was a dispute over whether the word of a government given to its own legislature could survive contact with a negotiation, and every parliamentarian who spoke in the trust vote debate returned to it, because the confidence being voted on was, in the plainest sense, confidence in that word. The crisis reading takes its name from this accumulation. By July 2008 the deal carried four years of programme language, two years of assurances, a foreign statute, a committee’s contested findings and a party secretary’s published lines, and the vote that settled it was asked to settle all of them at once.

Component Three: The Withdrawal of 9 July 2008

The third component is the shortest in time and the heaviest in consequence. The committee chaired by Pranab Mukherjee bought the coalition one winter. Its findings, shared with the Left in November 2007, recorded the government’s case that the 123 text protected every interest the assurances of 2006 had named, and recorded, in parallel, the Left’s refusal to withdraw its objection to the Hyde Act beneath the text. Manmohan Singh, in an interview given in August 2007, had already supplied the crisis with its epigraph, telling a newspaper that he had told the Left leadership that it was free to withdraw its backing if the deal was so objectionable, and that the government would not be deflected. The remark, remembered verbatim by every participant in the later quarrel, signalled that the Prime Minister had priced the coalition below the agreement, and from that season the question inside the arrangement was not whether the deal would proceed but which side would move first and be recorded as having broken the coalition over it. The Left’s own machinery moved in June 2008: the central committee of the Communist Party of India (Marxist), meeting from 20 to 22 June 2008, authorised its polit bureau to withdraw the parties’ backing at the moment the government approached the agency in Vienna to operationalise the safeguards agreement, the step that would set the international sequence in motion.

The trigger came from Tokyo. Manmohan Singh travelled to Japan for the Group of Eight summit at Toyako in the first week of July 2008, and on 7 July, speaking to the journalists travelling with him, he said that the government would approach the agency very soon, words that the Left read, as they were plainly intended to be read, as the announcement that the operational step was upon them. On 8 July the four Left parties, the Communist Party of India (Marxist), the Communist Party of India, the Revolutionary Socialist Party and the All India Forward Bloc, announced the withdrawal of their outside backing, and on 9 July 2008 their leaders delivered the letters of withdrawal to President Pratibha Patil at Rashtrapati Bhavan. The letters ended an arrangement of four years and two months. In their statements the parties were exact about their ground: the government, they said, had chosen the deal over the programme on which the coalition rested, had refused to place the 123 text and the safeguards draft before Parliament for a sense of the House, and had treated the Hyde Act’s discipline as a detail while treating the Left’s objections as an inconvenience. Karat’s formulation, repeated across the party’s publications that month, was that the withdrawal was not over a reactor contract but over the character of the state that would sign it: a government that accepted certification by a foreign statute, the Left held, had altered India’s strategic posture without a mandate, and no reading of the Common Minimum Programme could be made to cover the alteration.

The government’s counter-statement was equally exact and is essential to the even-handed record. The Congress leadership held that the Left had been consulted at every stage through the coordination committee and the Mukherjee committee, that the Prime Minister had reported to Parliament repeatedly, that the deal’s terms had improved under Left pressure in ways the Left declined to claim, and that a party exercising power without office, through outside backing, had converted consultation into a veto over foreign policy that no coalition document had granted it. Sonia Gandhi, as chairperson of the alliance, and Pranab Mukherjee, as the government’s manager of the file, carried this case to the country in the weeks that followed. Between the two statements sits the institutional fact that the crisis reading exists to capture: the arrangement of 2004 had never specified what would happen when consultation failed, and so the failure, when it came, could be resolved only by the device the Constitution keeps for exactly this vacancy, a counted vote on the floor of the House on the question of confidence itself. The government announced that it would seek that vote at a special session, and Manmohan Singh moved the motion in his own name.

What numbers made a confidence motion unavoidable after 9 July 2008?

The withdrawal removed fifty-nine Left members from the government’s column in a House of 543, leaving the alliance and its remaining backers short of the majority mark of 272 on any strict party count, so the Constitution’s test of confidence on the floor became the only instrument that could establish whether the government still governed.

The thirteen days between the withdrawal and the division were a market conducted in public, and the record of them survives in unusual completeness because every participant had an interest in being quoted. The Samajwadi Party, with thirty-nine members and a history of hostility to the Congress that ran back decades, executed the pivot on which the result turned. Its leaders, Mulayam Singh Yadav and Amar Singh, stated their grounds in sequence: the party had been briefed on the agreement, had consulted the former President A. P. J. Abdul Kalam, whose scientific standing in the missile programme gave his reassurance a weight no ministerial assurance could carry, and had concluded that the deal served the country’s energy interest and that the alternative to backing the government was to strengthen the Bharatiya Janata Party, which the party existed, in its own account, to keep from office. The party’s stated reasons were placed on the record by its leaders in interviews and in the House, and they are the reasons the even-handed account must carry, alongside the opposition’s stated suspicion, argued by the Left and the Bharatiya Janata Party alike, that a party under investigation pressure in a disproportionate assets case and a party leadership facing the Bahujan Samaj Party’s rise in Uttar Pradesh had priced its votes in a currency other than argument. Both statements belong to the record; the investigations of the assets case ran on their own track, and no finding of any forum has joined the party’s July 2008 decision to an inducement, a distinction the later cash investigations make it necessary to state with care.

Against the Samajwadi pivot stood the opposing consolidation. The Bahujan Samaj Party of Mayawati, with seventeen members, announced that it would vote against the motion, its leader stating that the deal served American interests and that the government had forfeited the confidence of the House by its conduct over the preceding months, a statement that joined the party to the Left and the National Democratic Alliance in the division lobby for that week without joining it to either in any programme. The Left and the Bharatiya Janata Party, adversaries on nearly every question the House had considered since 2004, thus found themselves voting the same way for opposite stated reasons, the Left to punish an alignment it opposed and the opposition to fell a government it wished to replace, and the strangeness of the pairing was itself an argument each side used against the other: the government charged that the Left had made common cause with the communal forces its programme existed to resist, and the Left answered that a government saved by purchased arithmetic would discover the charge to be a confession. Underneath the exchanges, the whips of every party worked the small numbers that the arithmetic section below sets out, the undecided members, the ailing members summoned from hospital beds, the two members brought from custody to vote, and the cross-currents inside parties whose leaderships had announced one position and whose members, in at least fourteen cases on the opposition side, would vote or stay away in the other sense.

Beneath the two large pivots, the Samajwadi Party’s arrival and the Bahujan Samaj Party’s refusal, the thirteen days were filled by the small numbers whose names rarely survive in the general accounts and whose votes the division list preserves. The Jharkhand Mukti Morcha, with five members, declared for the government after its leader Shibu Soren’s negotiations with the Congress leadership, a backing the opposition charged had been purchased with the promise of office and the party defended as the settled inclination of a constituent of the alliance since 2004; Soren entered the cabinet as minister of coal in August 2008, a sequence both camps cited, in their opposite senses, for years. Single members and pairs, from the northeastern parties, from the smaller formations of the Hindi belt and from the ranks of the independents, were courted by both whips with an attention inversely proportional to their size, and the record of the division shows the harvest scattered in both lobbies. The government also carried into the vote the human arithmetic of a House in which illness and custody are not metaphors: members were brought from hospital beds in Delhi and from their constituencies on stretchers and in wheelchairs, and two members facing criminal charges were produced from custody to record their votes, spectacles the press recorded and the participants defended as the ordinary duty of a member on a confidence question. The crisis reading insists on these details because they are the crisis at ground level: a foreign agreement whose fate had come to depend on whether a member could be carried into the lobby is an agreement whose domestic passage had consumed every reserve the system kept for ordinary business, and the photographs of that fortnight, of ambulances at the Parliament House gate and tellers counting past midnight arithmetic, belong to the episode’s evidence as surely as the texts of the Hyde Act.

The Demand for a Sense of the House

Inside the thirteen days sat a procedural quarrel that the crisis reading treats as substantive, because it decided the form the settlement of the quarrel would take. The Left’s central public demand, after the withdrawal, was that the government place the safeguards text and the state of the agreement before Parliament and take the sense of the House before approaching the agency’s board, a demand framed as the minimum respect owed to a legislature whose majority the government had just lost. The government’s answer, given by the Prime Minister and by Mukherjee, was that the conduct of a negotiation was an executive function, that Parliament had been informed at every stage, that the texts of concluded agreements would be laid before the House, and that the Constitution supplied exactly one instrument for testing a government’s title to continue, the confidence motion, which the government had already announced it would move. The exchange was conducted through letters and statements across 9 to 11 July 2008, and on 10 July the cabinet’s committee on political affairs authorised the approach to the agency, whose secretariat circulated the draft safeguards text to the board the same week. The Left read the sequence as proof that the confidence vote was being staged after the act it was meant to judge, the international step taken first and the domestic ratification arranged to follow it, and Karat’s statements of those days charged the government with presenting Parliament with a completed fact. The government read the same sequence as the ordinary order of diplomacy, in which a board’s consideration of a safeguards text committed no one until the state concerned accepted it, and charged the Left with demanding a veto dressed as a discussion.

The quarrel mattered beyond its fortnight because it named the vacancy at the centre of the arrangement of 2004. No clause of the Common Minimum Programme, and no convention of the coordination committee, had settled whether a backer outside the cabinet could require a parliamentary proceeding as the price of an executive act, and the Constitution, which the government’s spokesmen cited, settles the question of confidence and is silent on the question of prior restraint. Into that silence both sides projected their theories of the coalition, and the special session of 21 and 22 July was convened to answer the constitutional question while the procedural one was still being argued in the newspapers. It is a measure of the crisis’s depth that the House, when it met, debated the agreement, the alignment, the statute, the assurances, the Speaker’s position and an allegation of purchase, and at the end answered, by 275 votes to 256, only the question the Constitution had asked. The other questions were left to the forums the following sections record, and the forums, between them, took five years to leave them where the House had left them.

Component Four: The Debate, the Cash and the Division of 22 July 2008

The special session sat on 21 and 22 July 2008, and the debate it held is the documentary centre of the crisis. The motion stood in the name of the Prime Minister, and the Leader of the Opposition, L. K. Advani, opened against it with the opposition’s consolidated case: the government had lost its majority the day the Left left, had negotiated a strategic surrender in instalments, and was surviving, in Advani’s stated charge, by methods that the House would see exposed before the vote. Sixty members spoke across the two days besides the two leaders. P. Chidambaram, the finance minister, carried the government’s substantive defence of the agreement, its energy arithmetic and its safeguards terms, in the debate’s most closely argued speech. Lalu Prasad Yadav, the railway minister, carried its political defence in Hindi, turning the opposition’s charge of surrender into a charge of obstruction. The speech that outlived the debate in public memory came from a first-generation participant in the new politics of Jammu and Kashmir: Omar Abdullah, speaking for the National Conference, declared that he was a Muslim and an Indian and proud to be both, rejected the suggestion that the deal injured Indian Muslims or served an American design against Muslim states, and announced that he would vote for the motion as an Indian first, a statement quoted across the press of the following week as the debate’s moral summary by the government’s supporters and dismissed by its opponents as theatre laid over arithmetic.

Who presided over the confidence debate, and what did his party do to him on 23 July 2008?

Somnath Chatterjee, the Speaker, a member of the Communist Party of India (Marxist) for four decades, presided and refused his party’s direction to resign and vote against the motion, holding the Speaker’s office to stand above party; on 23 July the party’s polit bureau expelled him for seriously compromising its position.

The presiding officer’s position concentrated the crisis into a single conscience. Chatterjee had been elected Speaker unanimously in 2004, had resigned his party membership of the House’s leadership in the convention of the office, and had maintained, as the withdrawal approached, that he would decide his course when the moment came. His party’s position, conveyed by its general secretary and by emissaries that included his old comrade Jyoti Basu’s circle, was that the Speakership was a party nomination and that honour required him to vacate the chair, return to the ranks and vote with the Left against the motion, a single vote that would also have carried the symbolism of the chair itself deserting the government. Chatterjee’s answer, given in his statement of 23 July 2008 after the expulsion, was that the Speaker does not and cannot represent any political party while acting as such, and he called the day of his expulsion one of the saddest of his life. The expulsion was executed by the polit bureau under the summary procedure of the party constitution, and it stands in the record as the crisis’s clearest measure of the stakes the participants believed were in play: a party expelled its most senior parliamentarian, a man of ten elections, rather than permit the chair to be seen to stand apart from the attempt to bring the government down, and the man accepted the expulsion rather than resign an office he held to be held in trust for the House.

On the afternoon of the second day, with the Prime Minister yet to reply, the debate produced the image by which the crisis is remembered. Three members of the Bharatiya Janata Party, Ashok Argal, Faggan Singh Kulaste and Mahavir Bhagora, came to the well of the House carrying bags from which they produced bundles of currency notes, stated by them to amount to one crore of rupees, and placed the money before the House with the allegation that it had been delivered to Argal’s residence that morning by Sanjeev Saxena, an aide to the Samajwadi Party leader Amar Singh, as an advance against a promised three crore of rupees to each of them if they abstained in the division. The allegation, stated in the chamber and repeated in the members’ written complaint to the Speaker, named Amar Singh as the author of the offer and named intermediaries through whom the approach had been arranged over the preceding days. The Speaker adjourned the House amid the uproar, the Prime Minister’s reply to the debate could not be delivered in the disorder that followed and was laid on the table of the House, and the division proceeded in an atmosphere in which the question before the members had visibly doubled: the House was voting on confidence in the government and, in the same act, on the credibility of an accusation of purchase made in its own well an hour before.

What happened to the television sting recording after 22 July 2008?

The recording of the alleged delivery, made by the CNN-IBN channel working with the three members, was withheld from broadcast while the channel answered the inquiry committee’s summons, was shown to the committee, and was broadcast on 12 August 2008, after which its contents entered the public argument as evidence each side read in its own sense.

The text that was laid on the table in the Prime Minister’s name survives in the proceedings and deserves a place in the record, because it is the government’s fullest single statement of its case and it was never heard. The reply restated the energy case for the agreement, the projection that nuclear power’s share of India’s electricity, under three percent in 2008, could not rise toward the government’s targets without imported fuel and reactors, and it restated the sovereignty case in the terms the assurances of 2006 had fixed: no bar on the strategic programme, no safeguards beyond the civilian list India itself would file, no obligation accepted that the signed texts did not contain. It answered the charge of alignment with the record of the government’s refusals, on Iraq’s reconstruction contracts, on the pipeline diplomacy with Iran that Delhi continued against Washington’s stated displeasure, and it closed with the appeal that the House judge the agreement by the country’s need and the government’s word, both of which, the text held, the preceding three years had vindicated. That the statement entered the record unspoken was more than a procedural casualty. The opposition’s spokesmen treated the unspoken reply as the crisis in miniature, a government that had negotiated its central undertaking in rooms its own legislature could not enter and defended it, at the last, in a document its own Prime Minister could not deliver over the noise of an accusation of purchase. The government’s managers treated the same fact as proof of the opposition’s design, to convert a debate it could not win on argument into a spectacle it could not lose on television. Both readings were carried into the division lobby by the members who heard them, and the tellers recorded their effect in numbers.

The division, when it came late on 22 July, returned the figures around which this article is built: 275 for the motion, 256 against, ten abstentions, a margin of nineteen. The analytical weight of those figures sits in their comparison with the strict party count, which scholars of the episode, James Chiriyankandath of the University of London among them in his study of the vote presented in November 2008, reconstructed as a government defeat: had every member voted the announced position of every party, the motion would have fallen by 268 votes to 271. The nineteen-vote margin was manufactured inside the division by two movements that the record documents in detail. Fourteen members of opposition parties crossed to vote for the motion, among them four of the Bharatiya Janata Party, two of the Marumalarachi Dravida Munnetra Kazhagam and single members of eight other parties, while eight further opposition members absented themselves or abstained against their parties’ instructions, four of them again from the Bharatiya Janata Party. Against that flow, seven members of the government’s own widened column, six of the Samajwadi Party and the Congress dissident Kuldeep Bishnoi, voted against the motion, and the two members who abstained by announced party decision, Mamata Banerjee of the Trinamool Congress and Vanlalzawma of the Mizo National Front, stood apart from both camps by design. The expulsions and suspensions that several parties visited on their defectors in the following weeks, the Bharatiya Janata Party’s above all, closed the division’s account in the only currency the House fully controls, membership of the parties themselves. The government had survived, and the terms of its survival, a margin assembled from defections in both directions under an unresolved allegation of purchase, became the inheritance that the next four years of investigation were asked to audit.

The Investigations: A Committee, a Chargesheet and a Court

The crisis did not end with the division, because the allegation made in the well of the House had an institutional life of its own, and the sequence of its investigations is the part of the record on which the three readings of the episode most sharply divide. The Speaker referred the three members’ complaint to a committee of the Lok Sabha constituted on 30 July 2008 under the chairmanship of V. Kishore Chandra Deo, a senior Congress member from Andhra Pradesh, with members drawn across the House’s divisions. The committee took evidence through the autumn, examined the sting recording in camera before its broadcast, heard the members who had displayed the money, the intermediaries named in the complaint and the channel’s representatives, and reported in December 2008. Its findings, stated as findings, were two. The committee found no evidence that money had been offered on behalf of Amar Singh or of Ahmed Patel, the Congress treasurer named in the surrounding allegations, and it held the video record of a car and a delivery at a residence to be insufficient to establish what had transpired or that any offer had been made for a vote or an abstention. And it recommended that the conduct of three men be investigated by the enforcement agencies: Sanjeev Saxena, who was recorded in the delivery; Sohail Hindustani, an activist named as an intermediary; and Sudheendra Kulkarni, an aide to L. K. Advani who had, by his own later published account, helped design the sting as an exposure operation. Members of the committee recorded notes of dissent that accused the majority of eagerness to acquit, and the report entered the public record as contested on its face, a finding of insufficiency that one camp read as exoneration and the other as choreography.

The dissent recorded inside the committee is worth its own paragraph, because it shows the forum dividing on method before it divided on verdict. The dissenting members charged that the committee had declined to summon the four persons they held to be indispensable to the inquiry: Amar Singh and Ahmed Patel, the men the complaint named, and Hashmat Ali, the driver of the car recorded at the delivery, and the unidentified man in the yellow shirt whom the recording showed assisting Saxena. They charged that the committee had been slow to obtain the call records of the telephones in the chain and slower to weigh them, and that its tracing of the money through bank withdrawals had been conducted on a method, the serial numbers of currency notes, that the record itself showed the banks did not maintain. The majority’s answer, carried in the report, was that a parliamentary committee proceeds on the evidence placed before it, that it had found no prima facie material on which to summon members of the other House against their recorded denials, and that the residue of suspicion was precisely what its recommendation of a police investigation existed to pursue. The disagreement was never reconciled, and it fixed the report’s reception: a document whose majority found insufficiency and whose dissenters found avoidance could be cited by every later forum, the police in their chargesheet and the court in its discharge, in support of opposite courses, and was.

The criminal track opened in January 2009, when a case was registered on the committee’s recommendation and the Delhi Police crime branch began the investigation that the parliamentary process could not conduct. It ran for over two years, through the general election of 2009 that returned the government with a larger majority and no dependence on the Left, and it surfaced again in public in March 2011, when The Hindu newspaper published a confidential cable of the American embassy in New Delhi, released in the WikiLeaks disclosures, reporting that an aide to the Congress leader Satish Sharma had shown an embassy officer two chests of currency said to be part of a fund assembled for the purchase of votes in July 2008. The cable was an embassy report of a claim made to its officer, unverified by the embassy itself, and the government treated it in those terms: Manmohan Singh, in statements to both Houses on 18 March 2011, rejected the allegation absolutely and firmly, held that no person from the Congress or the government had indulged in an unlawful act, and pointed to the committee’s finding of insufficient evidence. The opposition treated the cable as corroboration of the display in the chamber, and the parliamentary session of March 2011 was consumed by the quarrel, which ended, as the original debate had ended, without a resolution that both sides accepted. For journalists and editors, the season was a case study in the handling of unverifiable source material of the highest sensitivity, and the account of how journalists use VaultBook to protect their sources belongs to the record of professional practice that the episode’s reporting generated.

The police file reached a court in August 2011. The chargesheet filed on 24 August 2011 before the special judge charged Amar Singh, Sudheendra Kulkarni, Sanjeev Saxena, Sohail Hindustani and the two former members Faggan Singh Kulaste and Mahavir Bhagora with conspiracy and with offences under the Prevention of Corruption Act, alleging that one crore of rupees had been delivered at Argal’s residence on the morning of 22 July 2008 as an advance against nine crore promised for three abstentions; a supplementary chargesheet later named Argal himself, after the Speaker’s sanction to prosecute a sitting member was sought and given. Amar Singh was arrested on 6 September 2011 when he appeared before the court, and the spectacle of a principal of the July 2008 pivot in custody completed, for the government’s critics, a narrative arc from the well of the House to the lock-up. The narrative did not survive the court’s own examination. On 22 November 2013, the special court discharged Amar Singh and the three Bharatiya Janata Party members, accepting their account that their conduct had been that of whistleblowers exposing an attempt at purchase, and held the circumstantial chain against Saxena’s employers to be incomplete, while directing that Saxena himself face trial under the provision of the corruption statute addressed to a person who takes gratification in order to influence a public servant. No conviction of any principal emerged from the episode. The investigations thus closed in the posture in which the committee had left them five years earlier: an allegation recorded in the chamber and on tape, an offer and a delivery attested by their recipients and contested in their authorship, and a chain of forums, parliamentary, police and judicial, each of which found the evidence sufficient for a further step and insufficient for a verdict.

The Sting: Cameras, Intermediaries and the Whistleblower Question

The cash display cannot be understood as an improvisation, because the record assembled by the committee and the police shows a constructed operation behind it, and the construction is where the episode’s hardest questions sit. The three members stated, in their complaint and in their evidence, that they had been approached across the days before the vote by Sohail Hindustani, who presented himself as a conduit to persons able to pay for abstentions, and that they had resolved to expose the approach rather than accept or merely refuse it. Through Sudheendra Kulkarni, an aide to the Leader of the Opposition who had worked on the party’s campaign machinery, the members were connected with the CNN-IBN channel, whose journalists agreed to record the approaches and the delivery with concealed cameras, on the understanding, stated by the channel’s editors in their evidence, that broadcast would follow verification and legal advice. Sanjeev Saxena, identified as an aide in Amar Singh’s office, was recorded delivering the money at Argal’s residence at 4 Ferozeshah Road on the morning of 22 July, accompanied by a man in a yellow shirt whose identity the investigations never fixed to the committee’s satisfaction, and the three members carried the delivered bundles, one crore of rupees by their count, into the chamber that afternoon. Kulkarni, in his published account of the operation after the chargesheet, described the sting as a citizen’s exposure of a market in votes, designed with Advani’s knowledge circle and frustrated, in his telling, by the channel’s refusal to broadcast before the vote, which he held to have denied the House the evidence at the only hour it could have altered the result. The channel’s editors answered that the recording had been incomplete at the hour of the vote, that the identity of the man who brought the money had not been verified, and that their legal advice forbade a broadcast that could itself have been an intervention in the division. Each of these accounts was given under examination, and none was withdrawn, and the committee’s report, which recommended Kulkarni himself for investigation as a participant in the chain he claimed to have exposed, recorded the forum’s refusal to accept the whistleblower frame at the value its authors claimed for it.

The Delhi court’s order of November 2013 reversed that refusal, and the reversal is the sting’s final institutional fact. In discharging the three members and Amar Singh, the court accepted that the members’ conduct bore the character of a whistleblowing operation, an exposure conducted with a television channel and concluded with a disclosure on the floor of the House at the first opportunity the sitting allowed. The reasoning sits awkwardly beside the committee’s, and the awkwardness is instructive for the crisis reading. Two forums of the same state, examining overlapping evidence across five years, divided precisely on the question of what the operation was: a purchase interrupted, a purchase invented, or an exposure of an approach whose authors stood behind the aide who carried the cash and could not be reached by the chain the police assembled. The court held the chain incomplete against Saxena’s alleged principals and left Saxena, the one man recorded with the money in his hands, to face trial on the narrower charge. No forum at any stage held that no approach had occurred, and no forum held that an approach by the government or its parties had been proved, and between those two negatives the episode rests, a scandal with a verified centre, the cash on the table of the House, and unverified edges in every direction. For the argument of this article, the resting place is the finding. A confidence division conducted inside a sting operation, investigated by a committee of the House, a police force and a court without a settled verdict, is a domestic crisis by any definition the republic has used, and the readings that pass over the sting as colour are discarding the evidence that explains why July 2008, and not the signing ceremonies of October, is the date the episode is remembered by.

The Settlement Completed: Vienna, Washington and the Peace That Followed

The government that survived on 22 July moved through the international sequence at the speed the Left’s withdrawal had been intended to prevent. The board of governors of the agency approved India’s safeguards agreement on 1 August 2008, ten days after the division. The Nuclear Suppliers Group, meeting in Vienna across two sessions in August and September, granted India its waiver on 6 September 2008, permitting the group’s members to trade in civil nuclear material and equipment with a state outside the non-proliferation treaty for the first time, after a resistance conducted by Austria, Ireland and New Zealand, with China absenting itself from the final consensus, had been overcome by American and Indian diplomacy in the capitals concerned. The Congress of the United States approved the cooperation agreement in the first days of October, and on 10 October 2008, in Washington, the minister of external affairs, Pranab Mukherjee, and the Secretary of State, Condoleezza Rice, signed the agreement that brought the 123 text into force. The four-component trajectory was complete: the statement of 2005, the statute of 2006, the withdrawal of 9 July 2008 and the vote of 22 July 2008 had issued, inside fifteen weeks of the vote, in the settlement the trajectory had promised, and the sequence stands as the cleanest demonstration in the pillar that the domestic crisis and the foreign settlement were one event read from two capitals.

The signing itself supplied a small coda that the participants noticed. Condoleezza Rice had visited New Delhi on 4 October 2008 with the ceremony expected at the visit’s end, and the signature had to wait, because the approval legislation in Washington had passed the two Houses and still awaited the President’s signature, which it received on 8 October. Mukherjee therefore travelled to Washington, and the agreement that an Indian government had nearly fallen to conclude was signed in the other capital, in the State Department’s rooms, with the two principals recording in their statements the three years of negotiation and the parliamentary ordeal, on the Indian side, that the signatures closed. The geography of the ceremony was a fitting emblem of the trajectory: the text had been authored in both capitals, but it had been ratified, in every sense that cost anything, in one.

The American end of the sequence deserves its figures, because they complete the contrast on which the crisis reading rests. The approval legislation for the agreement passed the House of Representatives on 27 September 2008 by 298 votes to 117 and the Senate on 1 October 2008 by 86 votes to 13, margins assembled without a confidence question, without a coalition in peril and without a single member displaying money on either floor. The same settlement that had cost the Indian government its majority arrangement, its Speaker’s party membership and five years of investigation was carried in Washington as a foreign policy measure of the ordinary kind, debated on its merits and its non-proliferation price and concluded inside a fortnight of floor time. The asymmetry is not an argument against the settlement; it is the settlement’s most instructive property. A bargain between a presidential system with a disciplined majority on the question and a parliamentary coalition held together by outside backing will be priced in the second capital in a currency the first does not use, and the price, once paid, becomes part of the bargain’s meaning for everyone who must live under it. India’s negotiators of the following decade, on files from climate finance to defence procurement, worked in the knowledge their legislatures had acquired in 2008, that an international commitment is a domestic event with a division list attached.

The peace that followed inside India was real and is part of the record the crisis reading must carry honestly. The general election of April and May 2009 returned the Congress with 206 seats, its strongest result since 1991, and installed a second United Progressive Alliance government that owed nothing to the Left, whose combined strength fell to twenty-four seats, sixteen of them held by the Communist Party of India (Marxist) in a collapse its own documents attributed in part to the rupture of 2008. The deal’s commercial promise, by contrast, matured slowly: the civil liability statute of 2010, which kept a right of recourse against suppliers that the supplier states and their companies judged uninsurable, held up the reactor contracts the settlement had been sold to deliver, and the first American-supplied reactors remained unbuilt across the following decade, a lag the strategic reading must explain and the crisis reading predicts, since a settlement carried through a domestic crisis at the price of a coalition tends to arrive with its legislative aftermath priced for the politics that produced it. The longer institutional shadow fell elsewhere. The trust vote of 2008 normalised a pattern the coalition era had been building since 1989, the pattern the series traces from the post-1991 institutional-political context: governments whose majority is assembled rather than inherited treat every major bill as a confidence question in waiting, and the whips’ arithmetic of July 2008, with its cross-votes, its abstentions and its expulsions, became a rehearsal studied by every floor manager who followed. What the episode did to the three-tensions framework that organises this series, the renegotiation of founding settlements under a dominant executive, is set out in the account of the integrated three-tensions framework, and its constitutional scale is best seen against the benchmark of the most-consequential-constitutional-moment framework, beside which July 2008 stands as the coalition era’s nearest approach to a constitutional moment conducted entirely within the ordinary forms of the House.

The Crisis in the Coalition Form: What July 2008 Taught the System

The settlement’s completion closed the diplomatic file and left the domestic file open, and the entries made in it after October 2008 show the crisis working itself into the system’s habits. The first entry was partisan and immediate. The parties that had sold or withheld their members in July spent the autumn disciplining them: the Bharatiya Janata Party expelled members who had crossed or stayed away, the Samajwadi Party expelled its own dissenters who had voted against the motion its leadership had backed, and the Bahujan Samaj Party read its participation in the losing lobby as a credential for the national campaign it intended to run in 2009. The second entry was electoral. The verdict of May 2009, which returned the Congress with 206 seats, was read by the government’s managers as the country’s ratification of the stand taken in July, and the reading was available because the Congress campaigned, in part, on the deal and on the Prime Minister’s refusal to surrender it, the first occasion on which a foreign agreement had been carried to the electorate as a governing credential by the party that had staked a government on it. The Left’s collapse to twenty-four seats was read by its own central committee’s review as the price of a rupture the electorate had not followed, a self-assessment that closed the argument of 2008 in the only forum whose verdict neither side could appeal.

The Left’s own review of the 2009 verdict, adopted by the central committee of the Communist Party of India (Marxist) in June 2009, supplies the crisis with its most candid internal document. The review held the withdrawal itself to have been correct on the party’s stated grounds, and located the defeat in the sequel: the attempt to build a non-Congress, non-Bharatiya Janata Party alternative in the campaign had failed to convince the electorate, the vote had consolidated behind the two national formations in state after state, and the party’s strength in its own strongholds had been spent defending positions the rupture had made untenable. Read against the government’s reading of the same verdict, the review completes the adjudication this article owes the record. Both camps agreed, in their internal documents and in public, that July 2008 had been decided correctly by their own lights and had cost the side that forced the question its place in the arrangement of power. The disagreement that remains is over the price’s meaning, whether it purchased a necessary settlement or advertised a preventable surrender, and that disagreement, anchored in the party documents of 2008 and 2009, is the form in which the crisis survives as an argument rather than a memory.

The third entry was procedural, and it is the one the crisis reading exists to record. After July 2008, every floor manager in the Lok Sabha worked with a template the episode had cut: a government short of a majority on a question it could not postpone would announce its own confidence motion rather than wait for the opposition’s, would build its margin from named pivots whose price and stated reasons were placed on the record in advance, and would treat abstentions, the currency in which July 2008 was finally settled, as a purchasable and defensible quantity distinct from votes. The template had ancestors in the confidence votes of the nineteen nineties, and the account of that decade’s practice belongs to the pillar’s earlier articles, but July 2008 was its consolidation, because it was the first confidence vote in which the government’s survival turned on a foreign policy instrument and in which the margin’s manufacture was conducted, alleged and investigated in full public view. The liability statute of 2010 supplied the template’s footnote from the other side: a Parliament that had been asked to ratify a settlement by surviving it asserted itself, in the next session of the file, by writing supplier liability into the settlement’s commercial chapter against the stated preference of both signatory governments, as though the House were reclaiming, clause by clause, the ground it had ceded in the division lobby. Read together, the vote and the statute are the crisis’s complete domestic sentence, the executive carrying the country into a settlement by the arithmetic of survival, and the legislature taxing the settlement afterwards for the manner of its carrying.

Adjudicating the Three Readings

The named disagreement the brief assigns this article can be stated as a contest of centres. The strategic reading places the centre in the relationship: two democracies, estranged through the Cold War, concluded a settlement that admitted India to the nuclear order on terms no other state outside the treaty has obtained, and the domestic turbulence was the noise a large change makes in a noisy polity. The policy-trajectory reading places the centre in the file: a negotiation of exceptional technical density, conducted across four instruments and six years, whose every clause can be tracked from the statement to the waiver, and whose domestic politics was the ratification procedure of a parliamentary state working as designed. The crisis reading defended here places the centre in the House on 22 July 2008, and it defends that placement on evidence the other two readings cannot digest. The strategic reading cannot explain why the settlement required the destruction of the governing arrangement, the expulsion of a Speaker and a margin assembled from defections, since none of those was a cost of the relationship itself; they were costs of the domestic distribution of the decision, and they are invisible from the altitude of statecraft. The trajectory reading cannot explain why the file’s hardest problems were solved not at the negotiating table but in the division lobby, nor why the file’s own custodians, from the Prime Minister’s August 2007 interview to the Left’s June 2008 authorisation, behaved as though the negotiation were a coalition instrument first and a diplomatic one second.

The scholarly placements of the episode divide along the same lines, and naming them sharpens the adjudication. Talbott’s Engaging India, written before the deal and extended in its later commentary, supplies the strategic reading’s prehistory: the dialogue of 1998 to 2000 as the patient construction of a relationship that the deal merely crowned. Ganguly and Mukherji, in India Since 1980, place the agreement inside the longer reorientation of Indian statecraft after the Cold War, one entry in a ledger that includes the economic opening of 1991 and the defence framework of 2005, and their treatment gives the domestic contest a paragraph where this article gives it a thesis. Pant’s essays, written across the negotiation and its aftermath, follow the file with a closeness neither of the others attempts, and their centre of gravity sits in the trajectory: the texts, the holdouts in Vienna, the liability quarrel that followed. Against all three, the crisis reading asserts a claim about explanatory order. The relationship was real, the reorientation was real and the file was masterfully conducted, and none of the three explains the event that every participant treated as the episode’s climax and that the country’s memory retained: the night a government counted its members in public, under an allegation of purchase, to decide whether a foreign agreement would be allowed to exist. A reading that cannot explain its subject’s climax has misplaced its centre, whatever else it explains, and the correction this article offers is a relocation of the centre to the chamber where the episode was decided.

The complication owed to the crisis reading must be faced in its strongest form, because the reading can be accused of building its climax on the cash display, and the cash display is the least settled element of the entire record. If the allegation was theatre, if the money’s authorship was never established and its purpose never proved, then a reading that culminates in the trust vote may seem to inherit the theatre’s exaggeration. The answer lies in what the reading actually claims. The crisis is not the scandal; the crisis is the destruction of a governing arrangement by a foreign policy decision, the conversion of a negotiation into a confidence question, and the conduct of that question under an allegation of purchase that the state’s own forums then spent five years failing to resolve. Every element of that claim is established by documents whose authority no camp disputes: the statement and the statute, the letters of withdrawal delivered on 9 July 2008, the proceedings of 21 and 22 July, the division list, the expulsion order of 23 July, the committee report of December 2008, the chargesheet of August 2011 and the discharge order of November 2013. The investigations are not the reading’s weakness but its final exhibit, because a polity in which an allegation of vote purchase, made on the floor during a confidence division, can be processed for five years and closed without a verdict is a polity whose crisis the strategic and trajectory readings have no vocabulary to describe. The rupture between the alliance and the Left, the subject of the companion account centred on the rupture with the Left, is one component of that crisis, the third of four; this article has kept the whole sequence in frame, from the statement that made a coalition’s promise untenable to the court order that closed its last file, and it is the whole sequence, and no single day inside it, that the four-component map exists to make citable.

The four components, gathered as the article’s findable artifact, are set out in the Domestic Crisis Trajectory Table below, which names each component, its date and the domestic institution through which its consequences ran.

Component	Date	Instrument or event	Domestic institution engaged
One	18 July 2005	Bush Singh Joint Statement	Parliament and the coalition coordination committee
Two	18 December 2006	Hyde Act, Public Law 109-401	The Lok Sabha debate and the Mukherjee committee
Three	9 July 2008	Left withdrawal of outside backing	The presidency and the confidence procedure
Four	22 July 2008	Trust vote, 275 to 256, with the cash display	The Lok Sabha, its inquiry committee and the courts

The table is the argument in miniature. A foreign policy that begins in a joint statement and ends in a courtroom has travelled through every institution the republic keeps for its domestic quarrels, and the deal that India signed in Washington in October 2008 was, by the time of its signing, also a verdict on the coalition system that had governed India for two decades: that it could carry a decision of this size only by breaking, and that the breaking, recorded in division lists and expulsion orders and committee reports, is the most durable text the episode produced. The settlement endures in the statute books of two countries. The crisis endures in the memory of every party that has since counted its members before announcing a conviction, which is to say, in the working constitution of Indian politics itself.

A last word belongs to the pillar’s longer arc, because the crisis of 2008 is the point toward which its earlier episodes run and from which its later ones descend. The wars and treaties of the earlier articles in this sequence were foreign decisions with domestic consequences, absorbed by governments whose majorities were their own and whose parliaments ratified after the fact. The nuclear deal inverted the order: the domestic consequence, the destruction of the governing arrangement, preceded and conditioned the foreign completion, and the ratification was the event. In that inversion the episode marks the maturity of the condition the framework names, a republic in which foreign policy is no longer a reserve domain guarded by a permanent executive consensus but a contested field on which governments are made and unmade like any other. The men and women who fought the thirteen days understood the novelty before the scholars named it, and they fought accordingly, with letters to a President, an expulsion under a party constitution’s summary clause, a sting conducted by a television channel and a division in which the tellers’ list ran to defections in both directions. The settlement they produced, a statute in Washington, a waiver in Vienna and a signature in October, has outlasted every government that touched it. The crisis they produced has outlasted the settlement’s news, and it is the crisis, this article has argued, that the record of Indian politics will keep.

Frequently Asked Questions

Q: What was the 22 July 2008 trust vote?

The trust vote was a motion of confidence in the Council of Ministers that Prime Minister Manmohan Singh moved in his own name at a special session of the fourteenth Lok Sabha, after the Left parties withdrew their outside backing over the nuclear agreement. The House debated the motion on 21 and 22 July 2008 and divided late on the second day. The motion carried by 275 votes to 256, with ten members abstaining, a margin of nineteen that kept the United Progressive Alliance government in office and cleared the way for the safeguards agreement, the suppliers group waiver and the signing of the cooperation agreement in October 2008.

Q: What was the cash-for-votes scandal 2008?

The scandal was the allegation, made on the floor of the Lok Sabha during the confidence debate on 22 July 2008, that money had been offered to opposition members to abstain in the division. Three Bharatiya Janata Party members, Ashok Argal, Faggan Singh Kulaste and Mahavir Bhagora, displayed bundles of currency stated to amount to one crore of rupees, which they said had been delivered to Argal’s residence that morning as an advance against three crore each, and named the Samajwadi Party leader Amar Singh as the offer’s author. The allegation was investigated by a Lok Sabha committee, which found the evidence insufficient, and later by the Delhi Police, whose case ended in the discharge of the principals in November 2013.

Q: What is the Hyde Act 2006?

The Hyde Act is the Henry J. Hyde United States-India Peaceful Atomic Energy Cooperation Act of 2006, Public Law 109-401, signed by President George W. Bush on 18 December 2006. The Act authorised the President to waive the provisions of American atomic energy law that barred civil nuclear cooperation with India, subject to determinations, reports to the Congress and a statement of policy that covered India’s conduct on Iran, non-proliferation and export controls. In the Indian debate the Act became the contested text of the whole agreement, since the Left parties held its reporting and policy provisions to be a standing American discipline over Indian foreign policy, a reading the government rejected.

Q: What is the 18 July 2005 Bush Singh statement?

The statement is the joint declaration issued in Washington on 18 July 2005 by Prime Minister Manmohan Singh and President George W. Bush, which founded the civil nuclear bargain. The United States undertook to work toward full civil nuclear cooperation with India, to seek changes in American law and in the guidelines of the Nuclear Suppliers Group, and India undertook to separate its civilian and military nuclear facilities in phases, accept agency safeguards on the civilian side, maintain its moratorium on testing, work toward a fissile material treaty and align its export controls with the regimes it had not signed. Every later instrument in the trajectory, the Hyde Act, the 123 Agreement, the safeguards agreement and the waiver, executed or contested the undertakings exchanged in that statement.

Q: What did the 123 Agreement specify operationally?

The 123 Agreement, whose text was concluded on 27 July 2007 and released on 3 August 2007, specified the working terms of cooperation between India and the United States under the section of American law that governs such agreements. It granted India consent to reprocess American-origin spent fuel at a dedicated safeguarded facility, recorded American backing for an Indian strategic fuel reserve and for agency-arranged fuel supply in case of disruption, addressed the return of supplied material and the consequences of a future test in language both governments could defend at home, and set the cooperation inside the safeguards framework India had accepted. It entered into force after the suppliers group waiver, with the formal signing in Washington on 10 October 2008.

Q: What was Manmohan Singh’s stake in the trust vote?

Manmohan Singh staked his continuance in office and the agreement together, and treated them as a single question. He had told an interviewer in August 2007 that the Left was free to withdraw its backing if it found the deal objectionable, had authorised the approach to the agency in Vienna that triggered the withdrawal, and moved the confidence motion in his own name rather than waiting for the opposition to move against him. His stake was also a matter of record and reputation: the deal was the central external undertaking of his government, he had given assurances to Parliament on its terms in 2006, and a defeat would have ended his government, voided the undertakings given in his name in Washington and fixed the episode, in every later account, as his personal failure.

Q: When did the Left withdraw support from UPA-1?

The four Left parties announced the withdrawal of their outside backing on 8 July 2008 and delivered their letters to President Pratibha Patil on 9 July 2008, the date the withdrawal took formal effect. The decision had been authorised by the central committee of the Communist Party of India (Marxist) in June 2008, to be executed when the government approached the International Atomic Energy Agency to operationalise the safeguards agreement. The trigger was the Prime Minister’s statement to journalists travelling with him to Japan on 7 July 2008, that the government would approach the agency very soon. The withdrawal removed fifty-nine members from the government’s column and made the confidence vote of 22 July unavoidable.

Q: What did the Lok Sabha inquiry committee report in December 2008?

The committee reported that it had found no evidence that money had been offered on behalf of Amar Singh or Ahmed Patel, the two men at the centre of the allegations, and that the video record of a car and a delivery at a residence could not establish what had transpired inside or that any offer had been made for a vote or an abstention. It recommended that the enforcement agencies investigate three men, Sanjeev Saxena, Sohail Hindustani and Sudheendra Kulkarni, whose conduct the evidence did engage. Members of the committee appended notes of dissent accusing the majority of eagerness to acquit, so the report entered the record as a finding of insufficiency that remained contested on its face.

Q: Who chaired the committee that inquired into the cash display?

The committee was chaired by V. Kishore Chandra Deo, a senior Congress member of the Lok Sabha from Andhra Pradesh, and it is known in the literature of the episode by his name. The Speaker constituted the committee on 30 July 2008, eight days after the division, with members drawn across the House’s divisions, on the complaint of the three members who had displayed the currency. Deo’s committee took evidence through the autumn of 2008, examined the sting recording before its broadcast, heard the complainants, the intermediaries and the channel, and reported in December 2008. Deo defended the report in public for years afterwards, restating that no prima facie evidence connected the men his committee had cleared to the money.

Q: What did the Delhi Police chargesheet of August 2011 allege?

The chargesheet, filed on 24 August 2011 before the special judge in Delhi, alleged a conspiracy under the Prevention of Corruption Act and the Penal Code. It named Amar Singh, his aide Sanjeev Saxena, the intermediary Sohail Hindustani, Sudheendra Kulkarni, and the former members Faggan Singh Kulaste and Mahavir Bhagora, and alleged that one crore of rupees had been delivered at Ashok Argal’s residence on the morning of 22 July 2008 as an advance against nine crore of rupees promised for three abstentions in the division. A supplementary chargesheet later named Argal himself, after sanction to prosecute a sitting member had been sought. Amar Singh was arrested on 6 September 2011 when he appeared before the court, and the accused contested the charges as a circumstantial chain with its central links missing.

Q: What did the Delhi court decide in the cash-for-votes case in November 2013?

On 22 November 2013 the special court discharged Amar Singh and the three Bharatiya Janata Party members, Ashok Argal, Faggan Singh Kulaste and Mahavir Bhagora, accepting the account that their conduct had been that of whistleblowers who had exposed an attempt at purchase rather than conspirators in one. The court held that the circumstantial evidence did not complete a chain connecting Saxena’s delivery of money to any direction by Amar Singh, and that Sudheendra Kulkarni’s part had been confined to connecting the television channel with the members to record the approach. It directed that Saxena alone face trial under the provision of the corruption statute addressed to a person who takes gratification in order to influence a public servant, and no conviction of any principal followed.

Q: What did Omar Abdullah say in the trust vote debate?

Omar Abdullah, speaking for the Jammu and Kashmir National Conference on 22 July 2008, rejected the argument that the nuclear agreement injured Indian Muslims or formed part of an American design against Muslim countries. He declared that he was a Muslim and an Indian, and proud to be both, said that no external power would decide his position on an Indian question, and announced that he would vote for the confidence motion. The speech was short, delivered without notes in the debate’s closing hours, and was quoted across the press of the following week more widely than any other intervention in the two days, with the government’s supporters treating it as the debate’s moral summary and its opponents treating it as eloquence placed over arithmetic.

Q: Why was Manmohan Singh’s reply to the debate laid on the table of the House?

The Prime Minister’s reply was overtaken by the disorder that followed the display of currency in the well of the House on the afternoon of 22 July 2008. The Speaker adjourned the sitting amid the uproar, members of the opposition pressed the allegation and the demand for the government’s answer in the chamber, and when the House resumed, the conditions for a delivered reply, heard in silence by convention, did not obtain. Manmohan Singh therefore laid the text of his reply on the table, where it entered the record of the proceedings without being spoken, and the House proceeded to the division. The episode left the government’s formal answer to two days of charges unread aloud, a gap the opposition cited as evidence of a defence the Prime Minister could not make in person.

Q: How did India’s IAEA votes on Iran enter the parliamentary argument over the Hyde Act?

India voted at the agency’s board in Vienna on 24 September 2005 to find Iran in non-compliance with its safeguards obligations, and on 4 February 2006 to report the Iranian file to the Security Council. The Left parties argued that the votes were the first instalment of the alignment’s price, cast under American pressure in the season the nuclear bargain was being negotiated, and the Hyde Act’s statement of policy, which named India’s cooperation on Iran as an object of American policy, was read in the Lok Sabha as the codification of that pressure. The government answered that the votes rested on the Iranian record and on India’s own non-proliferation interest, and that they preceded the Act by a year and more. The argument was never resolved on its facts, and it set the pattern for every later dispute over the statute’s meaning.

Q: Which states held out against the NSG waiver at Vienna in September 2008?

The sustained resistance inside the Nuclear Suppliers Group came from Austria, Ireland and New Zealand, three small member states whose governments argued that a waiver for a state outside the non-proliferation treaty, without a test ban commitment or a fissile production halt, would damage the regime the group existed to guard. China absented itself from the final consensus rather than record an objection. The negotiations ran across two sessions, in August and on 4 to 6 September 2008, with American and Indian officials working the capitals of the holdout states between the sessions, and the waiver was adopted by consensus on 6 September 2008 after the holdouts accepted a record of Indian commitments on testing and non-proliferation read into the group’s proceedings. The episode is the standing example of small states extracting a price in a consensus body.

Q: When was the 123 Agreement signed, and who signed it for India and the United States?

The agreement was signed in Washington on 10 October 2008, by Pranab Mukherjee, the minister of external affairs, for India, and by Condoleezza Rice, the Secretary of State, for the United States. The signing came after the agency’s board had approved India’s safeguards agreement on 1 August 2008, after the Nuclear Suppliers Group had adopted its waiver on 6 September 2008, and after the Congress of the United States had passed the approval legislation for the agreement in the first days of October. The text signed that day was the text concluded by the negotiators on 27 July 2007, so the ceremony closed a trajectory of more than three years that had run from the joint statement of July 2005 through the Hyde Act, the withdrawal of the Left and the trust vote.

Q: What did the 2011 cable disclosure published by The Hindu add to the cash-for-votes record?

The disclosure added an American embassy report of a claim. The cable, sent from the embassy in New Delhi in July 2008 and published by The Hindu in March 2011 from the WikiLeaks release, recorded that an aide to the Congress leader Satish Sharma had shown an embassy officer two chests of currency, said to form part of a fund of fifty to sixty crore of rupees assembled for the purchase of members’ votes in the confidence division. The embassy had not verified the claim, and the officer’s cable reported it as a statement made to him. Manmohan Singh, answering in both Houses on 18 March 2011, rejected the allegation absolutely and firmly and pointed to the committee’s finding of insufficient evidence, while the opposition treated the cable as corroboration of the display in the chamber, and the quarrel consumed the session without a resolution either side accepted.

Q: What reasons did the Samajwadi Party give for backing the government in July 2008?

The party’s leaders, Mulayam Singh Yadav and Amar Singh, gave three reasons in public and in the House. The party had examined the agreement and had consulted the former President A. P. J. Abdul Kalam, whose reassurance on the agreement’s safety and on India’s strategic freedom the party treated as settling the technical question. The deal served the country’s energy interest, which the party held to outweigh its quarrels with the Congress. And the alternative to backing the government was a division that could bring the Bharatiya Janata Party toward office, which the party’s stated purpose in national politics was to prevent. The opposition parties recorded their disbelief of the stated reasons and alleged inducements of another kind, and no finding of any forum has joined the party’s decision to an inducement, so the stated reasons stand in the record beside the suspicion, unresolved.

Q: Why did the Bahujan Samaj Party vote against the confidence motion?

Mayawati, the party’s leader and the chief minister of Uttar Pradesh, announced that her seventeen members would vote against the motion, and stated the party’s grounds in the same declaration. The agreement, in the party’s stated reading, served American interests and had been negotiated without taking Parliament or the country into confidence, and the government had forfeited the House’s trust by its conduct across the preceding months, including its handling of the state government in Uttar Pradesh. The decision placed the party in the same division lobby as the Left and the National Democratic Alliance for that week, without any programme shared with either, and it formed part of the party’s wider positioning for the general election of 2009, in which it presented itself as an independent claimant to national office rather than an auxiliary of either camp.

Q: What did President Bush’s signing statement say about the Hyde Act’s conditions?

In the statement issued with his signature on 18 December 2006, President Bush recorded that his administration would construe the Act’s provisions that touched the conduct of foreign affairs, including its passages on Iran and its policy declarations, as advisory, consistent with the President’s constitutional authority over the foreign relations of the United States. The gloss mattered in New Delhi out of proportion to its legal weight in Washington. Indian officials cited it for years as evidence that the Act’s most contested passages did not bind American conduct, still less India’s, while the Act’s Indian critics answered that a courtesy extended by one president at a signing ceremony could be withdrawn by any successor, and that a discipline which depended on such a gloss was a discipline still. The exchange fixed the pattern of the statute’s Indian career, in which every clause was argued twice, once as law and once as signal.