The Price of an Institute

Ninety thousand francs

The most secure beginning is a table. In the audited statement of IHÉS income for the year ending December 31, 1969, a line reads “Ministère des Armées.” The budgeted figure is 100,000 francs. The amount recorded as received is 90,000. Other lines list French and European members, the Ministry of National Education, the American committee for IHÉS, and the General Delegation for Scientific Research. The document was accompanied by a Price Waterhouse report dated February 27, 1970.11Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4, “État des produits”.

Numbers do not explain themselves, but they discipline explanation. The 90,000 francs represented about 4.17 percent of the 2,160,218.24 francs actually received as contributions and subsidies. Measured against total recorded income of 2,523,216.99 francs, including publication sales, housing charges, cafeteria receipts, and other resources, it represented about 3.57 percent. Jackson’s later summary, following the historian David Aubin, says military sources never supplied more than about five percent of the budget in the relevant years. The audited 1969 figures are consistent with that summary.22Jackson, “As If Summoned from the Void, II,” 1197–98.

The choice of denominator is not a technical nuisance. “Four percent” describes the share among institutional contributions and subsidies. “Three and a half percent” describes the share among all income. “Five percent” is a later approximate upper bound across years. A book that gives one number without a denominator manufactures precision. A book that rounds the line to insignificance makes a moral judgment under cover of arithmetic.

The audit and the April minutes refer to different fiscal moments. The audited financial statements record 90,000 francs actually received during 1969. The minutes of April 22, 1970, record the board’s decision about a proposed 90,000-franc grant for 1970. The matching amounts do not prove that one document describes the same payment as the other. The chapter therefore keeps receipt, proposal, and vote separate.33Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1.

This distinction also limits causal language. The 1969 line establishes prior military-origin income; the April vote establishes the policy decision immediately preceding resignation.

Ninety thousand francs was a small part of the institute’s income. It was also almost half the gap identified in a pre-budget discussed by the board on December 10, 1969. The director presented irreducible projected expenditure of 2.7 million francs and then-certain revenue of 2.5 million, a difference of 200,000. The Ministry of National Education had promised to reconsider its subsidy so that the program might be balanced. In that immediate setting, 90,000 francs was neither the foundation of IHÉS nor loose change.44Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of December 10, 1969, A3.1.4.1.1/1, item 2.

This is the scale on which the conflict should be read. Grothendieck did not discover that the institute was secretly a military laboratory. The surviving documents checked for this chapter do not identify a weapons project performed by him, by the algebraic-geometry seminar, or by IHÉS. The conflict concerned money originating with the Ministry of the Armies and the meaning of accepting it. The distinction between origin and documented research purpose is essential. It did not make the origin morally irrelevant to Grothendieck.

The independence that required support

IHÉS was created in 1958 as an independent institute inspired partly by the Institute for Advanced Study in Princeton. Under founder and director Léon Motchane, its stated mission provided exceptional researchers with conditions for theoretical work outside ordinary university demands. The institute assembled permanent professors in mathematics and theoretical physics, welcomed visitors, ran seminars, and published research. Grothendieck and Dieudonné gave its early mathematical program international visibility.55Institut des Hautes Études Scientifiques, “Alexander Grothendieck, Permanent Professor from 1958 to 1970”; Jackson, “As If Summoned from the Void, I,” 1049–51.

Independence had several meanings. IHÉS was not an ordinary state university department. Its permanent researchers had unusual freedom in choosing problems and organizing their time. The institute cultivated international support and an international visitor program. Yet freedom from a single administrative hierarchy did not mean freedom from finance. Salaries, buildings, travel, secretarial work, printing, housing, and the cafeteria required regular income.

The 1969 statement displays a mixed economy. The largest line came from the Ministry of National Education. French and European members contributed another portion. An American support committee supplied funds. A government research delegation appeared alongside the Ministry of the Armies. “Private” and “public” did not divide the table into clean moral categories, nor did “international” mean nonpolitical. The institute’s independence rested on plurality and negotiation, not on an absence of patrons.

That arrangement had been one of Motchane’s central achievements. He persuaded industrial, governmental, and foreign supporters to finance theoretical research whose value could not always be described as an immediate application. Later judgments of his conflict with Grothendieck must retain that institutional work. To reduce Motchane to the administrator who accepted one disputed grant would be as misleading as reducing Grothendieck to a professor who objected to four percent of a budget.

The institute also depended on reputation. Grothendieck’s seminar attracted visitors and students; EGA appeared in Publications Mathématiques de l’IHÉS; the Fields Medal amplified an already international standing. His later memoir framed him and Dieudonné as participants in the institute’s scientific founding and described an expectation of remaining there. This retrospective account establishes his later interpretation, not his precise view at every earlier date. It suggests why an institutional budget line could carry significance beyond its amount.66Grothendieck, Récoltes et Semailles (Université des Sciences et Techniques du Languedoc and Centre National de la Recherche Scientifique, 1986).

The relation was reciprocal and unequal in different domains. Motchane controlled administration and fundraising within a board structure. Grothendieck controlled no bank account by virtue of mathematical distinction, but his presence carried scientific and reputational weight. Neither could simply define “the institute” alone. The coming dispute was therefore not a private consumer refusing a product. It was a conflict over who could set a condition for collective identity.

A contracting year

The January 1970 five-year financial program makes the pressure concrete. It called 1970 a year of regression relative to 1969, caused by government budget restrictions. The plan compared a constrained first version of the 1970 budget with an expanded alternative dependent on exceptional resources, especially German or British, and sought more international support over the following years.77Institut des Hautes Études Scientifiques, Archives de l’IHÉS, “Programme financier de l’Institut de 1970 à 1974,” January 1970, A2.1.7.3.3/6, pp. 1–3.

The document is aspirational and administrative. It translates scientific life into “professor-months,” salaries, visitor stays, equipment, and maintenance. Such translation can appear alien to research, but without it there is no seminar room, no long visit, and no time protected for mathematics. Grothendieck’s program had benefited enormously from the ability to gather people for months rather than hours.

The plan also reveals why a sudden refusal of funds could look different from different desks. To a permanent mathematician drawing an ethical boundary, the provenance of a small share might compromise the whole. To a director trying to keep commitments, withdrawal of a line could shorten visits or deepen a deficit. To physicists concerned with their own visitor program, earmarking the money might appear a practical separation. None of these perspectives can be inferred from job title alone; fortunately, the April minutes record versions of all three.

Later biographical narratives often describe the military money as hidden until Grothendieck discovered it. The audited statement shows that it appeared as a named income line. That does not establish who had seen the line, when, or what had been disclosed in earlier years. “Secret” should therefore be avoided unless a source documents concealment. The stronger fact is that financial transparency inside the professoriate was contested or insufficient for Grothendieck, since the issue became a condition of his remaining.

Cartier later identified DRET—the Direction des Recherches et Études Techniques—as a small agency associated with the Ministry of Defense and remembered a modest funding line. The 1969 statement itself says “Ministère des Armées,” not DRET. It would be careless to collapse every relevant payment into an exact agency label without matching records. The chapter therefore uses the wording of each source: Ministry of the Armies in the accounts and minutes; DRET when reporting Cartier’s later recollection.88Cartier, “A Mad Day’s Work,” 391–92; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1.

The late-1969 reconstruction

Jackson, summarizing Aubin’s institutional history, gives a sequence not fully visible in the records inspected here. That reconstruction says the permanent professors pressed Motchane in late 1969 to stop accepting military-origin funding and that a subsequent grant reversed the agreement. In his May 25 letter, Grothendieck contemporaneously asserted written assurances from Motchane and an oral promise from the board president. His letter of May 25, 1970, proves that he made those claims; it does not independently prove their terms or date.99Alexandre Grothendieck, Aux membres du Comité scientifique [de l’IHÉS], 1970, https://csg.igrothendieck.org/wp-content/uploads/2024/10/L202u.pdf, opening paragraph; Jackson, “As If Summoned from the Void, II,” 1197–98; David Aubin, “A Cultural History of Catastrophes and Chaos: Around the Institut Des Hautes Études Scientifiques, France, 1958–1980” (PhD diss., Princeton University, 1998).

The broader reconstruction is not on the same evidentiary level as the Price Waterhouse figures, April 22 vote, or his letter of May 25. The documents constituting the asserted assurances have not been recovered, and their exact wording is not established. The narrator therefore reports the allegation and reconstruction without stating that a verified promise was violated.1010Grothendieck, Aux membres du Comité scientifique [de l’IHÉS].

The direct records independently establish several surrounding facts. Military-origin income existed in 1969. The 1970 pre-budget faced a 200,000-franc gap. The January plan described contraction and sought new sources. The April 22 minutes report a conflict in the April 6 scientific committee. The board then approved the proposed 90,000-franc Ministry of the Armies grant. Those facts give Aubin’s sequence context without proving every link.1111Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of December 10, 1969, A3.1.4.1.1/1; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, “Programme financier de l’Institut de 1970 à 1974,” January 1970, A2.1.7.3.3/6; Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1.

The distinction separates Grothendieck’s documented condition and contemporary allegation from the truth and terms of the alleged agreement. The former is directly evidenced; the latter remains C/U.

The same rule applies to financial necessity. The documents show pressure. They do not prove that accepting this specific 90,000 was the only way to keep IHÉS open, nor that Motchane believed there was no alternative. His administrative motives should not be written as private certainty. The board chose acceptance in a constrained budget; that is enough.

April 6

The April 22 board minutes report that Grothendieck objected during the April 6 scientific committee to grants from the Ministry of the Armies and made refusal of such grants a condition of his continued presence as a permanent professor. No April 6 transcript independently confirms the board’s summary.1212Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, pp. 2–3.

That formulation deserves close reading. The minutes distinguish the grants’ undisputed institutional source from Grothendieck’s characterization of their military significance. The board record therefore supplies its own framing of his position rather than his exact words.

The minutes then use condemnatory language about his conduct at the meeting. This chapter does not repeat the label as neutral description. Minutes are produced by an institution to record and justify decisions; they are strong evidence that the board considered the behavior unacceptable and weaker evidence for an impartial characterization of an argument whose other transcript is absent. There are no verbatim exchanges from the April 6 meeting in the available record.

According to the same minutes, Grothendieck sent scientific-committee members a letter a few days later refusing further participation in the committee’s work. The board president treated that letter as a breach leading to cessation of his functions, while Motchane suggested a period for reflection. Because the letter is unavailable, its exact wording, scope, and intended duration remain unknown.1313Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, p. 3.

This moment precedes the May 25 decision. By April 22 the institutional relationship was already close to rupture. The board president saw a functional ending, while Motchane proposed delay. The surviving record therefore does not support a simple story in which a director immediately expelled a dissident professor.

The earlier committee-withdrawal letter remains unavailable, so its wording and intended duration are unknown. A separate May 25 letter to the Scientific Committee survives: Grothendieck announced his decision and a probable October 1 departure. His June 9 letter to the director confirmed October 1 as effective.1414Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, p. 3; Grothendieck, Aux membres du Comité scientifique [de l’IHÉS]; Alexandre Grothendieck, Letter to the director of the IHÉS confirming his departure, 1970, https://csg.igrothendieck.org/wp-content/uploads/2024/10/L203u.pdf.

Three positions in one page

The minutes record a letter from the permanent physicists Louis Michel and David Ruelle. They generally associated themselves with the wish not to receive military funds. They nevertheless argued that the unplanned loss of the proposed 90,000 francs for 1970 could create a difficult situation. Their proposed accommodation was to assign the money personally to the institute’s physicists and use it for invitations of physicists.1515Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, p. 3.

Their position cannot fairly be summarized as support for military funding. They shared the preference against it and differed over immediate institutional action. Nor can it be summarized as agreement with Grothendieck. They accepted a transitional use he rejected. The difference was ethical and practical: whether provenance could be contained by purpose and whether an abrupt refusal was required.

The board president responded that public subsidies could be made only to IHÉS, not to private persons. He also asserted that support from the Ministry of the Armies entailed no research of a military nature. The board then decided to accept the 90,000 francs and, with the director’s agreement, give priority to invitations of physicists.1616Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, p. 3.

The page thus preserves three positions. Grothendieck made refusal a condition of remaining. Michel and Ruelle preferred no military money but proposed accepting the current grant for a bounded purpose because sudden loss would be difficult. The board accepted the money, denied that it directed military research, and earmarked priority within the legal recipient institution.

These were not three opinions about the same empirical fact. They embodied different theories of complicity. For Grothendieck, the source crossed a boundary regardless of the immediate research purchased. For Michel and Ruelle, use and transition could matter enough to justify temporary acceptance. For the board, absence of a military research condition and institutional need made acceptance compatible with scientific independence.

The minutes do not state whether the physicists anticipated further grants, treated the proposal as a one-time bridge, or assessed the Ministry’s broader role. Their letter records a general wish to avoid military funds and a practical argument about the proposed 1970 amount; no further motive is inferred.

The board’s assertion of no military research also requires exact grammar. It is evidence of what the president said. The financial records in the available record show no weapons deliverable. That combination justifies writing that the documented dispute concerned provenance rather than a specified weapons project. It does not justify a universal claim that defense funding could exert no influence or that every relevant arrangement has been found.

A decision recorded through several voices

The April 22 minutes do not present a verbatim debate. They combine the director’s account of the April 6 scientific-committee conflict, the board president’s legal interpretation, a letter from Michel and Ruelle, and the board’s decision. Each layer has a different evidentiary status.

The director’s report is direct evidence that he informed the board and that the issue reached the scientific committee. It is not an independent transcript of Grothendieck’s words. The president’s interpretation establishes how the board leadership understood committee withdrawal; it does not establish that every legal implication was uncontested. Michel and Ruelle’s letter gives a stated position through the minutes, but the original letter should still be checked if recovered. The vote is the board’s own act and can be stated directly.1717Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, pp. 2–3.

This layered reading prevents the official document from becoming an omniscient narrator. It also prevents the opposite error of dismissing it because the board had an interest. Interested records can provide strong evidence for acts within the record-maker’s authority, especially sums, agenda items, and votes.

The sequence within the document is limited but important. Grothendieck’s condition preceded the board decision as reported. His withdrawal from the committee followed the April 6 conflict. The April 22 board accepted the proposed grant and prioritized physicist invitations. The later resignation date followed. No source in the available record supports inserting an undocumented private meeting between those steps.

The legal recipient question narrowed the compromise. Michel and Ruelle proposed that the funds be assigned personally to physicists. The president said a public subsidy could be paid only to IHÉS. The board then preserved the intended use through priority rather than personal allocation. This was an administrative modification of the proposal, not acceptance of Grothendieck’s demand.

The phrase no military research also belongs to a specific speaker and setting. The president used it to distinguish the grant from directed work. The minutes do not include the grant instrument, so the chapter states the board’s position and records the contract as missing. It does not infer a weapons project, and it does not certify every condition of the support.

This care changes the conflict’s narrative scale. The disagreement was not between a mathematician who saw a secret weapons contract and administrators who denied its existence. The surviving record concerns a named military-origin subsidy, an assertion of scientific nondirection, a proposed use for visitors, and a categorical objection to provenance.

The board’s small meeting also matters procedurally. The minutes list the participants and declare a quorum; professors’ positions entered through reports and letters rather than their presence as voting administrators. This was legally an institutional decision and socially a decision about a community larger than the room.

By keeping these levels separate, the chapter can state the result firmly without treating the board’s account of conduct as neutral. It can also represent Michel and Ruelle’s compromise without turning it into indifference to military funding.

Fungible money, bounded work

The word fungible describes money’s capacity to substitute. If a 90,000-franc grant pays for visiting physicists, other income that would have paid for those visitors becomes available elsewhere. Earmarking can identify use without isolating effect. This is one reason Grothendieck could reject the proposed accommodation even if he accepted the board’s claim that no military project was attached.

The opposite argument is also coherent. Institutions routinely receive money from governments with many functions. If scientific work remains investigator-chosen, results are public, and no military task is specified, provenance may not determine content. A designated visitor program can be judged by what it enables. On this view, refusing a small, unrestricted line during a shortfall may damage independent science without weakening military power.

The evidence does not appoint the biographer to decide which theory was morally correct. It does require showing why neither was a semantic trick. Grothendieck was not confused about earmarking; his objection reached the institutional relation. Michel and Ruelle were not claiming that the words “physicist invitations” changed where the funds came from; they were weighing source, use, and immediate consequences differently.

The Cold War and the Vietnam War made the question acute. Grothendieck had refused Moscow in political protest and had taught in North Vietnam during bombardment. French mathematicians around him held a range of left, antiwar, and antinuclear commitments. A Ministry of the Armies line in the budget did not enter a politically blank institute.

At the same time, theoretical physics had a historically complicated relation to states and military research. IHÉS was modeled partly on Princeton’s institute, whose postwar prestige included figures such as Robert Oppenheimer. None of this proves that the disputed grant controlled research. It explains why claims of pure separation could be ethically contested even when legal conditions protected theoretical freedom.

The title “The Price of an Institute” therefore has two meanings that should not be confused. There was a financial price: millions of francs, raised from mixed sources, to maintain the conditions under which EGA, SGA, and theoretical physics flourished. There was also Grothendieck’s documented condition that the institute refuse the military-origin grant if he were to remain. The same 90,000 francs entered both accounts.

Motchane and Grothendieck

Later accounts agree that the relationship between Motchane and Grothendieck had deteriorated before the final break. They differ in interpretation and evidentiary discipline. Scharlau suggested that the budget dispute may have provided an occasion for Motchane to rid the institute of a professor he regarded as intolerably disruptive, while also arguing that Grothendieck was already approaching a broader turn. This is Scharlau’s interpretation, not a fact established by the April minutes.1818Scharlau, “Who Is Alexander Grothendieck?” 933–35.

Cartier situated the Ministry of Defense subsidy within a wider set of possible pressures, including Grothendieck’s earlier political action and his doubts about scientific work. Cartier repeatedly used modal language. A biographer must preserve that uncertainty rather than convert a colleague’s interpretation into a diagnosis.1919Cartier, “A Mad Day’s Work,” 391–93.

Jackson’s narrative is closer to the institutional sequence and includes named recollections from Deligne and others. It says Grothendieck tried to persuade permanent professors to resign with him and was deeply disappointed when they did not. The claim that all professors had earlier shared misgivings is consistent with Michel and Ruelle’s April letter. The claim that they had all demanded an end to such funding comes through Aubin’s reconstruction and awaits direct documentary confirmation.2020Jackson, “As If Summoned from the Void, II,” 1197–98.

None of these accounts supports reducing the conflict to temperament. The board minutes record an institutional objection to Grothendieck’s conduct, while the audit and vote establish a substantive disagreement independent of that characterization. This account therefore reports the conduct judgment as the board’s and does not use personality to explain the decision.

Later witnesses who shared parts of Grothendieck’s politics criticized some tactics. Those A2 judgments remain attributed and do not convert the funding refusal into a diagnosis or a general character verdict.2121Jackson, “As If Summoned from the Void, II”; Scharlau, “Who Is Alexander Grothendieck?”; Cartier, “A Mad Day’s Work”.

Motchane’s suggestion of a reflection period on April 22 is important for balance. It shows that the director did not, at that recorded moment, demand immediate termination. The minutes do not state his motive. They establish only that he proposed delay after the president treated the committee letter as cessation.

Grothendieck’s sense of being a scientific co-founder makes the refusal more intelligible but not automatically binding on others. The legal board, director, physicists, and other permanent professors retained their own responsibilities. A person can believe an institution partly embodies his life’s work and still lack the authority to define its funding policy alone.

Independence through several patrons

The 1969 income table suggests one institutional defense of mixed funding. No single source supplied the entire budget. The Ministry of National Education was the largest public contributor; French and European members, an American committee, a research delegation, the Ministry of the Armies, and operating income supplied other parts. Diversification could reduce the leverage of any one patron.2222Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4, “État des produits”.

That defense concerns concentration, not provenance. A small contribution may carry little bargaining power and still create a relation that a professor rejects. Grothendieck’s condition addressed the relation itself. The board’s recorded position addressed whether the grant directed research. The two criteria can produce opposite decisions from the same table.

Unrestricted or broadly designated support is particularly complex. Its scientific advantage is that researchers choose topics rather than deliver a contracted result. Its ethical difficulty is fungibility: the grant frees other resources and associates the recipient with the donor even when no project is specified.

The April proposal for physicist invitations tried to use designation as a boundary. The board’s legal response kept the institute as recipient while preserving priority for that activity. This arrangement could limit immediate use, but it could not make the origin private or remove fungibility.

Transparency offered another possible boundary. If professors, visitors, and readers knew the donors and amounts, they could judge participation. The audit records the line, but the available record does not show when the information reached permanent professors or how widely the report circulated. An accessible later archive is not evidence of contemporaneous disclosure.

Plural funding also depended on administrative labor. Motchane’s role included securing support from industrial, governmental, and international sources for research without obvious short-term return. The achievement matters to the conflict because refusal had operational consequences. It does not decide which donors the institute should have accepted.

The scientific committee and board embodied different responsibilities. Professors could judge the relation’s compatibility with scientific life. Administrators held legal and fiduciary authority. A stable policy would need a procedure connecting those judgments. The April record instead shows positions carried through reports, letters, a condition, and a vote.

Grothendieck’s resignation made one form of accountability possible when governance did not produce his preferred result. He withdrew his participation rather than claim authority to reverse the board unilaterally. The act imposed costs on the institute and collaborators, but it accepted the personal cost of the condition he had stated.

The portfolio model and the provenance model remain analytically distinct. The first asks whether any donor can direct the institute. The second asks whether some donors should be refused even without direction. The 1970 break occurred because diversification did not answer Grothendieck’s objection.

This distinction prevents the later history from reducing independence to one metric. Financial plurality, investigator control, public disclosure, faculty participation, and donor eligibility are separate institutional properties. IHÉS could satisfy some and remain contested on another.

The value created by the money

The conflict is sometimes narrated as though money stood outside mathematics until it contaminated it. The institute’s records show the opposite. Money had always been inside the conditions of production. It paid for the time in which Grothendieck could prepare hundreds of pages. It enabled visitors to remain at Bois-Marie. It supported secretarial labor and publication. It maintained residences where families and scholars stayed. The absence of a commercial product did not make the work costless.

Grothendieck gave ideas, distributed prenotes, and produced extensive expositions used by a community. That circulation depended on an institution able to subsidize work whose recipients did not pay its full cost. Public and philanthropic money made the work materially possible.

This dependence does not refute his objection. It sharpens it. If free mathematics requires patrons, then the selection of patrons becomes part of the institution’s ethics. Refusal cannot abolish dependence; it can choose among dependencies and accept the consequences of a smaller budget. Grothendieck’s position required precisely that willingness to lose resources.

For a director, choosing among dependencies may involve preserving diversity so that no source dominates. The 1969 table can be read as such a portfolio. Yet a small source can have symbolic importance disproportionate to its amount. Governments do not fund only through control clauses; association itself can confer legitimacy. Grothendieck’s refusal treated acceptance as a public relation rather than a private transfer.

The board treated research purpose as the decisive boundary. It said the grant carried no military research and directed priority to visitors. That is a defensible institutional criterion, but it answers a narrower question than Grothendieck asked. The two sides could therefore repeat facts without approaching agreement.

A record with a voice

The April 22 minutes are unusually valuable because they preserve sequence, sums, and positions. They are also a crafted voice. The board met at its Paris office. The document presents the director’s report, the president’s judgments, letters read aloud, and the decision. Grothendieck was not listed among those present. His position reaches the page through the board’s summary.

That asymmetry does not invalidate the minutes. Official records are often the best evidence for what an institution decided. It requires separating record types. The vote to accept 90,000 francs is a board act. The statement that Grothendieck made refusal a condition is a report of his position. The adjective applied to his behavior is an institutional judgment. A responsible chapter can rely strongly on the first, cautiously on the second, and only attributively on the third.

The earlier committee-withdrawal letter, April 6 minutes, grant instrument, and documentation of the late-1969 agreement remain gaps. Grothendieck’s letters of May 25 and June 9, 1970, supply the later decision and the date of his departure.

The later Récoltes et Semailles supplies Grothendieck’s own perspective, but from more than a decade away and within a work concerned with the moral history of mathematics. He repeatedly treated military funding as decisive. That testimony establishes how he understood the break and is corroborated by the April condition. It cannot recover every tactical exchange or settle what Motchane promised.

This layering of sources is itself part of the story. The financial statement says what money was recorded. The board minutes say what the institution decided. Later colleagues say how the conflict appeared to them. Grothendieck says what it meant in the history of his life. A biography that blends them into one omniscient account loses the very differences that make the episode intelligible.

Before the letter of resignation

By the close of the April 22 meeting, the practical issue had been decided against Grothendieck’s condition. The board accepted the proposed grant. It declined the suggestion that the money be given personally to physicists because public funds had to go to the institute, but it incorporated the intended use by giving their invitations priority. Motchane’s call for reflection delayed an immediate conclusion about Grothendieck’s post.2323Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1, p. 3.

The board decision did not mechanically determine the later sequence. His letter of May 25 records his decision and a probable date of October 1; his letter of June 9 confirms October 1. The surviving documents do not supply a day-by-day narrative between April 22 and May 25.2424Grothendieck, Aux membres du Comité scientifique [de l’IHÉS]; Grothendieck, Letter to the director of the IHÉS confirming his departure.

It is tempting to end this chapter with a slammed door. There is no evidence for one. The available documents end in administrative prose: a grant accepted, a priority assigned, an assembly to be convened. The emotional and ethical magnitude lies beneath language designed for recordkeeping.

The board’s stated criterion was the absence of directed military research. Grothendieck’s documented condition rejected the military-origin grant itself. Michel and Ruelle’s letter combined a general wish against such funding with a transitional allocation proposal. These positions were not reconciled.

An institute created to protect theoretical freedom had encountered an issue its founding ideal did not answer: who defined freedom from direction, which sources could secure it, and which members judged those sources. The price could be entered in a ledger. The cost of refusing or accepting it could not.

Reading the audit line by line

The 1969 financial statement deserves more than extraction of one controversial line. It distinguishes budget from actual receipts. French and European members were budgeted at 444,500 francs and produced 404,550. The Ministry of National Education was budgeted at 1.7 million and recorded at 1.4 million for the year, with notes explaining accounting across exercises. The American committee line recorded 135,668.24 francs against 150,000 budgeted. The Ministry of the Armies recorded 90,000 against 100,000. The research delegation supplied its budgeted 130,000.

The pattern shows shortfalls in several sources. The military-origin line was not uniquely uncertain. Total received contributions and subsidies fell below the budgeted total of 2,524,500. Other income partly compensated, but total charges still exceeded products by 35,863.54 francs according to the statement. The institute entered 1970 without a large cushion.

These details matter because “Motchane needed the money” is too vague. Administrators planned against commitments while major receipts arrived below budget. A line worth 90,000 could cover several visitor-months or salaries. Rejecting it required either replacing it, cutting activity, or accepting a larger deficit.

The audit does not say which expense the 1969 Ministry line financed. Income statements pool resources unless a restriction is separately recorded. The April 1970 board’s proposed priority for physicist invitations concerns the new grant under debate, not necessarily the disposition of every 1969 franc. A caption must not merge the two years.

Nor does the audit reveal the grant’s contractual conditions. The Ministry of National Education note says its convention imposed conditions and that management believed those conditions were met. No comparable explanatory note appears beside the Ministry of the Armies line in the extracted table. The board president’s April statement supplies an institutional claim of no military research, but the actual grant instrument remains an uncollected gap.

This absence is important for both sides. Critics cannot assert a classified deliverable without a contract. Defenders cannot prove the money was wholly unrestricted without one. The strongest wording stays with provenance, amount, and the board’s recorded assertion.

The arithmetic can be reproduced:

$$\frac{90{,}000}{2{,}160{,}218.24}\approx 0.04166,\qquad \frac{90{,}000}{2{,}523{,}216.99}\approx 0.03567.$$

Rounded, those are 4.17 and 3.57 percent. If a designer turns them into a bar chart, both denominators should be labeled. A truncated axis that makes the military line visually equal to National Education would be propaganda. A chart so compressed that the line vanishes would also mislead.

The audited table provides something later memories cannot: proportion. Memory records betrayal, pressure, and principle. The ledger records scale. Neither supersedes the other.

Governance of scientific freedom

IHÉS separated roles among a director, board of administration, scientific committee, permanent professors, and external supporters. The exact statutes changed and internationalization was under discussion in 1969–1970. The conflict crossed these bodies because funding and scientific identity did not belong to the same office.

The scientific committee could express professors’ judgment about what support was compatible with research. The board held legal responsibility for accepting public funds and budgets. Motchane connected fundraising, daily direction, and long-term planning. Grothendieck’s scientific authority gave his condition immense weight without giving him a formal veto recorded in the minutes.

The April board meeting itself was small. The minutes list president André Grandpierre and Jacques Ballet, with Motchane attending as director, and state that a quorum was present. The document records no permanent professor in the room. Letters and the director’s report carried professorial positions into the decision.

This governance fact helps explain why participants could experience the result differently. From the board’s viewpoint, it exercised assigned fiduciary responsibility. From Grothendieck’s viewpoint, administrators overrode a condition touching the institute’s scientific honor. From Michel and Ruelle’s viewpoint, a compromise proposed by permanent scientists was translated into an institutional earmark.

The board president’s treatment of Grothendieck’s committee letter as a contractual rupture also asserted governance. Withdrawal from the scientific committee could not, in his view, be separated from permanent professorial duties. Motchane’s request for delay held open a different administrative response. The minutes do not record a vote terminating the appointment on April 22.

This sequence means the later resignation addressed both a donor relationship and a conflict over institutional authority. Grothendieck could set a condition for his participation. The board could reject the condition. Neither could make the other’s domain disappear.

An institute dedicated to free inquiry therefore required procedures for disagreement about support. The records do not show a pre-existing conflict-of-interest policy, disclosure rule, or faculty veto governing military-origin funds. The argument appears to have been resolved through personal and corporate positions rather than a published policy.

Modern readers may expect an ethics committee or donor-transparency page. Projecting those mechanisms backward would be anachronistic. Their absence, if confirmed, is historically revealing: institutions of pure science were still developing ways to govern the political meaning of heterogeneous funding.

Knowledge and consent

One unresolved question is what the professors knew before late 1969. The audited statement is clear after the fact. It does not prove when draft budgets or donor lists were circulated to permanent faculty. Jackson’s phrase that Grothendieck “discovered” the funding can suggest a concealed fact, but concealment requires evidence about expected disclosure.

Several possibilities fit the current record. The line may have appeared in administrative accounts not routinely read by scientists. Earlier grants may have been known generally but their exact source or continuation disputed. The issue may have become urgent only when a new 1970 grant was proposed after an agreement to refuse it. This account does not choose.

Consent to employment is not a one-time act if institutional conditions change. Grothendieck had joined an independent institute more than a decade earlier. If funding sources mattered to his willingness to remain, meaningful consent required information about them. His categorical response can be read as an attempt to make that condition explicit.

The board could reply that professors cannot individually approve every donor. Collective governance exists to prevent one member from controlling a budget. That answer is procedurally strong and ethically incomplete if no mechanism permits conscientious refusal or policy review.

Michel and Ruelle’s letter shows informed disagreement among scientists. They knew the source, shared a preference against it, and proposed a temporary allocation. Their participation makes the April decision more than an administrator-versus-mathematician confrontation. Scientific conscience did not speak with one voice.

Deligne’s later refusal to resign adds another layer. A newly appointed mathematician could oppose military influence and still judge departure wrong or unnecessary. The source records his decision and Grothendieck’s disappointment, not Deligne’s complete reasoning. Inferring consent to the grant from refusal to resign would be as careless as inferring endorsement of every institutional policy from employment.

The ethics of consent also concern visitors and readers. The archive does not establish who beyond the governance bodies knew that part of IHÉS income came from the Ministry of the Armies or whether that knowledge affected decisions. Grothendieck’s public break made the relation visible beyond those bodies.

Cold War categories

The phrase “military funding” contains several possible relations. A defense ministry may commission a specific weapons project. It may support basic physics believed strategically useful. It may give an unrestricted institutional subsidy. It may finance travel or equipment through an agency whose portfolio includes civilian research. The moral and practical implications differ.

The April minutes place the grant toward the less directive end of this spectrum as the board represented it: no military research, priority for physicist invitations. Grothendieck’s objection rejected the idea that absence of a deliverable ended responsibility. The Ministry of the Armies was part of a military apparatus whose purposes he opposed.

Cold War states invested in theoretical science partly because fundamental knowledge could acquire strategic value. Nuclear physics made the relation impossible to treat as hypothetical, even when a particular theorem had no immediate application. Governments also supported pure mathematics for prestige, education, and long-term capacity. One grant could participate in several rationales.

IHÉS’s mix of mathematics and theoretical physics sharpened the issue. Algebraic geometry appeared remote from weapons. Theoretical physics carried a more visible historical connection to military patronage, though the institute’s physicists pursued fundamental work. The proposal to use the funds for physicist visitors may have looked administratively natural and symbolically unacceptable for the same reason.

This chapter has no evidence that Michel or Ruelle conducted military research under the grant. Their names should not be placed beside the money in a way that implies it. The minutes identify them as proponents of a transitional allocation, not recipients of a weapons contract.

Grothendieck’s Vietnam experience gave military abstraction concrete images, but one should not narrate a direct mental chain without his dated words. His report described lectures interrupted by air raids and teachers killed after a bomb strike. Three years later he rejected money from a ministry of armies. The chronological juxtaposition is factual; its causal weight remains interpretation.

The French Ministry of the Armies was not the United States military conducting the bombing in Vietnam, and the Soviet state he protested in 1966 was another power. Grothendieck’s antimilitarism crossed blocs. This breadth prevents the funding refusal from being reduced to a partisan objection against one war.

Four possible tests

The participants’ disagreement can be clarified through four ethical tests, none of which the sources show them formally adopting.

The direction test asks whether the donor controls topics, methods, secrecy, or publication. The board’s “no military research” statement addresses this test. If accurate, the grant did not direct content.

The provenance test asks whether accepting resources from a given institution creates complicity regardless of control. Grothendieck’s condition addresses this test. The Ministry’s role in organized military power was sufficient reason for refusal.

The fungibility test asks whether earmarking truly isolates money. Michel and Ruelle proposed a bounded use; Grothendieck could reasonably answer that the subsidy freed other resources. The board’s legal response—that grants went to IHÉS, not individuals—made institutional fungibility explicit.

The consequence test asks what acceptance or refusal would do. Acceptance preserved visits and budget flexibility; refusal might signal independence and pressure the institute to find other patrons. Participants could disagree about which consequence mattered and whether symbolic action would influence military systems.

These tests do not produce one algorithmic answer. They show that the parties were not simply debating the definition of “military.” Each prioritized a different relation between money and work.

A fifth test, transparency, cuts across the others. Disclosure of donors and amounts to faculty and visitors remains relevant even when an institution accepts the funding. The audited statement supplied disclosure in an administrative record, but its timing and accessibility remain unclear.

Modern institutions use policies because personal confrontation is a costly way to answer recurring questions. In 1970 the absence of a shared test turned disagreement into an ultimatum. Grothendieck made remaining conditional; the board chose acceptance; individual colleagues had to decide whether to follow.

Counterfactuals the record cannot bear

The budget records show strain without establishing that IHÉS would have collapsed without 90,000 francs. The five-year plan identified possible German and British resources but no guaranteed replacement. No source determines whether Grothendieck would have stayed after refusal of only the 1970 grant.

A clearer policy might have offered a phase-out, segregated fund, or faculty disclosure rule, but the record does not show that Grothendieck would have accepted continued military-origin support. The counterfactual cannot be presented as an obvious missed solution.

The April letter indicates that other professors distinguished source from use, but it does not identify hypothetical thresholds for resignation. Their actual choice cannot predict every alternative.

Grothendieck’s Moscow and Vietnam actions predated the conflict, and broader concerns were developing. The record cannot reconstruct the precise organizational path his ecological activism would have taken without the funding dispute.

Counterfactual restraint is not timidity. It keeps attention on choices people actually faced: a proposed 90,000-franc grant during budget pressure and a stated refusal to remain if it was accepted.

Histories of the break

The episode has accumulated a hierarchy of retellings. Grothendieck’s retrospective work gives it moral and spiritual meaning. Cartier offers a colleague’s broad interpretation. Aubin reconstructs institutional finance and conflict. Jackson synthesizes archive history and interviews. Scharlau compares explanations and advances his own. Popular biographies often compress these into one sentence.

Compression usually drops attribution first. It can turn an uncertain discovery date into an immediate resignation, promote a reconstructed agreement into a verified breach, use an approximate percentage as moral dismissal, or make a minority income line sound like institutional ownership.

The evidence-led sentence is longer: audited accounts record a 1969 receipt of 90,000 francs from the Ministry of the Armies; a later reconstruction reports a disputed agreement; the April minutes document Grothendieck’s condition and the board’s decision; his letters document the May 25 decision, June 9 confirmation, and October 1 effective departure.

Length here is not stylistic indulgence. It preserves chronology and agency. Each clause has a source and a different confidence level.

The sources also have interests. Grothendieck sought to explain a turning point. The board sought to record and justify governance. Colleagues sought to remember a difficult person and period. Historians sought causal coherence. A biography should use their work without pretending to stand nowhere.

Fair wording does not split every difference. The Ministry line existed. The board accepted it. Grothendieck made refusal a condition. Those are not matters of perspective. Fairness enters where records characterize conduct, promise, necessity, and motive.

The institution each side wanted

Grothendieck’s position implied an institute whose scientific freedom included freedom from military patronage. Such an institute might be smaller, more dependent on civilian public funds, or required to cultivate other donors. Its integrity would be demonstrated by sources refused.

The board’s position implied an institute whose freedom was secured by nondirection and mixed support. A donor’s governmental identity did not determine research if no military work was required. Integrity would be demonstrated by scientific autonomy in use.

Michel and Ruelle’s compromise implied an institute able to recognize an undesirable relation, protect current commitments, and transition through earmarking. Integrity would include prudence toward visitors and programs already planned.

Motchane’s recorded role joined the second and third. He accepted the board framework and suggested reflection before ending Grothendieck’s functions. Beyond that, motive remains uncertain.

These models share more than polemic admits. All value scientific work not dictated by military objectives. They disagree over whether source itself constitutes direction or complicity and over how quickly principle must alter budgets.

The break occurred because the shared value did not yield a shared institution. That is more historically precise than saying one side cared about freedom and the other about money.

Trust after an alleged promise

If Aubin’s reconstruction of a late-1969 agreement is confirmed, the conflict involved trust as well as policy. An agreement to refuse future funds would have allowed permanent professors to believe the question settled. Acceptance a few months later could then appear not as a renewed debate but as a reversal.

Financial emergency can lead administrators to revisit commitments. Ethical legitimacy depends on how openly they do so and who has authority to consent. A promise made by a director may be constrained by a board’s legal duties; a board may still owe explanation to those who relied on it.

Because the underlying agreement is missing, this chapter cannot specify whether it was categorical, temporary, collective, or informal. The word promise may be stronger than the document would support. Jackson describes an agreement followed by a reversal when summarizing Aubin. A precise account would have to inspect Aubin’s evidence before choosing the noun.

The April minutes show trust already depleted. Grothendieck did not ask only for discussion; he made continued presence conditional. The board did not postpone the grant pending reconciliation; it accepted it. Ultimatum and decision left little space for gradual repair.

Motchane’s suggestion of reflection indicates some remaining procedural space. It came after the president treated committee withdrawal as cessation and before the board’s funding vote ended the relevant compromise. Reflection without policy change may have been insufficient by design.

Scharlau speculated that Motchane used the dispute to remove Grothendieck. This account records that interpretation but does not adopt it or construct an inverse motive for Grothendieck. The documented funding disagreement remains sufficient to explain the institutional conflict without treating it as a pretext.2525Scharlau, “Who Is Alexander Grothendieck?” 933–35.

Trust is historically visible through behavior more readily than motive. Colleagues did not resign together. Grothendieck withdrew from the committee. The board interpreted his letter contractually. These acts show failed confidence in common resolution.

Institutional values in different records

The sources frame the disagreement in different vocabularies. The April minutes record a condition of continued employment, a contractual interpretation of committee withdrawal, a proposed allocation, and a vote. The financial records describe revenue, shortfall, and budget. Grothendieck’s later memoir presents the break as part of an ethical and spiritual history. These records do not support a retrospective psychological diagnosis.

Grothendieck’s documented condition treated the source of the money as incompatible with remaining at IHÉS. The minutes do not reduce that condition to personal income or a dispute over research assignments. They show an institutional boundary.

The board’s records emphasize continuity and legal authority. Michel and Ruelle’s letter combined a general preference against military-origin funds with concern about the practical effect of abrupt withdrawal. Their proposed compromise failed because it did not satisfy Grothendieck’s condition and could not redirect a public grant to private persons.

The difference among contract, budget, and retrospective testimony is substantive. Each source records the episode for a different purpose. None supplies an unmediated account of every participant’s inner state.

The biographer’s task is not to invent a neutral vocabulary purified of values. It is to identify whose vocabulary appears. “Subsidy,” “military funds,” “support without military research,” and “apparatus” frame the same relation differently.

Quotations would make those frames visible, but direct quotation requires exact source checks. This account paraphrases the minutes and marks the later typescript unverified rather than selecting rhetorically powerful lines from an OCR.

What the record establishes

The episode can be stated in order of decreasing confidence.

The strongest documentary facts are the 90,000-franc receipt in the 1969 audit, the 200,000-franc planning gap, the April record of Grothendieck’s condition and the board’s acceptance, and the sequence of his departure in his own letters.

Important gaps remain: the reported late-1969 agreement’s exact terms, the April 6 minutes, the earlier committee-withdrawal letter, the 1970 grant instrument, and exact DRET routing.

Claims that Motchane reversed a promise, that the relationship was irreparable before April, or that exhaustion and earlier displacement determined the break remain attributed interpretation. No evidence in the available record supports the claim that IHÉS assigned Grothendieck to weapons work.

The chapter’s source count also should not be mistaken for proof by accumulation. One audit can outweigh ten later summaries on an amount. Several independent memories can establish that colleagues perceived a conflict while remaining weak on the exact words of a private exchange.

Different accounts deserve fair space without false balance. The grant existed and was accepted; no alternative account disputes that. Balance applies to its meaning and the broader cause of departure.

That is the evidentiary price of writing the institute accurately: a longer sentence, a visible denominator, and unanswered questions left unanswered.

Why pure mathematics could not stand outside

Grothendieck’s field seemed far from weapons. Schemes and algebraic cycles had no documented operational role in the disputed grant. One response is that his resignation proves an unusually expansive conscience. Another is that it mistakes association for use. The historical importance lies in making the distance itself a question.

Pure mathematics depends on education, communication, and protected time supplied by states and institutions. Its results may acquire applications unpredictably. These facts make both complete innocence and automatic guilt hard to defend. A theorem is not morally identical to every later use; a researcher is not socially unrelated to every patron.

Grothendieck’s position emphasized relation over forecast. He did not need to show that EGA would guide a weapon. Acceptance connected the institute to a ministry whose function he opposed. This was an institutional ethics of association.

The board emphasized autonomy over association. If research topics, publication, and appointments remained free, the source did not determine scientific purpose. This was an institutional ethics of nondirection.

Both models continue to shape research funding debates. Their persistence does not make Grothendieck the precursor of every modern policy. It shows that the 1970 conflict identified a durable problem.

The institute’s theoretical physics mission made purity itself plural. Fundamental physics can be conceptually remote from engineering and historically entangled with military states. Mathematics can appear remote and later become technological. A donor category cannot perfectly predict use.

The ethical choice therefore involves institutional identity: which relations an institute is willing to enter even when content remains free. Identity is expressed not only by research produced but donors accepted and refused.

Fairness to persons, firmness about acts

Fairness to Motchane requires acknowledging the institute he built, the financial pressure documented, and his suggestion of reflection. It does not require softening the board’s acceptance of the grant.

Fairness to Grothendieck requires treating antimilitarism as a documented principle, not symptom or pretext. It does not require endorsing every tactic or inferred accusation.

Fairness to Michel and Ruelle requires recording their stated preference against military funds and practical compromise. It does not turn their choice into Grothendieck’s.

Fairness to Deligne and other professors requires leaving their motives unassigned. Refusal to resign is an act, not a confession of indifference.

Fairness to the board requires distinguishing a claimed absence of military research from the money’s undisputed provenance. It does not let institutional minutes define neutral character.

These rules make the narrative asymmetrical where evidence is asymmetrical. The audit has exact numbers; private intention does not. Firmness attaches to acts and records. Modality attaches to motive.

Such wording may frustrate readers seeking hero and antagonist. The episode is morally consequential without a simple villain. Its conflict was strong enough to end one of twentieth-century mathematics’ most productive institutional relationships.

Three financial comparisons

Three comparisons appear in the chapter, and each answers a different question. The first compares that 1969 Ministry of the Armies receipt with all received contributions and subsidies. The result is about 4.17 percent. This denominator measures the grant among institutional support lines.2626Institut des Hautes Études Scientifiques, Archives de l’IHÉS, Price Waterhouse & Co., financial statements for 1969, February 27, 1970, A3.1.4.2.4/4.

The second compares the same receipt with all recorded income, including publications, housing, cafeteria receipts, and other resources. The result is about 3.57 percent. This denominator measures the grant within the institute’s total income statement.

The third comparison places 90,000 beside the 200,000-franc gap in the December 1969 planning figures for 1970. The amount is 45 percent of that gap. This is a scale comparison, not an accounting identity: the audited 1969 receipt and the proposed 1970 budget belong to different fiscal records.

The actual 1969 income statement also records an excess of charges over products of 35,863.54 francs. That realized annual deficit differs from the 200,000-franc prospective gap. One describes completed accounts; the other describes a plan before revenue was fully secured.

The table further distinguishes 100,000 francs budgeted from the Ministry of the Armies and 90,000 received. That ten-percent shortfall within the line should not be confused with the line’s share of the whole budget.

These comparisons explain why later prose can produce apparently inconsistent percentages without any source disagreeing. The denominator and fiscal question change. A responsible note must state both.

No conversion into current currency appears here. Such a conversion would require a defined inflation series and would still not capture the institutional value of visitor months, salaries, or publication. The contemporary budget ratios are enough for the chapter’s purpose.

The arithmetic establishes material scale. It does not prove whether the relation was ethically acceptable, whether another donor was available, or whether the grant determined research. Those remain separate institutional and interpretive claims.

The exact figures also keep the narrative from oscillating between triviality and domination. The line was a minority source and a substantial portion of the immediate planning gap. Both statements are true under stated denominators.

The price stated exactly

The title can now be resolved into four prices.

The budget price was measurable: a multi-million-franc institution with a 200,000-franc projected gap and a disputed 90,000-franc grant.

The scientific price of refusal was possible contraction in visits and protected work. The records show pressure but do not quantify which activity would have been cut.

The ethical price of acceptance was, for Grothendieck, participation in a relation to the military establishment. That valuation came from his stated condition and later account.

The relational price was the end of his permanent IHÉS life and damage among colleagues. It cannot be reduced to francs.

Different actors accepted different prices. The board accepted ethical controversy to preserve resources and its autonomy criterion. Grothendieck accepted institutional loss to preserve his provenance criterion. Michel and Ruelle proposed spreading the transition cost. No common unit compares them.

The exact sum prevents grandeur from floating free. The moral disagreement prevents the sum from deciding value. This double control is the chapter’s purpose.

On April 22 the board chose. Grothendieck notified the committee of his decision on May 25; on June 9 he fixed October 1 as the effective departure. The next chapter turns to the obligations created by that interval.2727Institut des Hautes Études Scientifiques, Archives de l’IHÉS, minutes of the board meeting of April 22, 1970, A3.1.4.2.2/1; Grothendieck, Aux membres du Comité scientifique [de l’IHÉS]; Grothendieck, Letter to the director of the IHÉS confirming his departure.

The disputed line was small enough for compromise to look possible and large enough for refusal to hurt. That middle scale intensified rather than softened conflict. Had the grant dominated the budget, dependence would have been obvious. Had it been negligible, replacement might have been easy. At roughly four percent of subsidy receipts and nearly half the immediate planning gap, it invited opposite descriptions that were both arithmetically grounded.

“Only” and “as much as” are therefore banned before the number. Readers can see scale and moral argument separately.

The institute paid a price whichever adjective they chose. Acceptance lost Grothendieck. Refusal would have lost funds and perhaps plans. The board made one cost actual; the counterfactual cost remains unknown.

The ledger cannot price the relation it ended. It can prevent that ending from becoming legend without scale. Ninety thousand francs, one board decision, and one stated condition are enough to ground the breach; everything further requires named evidence.