Endnotes

The main text carries no note numbers. Each note below begins with a phrase quoted from the passage it documents, and the notes appear in the order those passages occur. To find a source, match the words of the passage to the corresponding phrase under that chapter.

Each note also states what its sources establish—and what they do not. Those limits are part of the citation rather than qualifications added around it.

Introduction — I Have Seen This Future Before

“In 2020, Congress enacted the FAFSA Simplification Act…” — Consolidated Appropriations Act, 2021, Pub. L. No. 116-260, div. FF, title VII, 134 Stat. 1182 (2020); Consolidated Appropriations Act, 2022, Pub. L. No. 117-103, div. R, 136 Stat. 819 (2022); U.S. Department of Education, “FAFSA Simplification Act Changes for Implementation in the 2023–2024 Award Year,” 87 Fed. Reg. 66,683 (November 4, 2022); U.S. Department of Education, “Early Implementation of the FAFSA Simplification Act’s Removal of Requirements for Title IV Aid Eligibility Related to Selective Service Registration and Drug-Related Convictions,” 86 Fed. Reg. 32,252 (June 17, 2021). These sources establish the enactment, changed implementation schedule, phased implementation, and the lag between some legal changes and applicantfacing questions or warnings; they do not imply that legislative nesting or staggered implementation was itself deceptive.

“By spring 2024, some families could not complete the new form…” — U.S. Government Accountability Office, FAFSA: Education Needs to Improve Communications and Support Around the Free Application for Federal Student Aid, GAO-24-107407 (Washington, DC: U.S. Government Accountability Office, September 24, 2024), 1–2, 19–23; U.S. Government Accountability Office, Department of Education: Preliminary Results Show Strong Leadership Needed to Address Serious Student Aid System Weaknesses, GAO-24-107783 (Washington, DC: U.S. Government Accountability Office, September 24, 2024), 1–7; U.S. Department of Education, Federal Student Aid, “Update on Technical Fix to the 2024–25 FAFSA Form for Individuals Without a Social Security Number (SSN),” Electronic Announcement GENERAL-24-19 (March 12, 2024). GAO documents identity-access problems, supportcenter failures, submission declines, and system weaknesses; the Department’s announcement documents the March 2024 access fix. The cited figures describe the rollout period and do not establish that every applicant encountered the same problem.

“The fiscal and technical record added more pieces…” — U.S. Department of Education, Office of Inspector General, FAFSA Simplification Act and Student Loan Forgiveness Initiatives Accounting, ED-OIG/I24DC0194 (Washington, DC: U.S. Department of Education, February 27, 2025), Results in Brief, 3, 9–10; U.S. Government Accountability Office, Department of Education: Gaps in Federal Student Aid Contract Oversight and System Testing Need Immediate Attention, GAO-25-107396 (Washington, DC: U.S. Government Accountability Office, September 3, 2025), 8–9, 16–18; U.S. Department of Education, Federal Student Aid, “Response to GAO Draft Report GAO-25-107396,” letter to Marisol Cruz Cain (August 6, 2025), 2–4. The sources distinguish initiative accounting, estimated contract value, obligations, requirement tracking, and the Department’s description of multiple interdependent technical components. A contract ceiling, an obligation, and a delivered public service are different objects.

“Sweden established its Parliamentary Ombudsman in 1809…” — The Swedish Parliamentary Ombudsmen (Riksdagens Ombudsmän), “History,” updated June 7, 2024, https://www.jo.se/en/about-jo/ history/; and “Tasks and Role,” updated March 24, 2023, https:// www.jo.se/en/about-jo/tasks-and-role/. The official history links the 1809 office to parliamentary control after autocratic rule and describes the long shift from a primarily prosecutorial role toward advisory, consultative, and preventive supervision. The current role page describes independence from the executive, complaint and own-initiative investigations, inspections, criticism, advisory opinions, referrals, and other bounded powers.

“New Zealand adopted the institution in 1962…” — Office of the Ombudsman, New Zealand, “Past, Present and Future,” https:// www.ombudsman.parliament.nz/about-ombudsman/past-present-andfuture/, accessed August 22, 2026; Office of the Ombudsman, Annual Report 2016/2017; and Office of the Ombudsman, Annual Report 2006/2007, part III. New Zealand identifies itself as the fourth country, and the first outside Scandinavia, to appoint an Ombudsman in 1962. Its official histories describe the initial focus on central-government administration and the later expansion of jurisdiction and functions. The 2006/2007 report describes the office as an inexpensive and informal route for complaints about government administration relative to court proceedings; the comparison does not imply that an Ombudsman substitutes for judicial review or binding adjudication.

“In Sweden, an Ombudsman’s decisions are generally not legally binding…” — The Swedish Parliamentary Ombudsmen, “Decisions,” https://www.jo.se/en/decisions/, accessed August 22, 2026; and “How to Complain,” https://www.jo.se/en/how-to-complain/, accessed August 22, 2026. The Ombudsmen state that their decisions are not legally binding and that they cannot alter a judgment or ordinary decision that should instead be appealed through the applicable route. They also report that public authorities in practice generally follow Ombudsman criticism and often use it to improve procedures. These characteristics illustrate one institutional design, not a universal requirement for every ombudsman model.

Chapter 1 — Four Hundred Years to One Garden

This chapter recounts Luminaran history and makes no factual claim about Earth. Under the Note on Evidence, invented material is not accompanied by invented sources.

Chapter 2 — When Power Tries to Accelerate History

“On September 11, 2001…” — National Commission on Terrorist Attacks Upon the United States, The 9/11 Commission Report (Washington, DC, 2004), especially executive summary and chaps. 2 and 12, https://9-11commission.gov/report/911Report.pdf; George W. Bush, “Address to a Joint Session of Congress and the American People,” September 20, 2001, George W. Bush White House Archives, https:// georgewbush-whitehouse.archives.gov/news/ releases/2001/09/20010920-8.html; NATO, “Invocation of Article 5 Confirmed,” October 2, 2001, https://www.nato.int/en/news-andevents/articles/news/2001/10/02/invocation-of-article-5-confirmed. The 9/11 Commission documents al-Qaeda’s Afghan sanctuary and alliance with the Taliban. The Bush and NATO sources establish the publicly stated U.S. demands and NATO’s contemporaneous collective-defense determination; they do not by themselves settle every legal or strategic question concerning the subsequent war.

“When U.S. and British military operations began on October 7…” — George W. Bush, “Presidential Address to the Nation,” October 7, 2001, George W. Bush White House Archives, https://georgewbushwhitehouse.archives.gov/news/releases/2001/10/text/20011007-8.html. The address states that the initial strikes targeted al-Qaeda training camps and Taliban military installations to disrupt Afghanistan’s use as a terrorist base. It also describes humanitarian assistance. The source establishes the administration’s stated objectives, not an independent assessment of every military effect or later mission.

“In December 2001, Afghan representatives meeting under United Nations auspices…” — United Nations Peacemaker, Agreement on Provisional Arrangements in Afghanistan Pending the Re-establishment of Permanent Government Institutions (Bonn Agreement), December 5, 2001, https://peacemaker.un.org/en/documents/agreement-provisionalarrangements-afghanistan-pending-re-establishment-permanent; United Nations Security Council Resolution 1386 (2001), December 20, 2001; Kofi Annan, “Remarks to the Security Council on Afghanistan,” January 15, 2004, United Nations. The Bonn Agreement established the Interim Authority and a transitional route toward a constitution; Resolution 1386 authorized ISAF initially in Kabul and surrounding areas to assist the Afghan Interim Authority. Annan’s 2004 statement records adoption of the constitution by the Constitutional Loya Jirga. These sources show meaningful Afghan institutional participation and international authorization while leaving open the degree of external influence over particular choices.

“By 2003, the mission had acquired a much larger future…” — “Joint Statement Between the United States of America and Afghanistan,” February 27, 2003, George W. Bush White House Archives, https:// georgewbush-whitehouse.archives.gov/news/releases/2003/02/ text/20030227-22.html; NATO, “NATO Expands Presence in Afghanistan,” June 29, 2004, https://www.nato.int/en/news-and-events/ articles/news/2004/06/29/nato-expands-presence-in-afghanistan; NATO, “ISAF’s Mission in Afghanistan (2001–2014),” updated retrospective. The sources document a stated shift toward long-term reconstruction, institution-building, and the use of Provincial Reconstruction Teams to extend central-government authority and facilitate development. They establish the expansion of stated mission and mechanism; they do not establish that all participants shared one motive or theory of state-building.

“The shift had begun even earlier through the partners chosen to fight the Taliban…” — Special Inspector General for Afghanistan Reconstruction, Corruption in Conflict: Lessons from the U.S. Experience in Afghanistan, SIGAR 16-58-LL (Arlington, VA: SIGAR, September 2016), especially the discussion of the 2001 light-footprint strategy and partnership with Afghan warlords and Northern Alliance forces, https://www.sigar.mil/Portals/147/Files/Reports/Lessons- Learned/SIGAR-16-58-LL.pdf; Dick Cheney, “Interview of the Vice President by BBC,” November 15, 2001, George W. Bush White House Archives, https://georgewbush-whitehouse.archives.gov/vicepresident/ news-speeches/speeches/vp20011115.html. These sources establish the deliberate use of Afghan partners and the later U.S. oversight finding that support to some strongmen contributed to impunity, patronage, and corruption. They do not imply that all Northern Alliance figures behaved alike or that a feasible alternative military strategy was available on identical terms.

“The United States’ own reconstruction inspector later described…” — Special Inspector General for Afghanistan Reconstruction, What We Need to Learn: Lessons from Twenty Years of Afghanistan Reconstruction, SIGAR 21-46-LL (Arlington, VA: SIGAR, August 2021), executive summary and the sections on timelines, sustainability, local context, and “Imposing Formal Institutions on an Informal Environment,” https://www.sigar.mil/Portals/147/Files/Reports/ Lessons-Learned/SIGAR-21-46-LL.pdf. SIGAR concluded that U.S. timelines often underestimated the duration and resources required, encouraged rapid spending and unsustainable outputs, and insufficiently accounted for Afghan political, social, institutional, and human-capital conditions. In its rule-of-law discussion, the report states that between 2003 and 2015 the U.S. government spent more than $1 billion on ruleof-law programming, approximately 90 percent on the formal legal system, while citing estimates that 80–90 percent of civil disputes had historically been handled through informal mechanisms. SIGAR is a U.S. reconstruction oversight institution synthesizing its own reports, interviews, and other evidence; its account is substantial but not a complete or neutral history of Afghanistan.

“The archive forced me to extend that clock backward…” — U.S. Department of State, Office of the Historian, Foreign Relations of the United States, 1977–1980, vol. XII, Afghanistan, Preface, https:// history.state.gov/historicaldocuments/frus1977-80v12/preface; “Memorandum Prepared in the Central Intelligence Agency,” June 11, 1980, doc. 288, https://history.state.gov/historicaldocuments/ frus1977-80v12/d288; and United Nations Good Offices Mission in Afghanistan and Pakistan, “Background,” https://peacekeeping.un.org/ sites/default/files/past/ungomap/background.html. The FRUS material documents the Soviet military intervention of December 1979 and a U.S. covert-action program that began with nonlethal support before the invasion and expanded after it to lethal equipment delivered through Pakistan with Saudi participation. UNGOMAP records completion of the Soviet troop withdrawal in February 1989 and continuing hostilities requiring a political settlement. These sources establish external military and covert/proxy involvement and the withdrawal sequence; they do not provide a complete causal history of the Afghan wars, erase Afghan agency, or imply that later conflicts followed mechanically from any one sponsor’s actions.

“By 2020, according to the World Bank…” — World Bank, Afghanistan Development Update: Towards Economic Stabilization and Recovery (April 2022), special section “Afghanistan’s Economic Development from 2001 to 2021,” pp. 9–12, https://thedocs.worldbank.org/en/ doc/5f01165822f3639224e0d483ba1861fc-0310062022/original/ ADU-2022-FINAL-CLEARED.pdf. The report states that in 2020 grants equaled around 45 percent of GDP, financed around half of the government budget, 75 percent of total public expenditure, and 90 percent of security expenditures. It also documents both development gains and structural dependence associated with exceptionally high grant inflows. These figures demonstrate fiscal dependence on grants, not the absence of Afghan institutions, agency, or domestic revenue effort.

“The same international support that created dependencies also financed real improvements…” — World Bank, Afghanistan Development Update, April 2022, “Afghanistan’s Economic Development from 2001 to 2021.” The report attributes international support to expansion of state capacity and services and documents substantial improvements in growth, health, education, revenue collection, and other humandevelopment indicators. The chapter uses these findings to resist a onedirectional failure narrative; it does not imply that all gains were evenly distributed, sustainable, or caused by foreign assistance alone.

“SIGAR’s later review of the Republic’s collapse…” — Special Inspector General for Afghanistan Reconstruction, Why the Afghan Government Collapsed, SIGAR 23-05-IP (Arlington, VA: SIGAR, November 2022), https://www.sigar.mil/pdf/evaluations/SIGAR-23-05-IP.pdf. SIGAR identifies six contributing factors, including dependence on U.S. military and contractor support, the Afghan government’s exclusion from U.S.- Taliban talks, Taliban unwillingness to compromise, President Ashraf Ghani’s narrow governing circle, centralization, corruption, and legitimacy problems. The report supports a multi-causal account of collapse; its factor selection and weighting remain open to historical and political debate.

“After Japan’s defeat in 1945…” — U.S. Department of State, Office of the Historian, “Occupation and Reconstruction of Japan, 1945–52,” https://history.state.gov/milestones/1945-1952/japan-reconstruction; U.S. Department of State, Foreign Relations of the United States, Diplomatic Papers, 1945, vol. VI, “Instructions to General of the Army Douglas MacArthur,” September 6, 1945, doc. 491, https:// history.state.gov/historicaldocuments/frus1945v06/d491; National Diet Library of Japan, “Birth of the Constitution of Japan,” https:// www.ndl.go.jp/constitution/e/. The sources document the occupation’s broad reform program, MacArthur’s supreme authority, the decision to exercise control through the Japanese government where workable, and the constitutional process. They establish an externally directed occupation operating through surviving Japanese institutions; they do not prove that occupation policy alone caused Japan’s later political or economic performance.

“The contrast becomes sharper when the occupation’s method is made explicit…” — U.S. Department of State, Foreign Relations of the United States, Diplomatic Papers, 1945, vol. VI, “Instructions to General of the Army Douglas MacArthur,” September 6, 1945, para. 2, https:// history.state.gov/historicaldocuments/frus1945v06/d491; Douglas MacArthur, statement of September 17, 1945, FRUS 1945, vol. VI, doc. 495, https://history.state.gov/historicaldocuments/frus1945v06/d495; see also U.S. Department of State, Foreign Relations of the United States, 1949, vol. VII, part 2, discussion of occupation administration, https://history.state.gov/historicaldocuments/frus1949v07p2/d99. The September 6 instruction explicitly directed that control be exercised through the Japanese Government to the extent it produced satisfactory results; MacArthur later emphasized that continuing Japanese civil administration avoided the need to reconstruct the entire governmental apparatus. These records demonstrate institutional continuity under supreme occupation authority. They do not establish that bureaucratic continuity was sufficient for success, that Japanese actors were fully autonomous, or that the Japanese conditions were transferable to Afghanistan.

Chapter 3 — The Law of Uneven Development

“India’s first general election, held across 1951 and 1952…” — Election Commission of India, Leap of Faith: Journey of Indian Elections, 2nd ed., section on the First General Elections (1951–52), https:// www.eci.gov.in/EBooks/Leap_of_faith_Second_Edition/; Election Commission of India, India Votes: The General Elections 2014, discussion of literacy, symbols, and ballot-box design in the first general election. The Commission records an electorate of about 173.2 million and literacy of about 16.6 percent and describes adaptations that made voting usable for a largely non-literate electorate. The case is used to show that institutional operation depended on deliberately built administrative conditions, not to claim that early Indian democracy lacked exclusion, conflict, or institutional weakness.

“From the 1960s through the 1980s, South Korea industrialized at extraordinary speed…” — Joan E. Cho, Seeds of Mobilization: The Authoritarian Roots of South Korea’s Democracy (Ann Arbor: University of Michigan Press, 2024), especially chaps. 1–5; National Museum of Korean Contemporary History, online exhibition on democratization and the June 1987 Democratic Uprising, https:// www.much.go.kr/online_exhi/vnv/exh_info.html. Cho argues that South Korea’s path was nonlinear: industrialization and tertiary education initially helped bolster authoritarian rule but later facilitated mobilization by workers and students. The case is used to show changing political effects under changing relations, not to claim that economic development mechanically produced the 1987 transition.

“Earth’s own scholarship has spent decades arguing over this problem…” — Adam Przeworski and Fernando Limongi, “Modernization: Theories and Facts,” World Politics 49, no. 2 (1997): 155–183, https:// doi.org/10.1353/wp.1997.0004; Daron Acemoglu, Simon Johnson, James A. Robinson, and Pierre Yared, “Income and Democracy,” American Economic Review 98, no. 3 (2008): 808–842, https:// doi.org/10.1257/aer.98.3.808. Przeworski and Limongi found that development did not increase the probability of transitions to democracy in their 1950–90 sample, while affluence increased democratic survival. Acemoglu et al. found no causal income effect after accounting for persistent country differences and argued that the observed cross-country relation is consistent with divergent political-economic paths. These studies are major contributions to a contested literature, not final proof that economic and social development are irrelevant to political change.

“Hong Kong’s Independent Commission Against Corruption became influential enough…” — OECD, Specialised Anti-Corruption Institutions: Review of Models, 2nd ed. (Paris: OECD Publishing, 2013), https://doi.org/10.1787/9789264187207-en. The OECD notes that the success of earlier specialized institutions, including Hong Kong’s anti-corruption commission, encouraged many countries to create similar bodies, while emphasizing that specialized agencies cannot fight corruption alone and depend on wider integrity, control, and publicadministration institutions. The comparison concerns institutional ecology, not a ranking of national anti-corruption performance.

“The World Bank’s 2017 World Development Report states the broader principle…” — World Bank, World Development Report 2017: Governance and the Law (Washington, DC: World Bank, 2017), main messages and chaps. 1–2, https://www.worldbank.org/en/publication/ wdr2017. The report emphasizes that policy effectiveness depends on credible commitment, coordination, cooperation, and power relations, and advises analysts to examine institutional function rather than form alone. It supports the chapter’s conditional account of institutional performance; it does not establish one universal set of enabling conditions.

“Development practitioners have given one recurring failure an awkward but useful name…” — Matt Andrews, Lant Pritchett, and Michael Woolcock, Building State Capability: Evidence, Analysis, Action (Oxford: Oxford University Press, 2017), chaps. 2–3, https:// doi.org/10.1093/acprof:oso/9780198747482.001.0001. The authors use “isomorphic mimicry” for adoption of organizational forms that resemble capable institutions without reproducing function, and “premature load bearing” for imposing demanding tasks before capability is sufficient. The concepts are used here as diagnoses of form– function mismatch, not as a claim that institutional borrowing is inherently misguided.

“This is where the machine earns a place in my argument…” — Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https://doi.org/10.1162/tacl_a_00288; Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474. These works establish technical foundations for crosslingual semantic comparison and retrieval over external records. They do not show that an AI system can determine institutional equivalence, developmental stage, political legitimacy, or what a society should adopt.

Chapter 4 — Political Systems Are Alive

“In November 1993, New Zealanders were asked…” — New Zealand Electoral Commission, “The History of MMP” and “The Introduction of MMP,” Elections NZ, accessed August 22, 2026, https://elections.nz/ democracy-in-nz/what-is-new-zealands-system-of-government/what-ismmp/the-history-of-mmp/ and https://elections.nz/democracy-in-nz/ what-is-new-zealands-system-of-government/what-is-mmp/theintroduction-of-mmp/; Electoral Act 1993, New Zealand Public Act 1993 No. 87; New Zealand Parliament, “From Legislative Machine to Representative Forum? Procedural Change in the New Zealand Parliament in the Twentieth Century.” The official sources describe dissatisfaction with first-past-the-post outcomes, the 1993 referendum, the first MMP election in 1996, the emergence of a multi-party parliamentary environment, coalition or minority government, and changes to parliamentary procedure and select committees. They support the claim that one democratic system changed its representation and bargaining machinery without becoming a different regime; they do not establish that MMP is normatively superior on every dimension.

“In the 1980s, rural governance reforms began to establish directly elected villagers’ committees…” — Organic Law of the Villagers Committees of the People’s Republic of China, adopted November 4, 1998, arts. 2–4 and 11–15, National People’s Congress English text, https://www.npc.gov.cn/zgrdw/englishnpc/Law/2007-12/11/ content_1383542.htm; Congressional-Executive Commission on China, bilingual legal text, https://www.cecc.gov/resources/legal-provisions/ organic-law-of-the-villagers-committees-of-the-peoples-republic-ofchina. The law describes villagers’ committees as mass organizations of self-government, assigns village Party organizations a leading role, directs township governments to guide and assist rather than interfere in matters of lawful self-government, and requires direct election, direct nomination, secret ballots, and public vote counting. The legal text establishes the bounded institutional design, not the quality or binding force of every election in practice.

“Research on Chinese villages found wide variation in implementation…” — Kevin J. O’Brien and Lianjiang Li, “Accommodating ‘Democracy’ in a One-Party State: Introducing Village Elections in China,” The China Quarterly 162 (June 2000): 465–489, https://doi.org/10.1017/S0305741000008213; Jean C. Oi and Scott Rozelle, “Elections and Power: The Locus of Decision-Making in Chinese Villages,” The China Quarterly 162 (June 2000): 513–539, https://doi.org/10.1017/S0305741000008237. The studies document the institutionalization of village elections while emphasizing uneven implementation and the continuing importance of Party branches and higher-level authorities in determining practical power. They are used to distinguish access to an elected office from the authority that office can exercise, not to characterize all Chinese villages or infer a national democratization trajectory.

“The 1982 Constitution stated that the president and vice-president would serve no more than two consecutive terms…” — Constitution of the People’s Republic of China, adopted December 4, 1982, art. 79, National People’s Congress English text, https://www.npc.gov.cn/ zgrdw/englishnpc/Constitution/2007-11/14/content_1372954.htm; Constitution of the People’s Republic of China, current English text, arts. 79 and 87, National People’s Congress, https://www.npc.gov.cn/ englishnpc/constitution2019/201911/t20191120_384295.html; Xinhua, “Constitutional Amendment Good for China’s Enduring Peace, Stability,” March 11, 2018, https://www.xinhuanet.com/ english/2018-03/11/c_137032173.htm. The earlier Article 79 included a two-consecutive-term ceiling for the state president and vice-president; the current text does not. Article 87 continues to limit the premier, vicepremiers, and state councillors to two consecutive terms. The chapter treats this as a change in one formal succession constraint, not as a complete description of Party leadership, the Central Military Commission, or China’s overall succession politics.

“This is where machine-scale comparison becomes more than convenience…” — Comparative politics already contains mature efforts to disaggregate political systems rather than rely on regime labels alone. V-Dem, The V-Dem Dataset, version 16 (March 2026), https://www.vdem.net/data/the-v-dem-dataset/, publishes country-year and countrydate data with hundreds of separately defined indicators and indices, including coder-level data produced by country experts; Center for Systemic Peace, Polity5 Project and Polity5 Dataset Users’ Manual v2018, https://www.systemicpeace.org/polityproject.html and https:// www.systemicpeace.org/inscr/p5manualv2018.pdf, code regime authority characteristics and transitions over long historical series. These projects establish that multidimensional comparative coding is neither new nor uniquely enabled by AI. The chapter’s proposed machine contribution is different and narrower: lower-cost retrieval, version alignment, and candidate lineage construction over heterogeneous primary records, including institutional changes that were not specified in advance as indicators. Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https://doi.org/10.1162/ tacl_a_00288; Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474. These technical works support cross-lingual semantic comparison and retrieval over external records; they do not establish institutional equivalence, political progress, or a superior comparative method. The machine may surface candidate relations. Scholars and responsible institutions must still judge whether the relation is valid, consequential, and worth learning from.

Chapter 5 — Beyond Democracy and Dictatorship

“Singapore’s Housing and Development Board was created in 1960…” — Housing & Development Board, Singapore, “Our History”; HDB Key Statistics 2024/2025, reporting that 76.0 percent of Singapore’s resident population lived in HDB flats at end-March 2025. The sources establish the origin, scale, and persistence of the housing institution; they do not establish that Singapore’s wider political order is superior, that public housing outcomes are attributable to one regime feature, or that the model transfers intact to other societies.

“After the Supreme Court unanimously required President Richard Nixon to surrender the tapes…” — U.S. Senate, “Select Committee on Presidential Campaign Activities”; National Archives and Records Administration, “Watergate and the Constitution” and “Exhibit: Nixon and Watergate.” The Senate record states that United States v. Nixon required production of the tapes, the House Judiciary Committee adopted three articles of impeachment, and Nixon resigned on August 9, 1974. The episode is used to illustrate distributed institutional correction and peaceful succession, not to claim that the U.S. system corrected every Watergate-related harm or that democratic correction is always timely or effective.

“The People’s Action Party has won every general election since independence in 1965…” — Steven Oliver and Kai Ostwald, “Explaining Elections in Singapore: Dominant Party Resilience and Valence Politics,” Journal of East Asian Studies 18, no. 2 (2018): 129–156, https:// doi.org/10.1017/jea.2018.15; Elections Department Singapore, “2025 Parliamentary General Election Results.” Oliver and Ostwald describe PAP dominance, voter perceptions of credibility, and the use of a highcapacity state in the party’s resilience; the official 2025 results confirm continued PAP government after the most recent general election. These sources support treating Singapore as a dominant-party, high-capacity case; they do not justify inferring that every administrative achievement follows from party dominance.

“Norway’s Government Pension Fund Global and fiscal policy rule…” — Norwegian Ministry of Finance, “The Norwegian Fiscal Policy Framework,” and Report to the Storting No. 7 (2025–2026), The Management of the Government Pension Fund in 2025. The framework transfers petroleum revenues to the Government Pension Fund Global and guides withdrawals over time by the Fund’s expected real return, currently estimated at 3 percent; the government reports broad support across successive governments and parliaments. Alan M. Jacobs, Governing for the Long Term: Democracy and the Politics of Investment (Cambridge: Cambridge University Press, 2011), identifies electoral risk, prediction, and institutional capacity as obstacles to long-term policy in democracies. These sources support the narrower claim that democratic systems can deliberately construct long-horizon institutions despite short-term political pressures.

“During the Great Leap Forward, that same vertical reach…” — Xin Meng, Nancy Qian, and Pierre Yared, “The Institutional Causes of China’s Great Famine, 1959–1961,” Review of Economic Studies 82, no. 4 (2015): 1568–1611, https://doi.org/10.1093/restud/rdv016; Wei Li and Dennis Tao Yang, “The Great Leap Forward: Anatomy of a Central Planning Disaster,” Journal of Political Economy 113, no. 4 (2005): 840– 877, https://doi.org/10.1086/430804; Ning He and Wenbing Wu, “The Shadow Cost of State Violence: Evidence From Bureaucratic Purges in China,” Comparative Political Studies (2026), https:// doi.org/10.1177/00104140251400341. The cited research identifies inflexible procurement, resource diversion, political pressure, inflated production reporting, and excessive extraction as mechanisms contributing to famine mortality. The case concerns a specific Mao-era institutional configuration; it is not evidence that central planning, hierarchy, or every authoritarian government mechanically produces the same outcome.

“Research on modern autocracies finds leadership succession to be a recurrent vulnerability…” — Anne Meng, “Winning the Game of Thrones: Leadership Succession in Modern Autocracies,” Journal of Conflict Resolution 65, no. (2021): 950–981, https:// doi.org/10.1177/0022002720978807. Meng finds that leadership succession is a fundamental challenge for authoritarian regimes while showing that constitutional succession rules can facilitate peaceful transfer and regime continuation. The study supports treating succession as an institutional capability rather than an invariant property of a regime label.

“A recent study of its health-policy history describes a technocratic state with constrained civic freedoms…” — Ijlal Naqvi, Federico M. Rossi, and Rayner Kay Jin Tan, “Grievance Politics and Technocracy in a Developmental State: Healthcare Policy Reforms in Singapore,” Development and Change 55, no. 2 (2024): 244–275, https:// doi.org/10.1111/dech.12821. The process-tracing study identifies episodes in which social grievances, electoral signals, experts, and relatively independent groups influenced health-policy revision within Singapore’s constrained political environment. It does not establish that such feedback is equivalent to liberal-democratic contestation or that all policy domains are equally correctable.

“It can search across large, multilingual records for functionally similar arrangements…” — Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https://doi.org/10.1162/ tacl_a_00288; Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474. These works provide technical foundations for cross-lingual semantic comparison and retrieval over external records. They do not establish that an AI system can determine institutional equivalence, regime quality, political legitimacy, or the safeguards a society should adopt.

Chapter 6 — Many Gardens, Peacefully

“In April 1955, representatives of twenty-nine Asian and African governments met in Bandung, Indonesia…” — Final Communiqué of the Asian-African Conference, Bandung, Indonesia, April 18–24, 1955, reproduced in United Nations, International Organization and Integration: Annotated Basic Documents and Descriptive Directory of International Organizations, 2nd ed., vol. II.K (the communiqué lists the five sponsoring governments and twenty-four additional participants); Museum of the Asian-African Conference, Ministry of Foreign Affairs of the Republic of Indonesia, “KAA 1955.” The sources establish the dates, participation, and agreed declaration. They do not show that the participating governments shared a regime type, alliance, or later foreignpolicy practice.

“The conference ended with ten principles for relations among states…” — Final Communiqué of the Asian-African Conference, section G, “Declaration on the Promotion of World Peace and Co-operation”; Museum of the Asian-African Conference, “Dasasila Bandung.” The ten principles include respect for fundamental human rights and the UN Charter, sovereignty and territorial integrity, racial and national equality, non-interference, self-defense, limits on great-power pressure, nonaggression, peaceful settlement, cooperation, justice, and international obligations. The communiqué was a political declaration; citation here does not imply that it created an enforcement system or prevented later conflict.

“The Universal Declaration of Human Rights begins not with a preferred constitution but with the equal dignity and rights of human beings…” — United Nations General Assembly, Universal Declaration of Human Rights, Resolution 217 A (III), December 10, 1948, preamble and arts. 1–5, 7–10, 28–30; United Nations, “Universal Declaration of Human Rights.” The Declaration states a universal standard of human rights and was drafted by representatives from different legal and cultural backgrounds. It is used here to distinguish human standing from a complete institutional template, not to claim that rights interpretation or implementation is uncontested.

“The Constitutive Act of the African Union contains this tension in unusually concentrated form…” — Constitutive Act of the African Union, adopted July 11, 2000, art. 4(a), (e)–(j), (m), (o)–(p), African Union / African Commission on Human and Peoples’ Rights. Article 4 pairs non-interference by member states with peaceful coexistence and a Union right, pursuant to an Assembly decision, to intervene in grave circumstances including war crimes, genocide, and crimes against humanity. The Act establishes principles and institutional standing; it does not establish that these principles have been uniformly implemented or that every proposed intervention is legitimate.

“The United Nations Charter carries a related structure at global scale…” — Charter of the United Nations, arts. 1–2 and 51, especially art. 2(1), 2(3), 2(4), and 2(7), and ch. VII; United Nations, “Chapter I: Purposes and Principles” and “United Nations Charter (full text).” The Charter combines sovereign equality, peaceful settlement, the prohibition on threat or use of force, limits on intervention by the Organization in domestic jurisdiction, self-defense, and collective-security mechanisms. The passage uses this architecture to distinguish default restraint from categorical non-action, not to resolve contested questions of international-law application.

“The Antarctic Treaty gives a remarkably literal example…” — Antarctic Treaty, signed at Washington December 1, 1959, entered into force June 23, 1961, arts. I–IV, United Nations Treaty Series 402, no. 5778; Secretariat of the Antarctic Treaty, “The Antarctic Treaty.” The twelve original signatories included the United States and Soviet Union; the treaty reserves Antarctica for peaceful purposes, continues scientific cooperation and exchange, and preserves existing legal positions on territorial claims while barring new or enlarged claims during its operation. The example establishes a bounded cooperative architecture among politically different states, not that the treaty resolved their wider rivalry or supplies a general model for all international disputes.

“Modern multilingual systems can translate across far more languages than any individual scholar can command…” — NLLB Team, “Scaling Neural Machine Translation to 200 Languages,” Nature 630 (2024): 841–846, https://doi.org/10.1038/s41586-024-07335-x; Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https:// doi.org/10.1162/tacl_a_00288. The first demonstrates machinetranslation coverage and evaluation across 200 languages; the second demonstrates cross-lingual semantic representations and similarity search across many languages. These technical results support translation and candidate retrieval, not cultural understanding, institutional equivalence, political legitimacy, or an AI right to recommend political reform.

Chapter 7 — The Naked King

“Congress enacted the FAFSA Simplification Act in 2020…” — Consolidated Appropriations Act, 2021, Pub. L. No. 116-260, div. FF, title VII, 134 Stat. 1182 (2020); FAFSA Simplification Act Technical Corrections Act, Pub. L. No. 117-103, div. R, 136 Stat. 819 (2022); U.S. Department of Education, “FAFSA Simplification Act Changes for Implementation in the 2023–2024 Award Year,” 87 Fed. Reg. 66,683 (November 4, 2022). These sources establish the statutory changes, later timetable, and phased implementation authority. They do not by themselves establish whether the redesigned application was practically simpler for every group of applicants.

“Federal law removed Selective Service registration and covered drug convictions…” — U.S. Department of Education, “Early Implementation of the FAFSA Simplification Act’s Removal of Requirements for Title IV Aid Eligibility Related to Selective Service Registration and Drug-Related Convictions,” 86 Fed. Reg. 32,252 (June 17, 2021). The notice instructed institutions to disregard specified questions and warning codes after the underlying eligibility consequences had changed. The example demonstrates staggered policy and interface versions, not concealment.

“Some families with a contributor who lacked a Social Security number…” — U.S. Department of Education, Federal Student Aid, “Update on Technical Fix to the 2024–25 FAFSA Form for Individuals Without a Social Security Number (SSN),” Electronic Announcement GENERAL-24-19 (March 12, 2024); U.S. Department of Education, Federal Student Aid, “FAFSA Issue Alerts,” entry concerning a parent with no SSN; U.S. Government Accountability Office, FAFSA: Education Needs to Improve Communications and Support Around the Free Application for Federal Student Aid, GAO-24-107407 (Washington, DC: GAO, September 24, 2024), 1–2, 19–23. GAO reported that nearly three quarters of calls went unanswered during the first five months and that first-time submissions were about 9 percent lower than in the prior cycle, with the largest declines among lower-income applicants. These figures describe the rollout period studied, not every applicant’s experience.

“In February 2025, the Department of Education’s inspector general reported…” — U.S. Department of Education, Office of Inspector General, FAFSA Simplification Act and Student Loan Forgiveness Initiatives Accounting, ED-OIG/I24DC0194 (Washington, DC: U.S. Department of Education, 2025). The report found that initiative-level accounting information was not readily available in the form requested and discussed late spend-plan submissions and limited initiative-specific detail. The finding concerns accounting visibility and management information; it is not a finding that the reported expenditures were unlawful.

“The Government Accountability Office reported that the contract had been initially valued…” — U.S. Government Accountability Office, Department of Education: Gaps in Federal Student Aid Contract Oversight and System Testing Need Immediate Attention, GAO-25-107396 (Washington, DC: GAO, 2025), especially 10–12, 16– 17, 24–31, 41–42; Federal Student Aid, “Response to GAO Draft Report.” GAO reported the contract’s initial estimated value, later estimated total after modifications, obligations as of May 2025, and oversight gaps. The Department emphasized that the FAFSA service depended on several interdependent systems and vendors and disputed aspects of a requirements-centered assessment. The figures are used to distinguish contract states and service architecture, not to infer waste, corruption, or single-vendor causation.

“Research on retrieval-augmented language systems and source attribution…” — Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474; Hannah Rashkin et al., “Measuring Attribution in Natural Language Generation Models,” Computational Linguistics 49, no. 4 (2023): 777–840, https:// doi.org/10.1162/coli_a_00486; Tianyu Gao, Howard Yen, Jiatong Yu, and Danqi Chen, “Enabling Large Language Models to Generate Text with Citations,” in Proceedings of EMNLP 2023, 6465–6488, https:// doi.org/10.18653/v1/2023.emnlp-main.398. These works establish technical foundations for retrieval over external records and sourceattributed generation and also treat citation quality as an evaluation problem. They do not establish that current systems can autonomously reconstruct a complete political reality, determine which sources are legally controlling, or assign political responsibility.

“Earth offers some evidence that timely visibility can alter behavior…” — Monika Bauhr, Ágnes Czibik, Jenny de Fine Licht, and Mihály Fazekas, “Lights on the Shadows of Public Procurement: Transparency as an Antidote to Corruption,” Governance 33, no. 3 (2020): 495–523, https://doi.org/10.1111/gove.12432. Using more than 3.5 million European government contracts from 2006 through 2015, the authors found that tender transparency was associated with lower corruption risk measured through single bidding, with the result driven largely by ex ante transparency that enabled monitoring by actors inside the bidding process. Single bidding is used as a corruption-risk proxy; the study does not establish corrupt intent in any individual contract or prove that transparency always changes behavior.

“Expected disclosure can also change where the record is created…” — Robert Hazell, Ben Worthy, and Mark Glover, “The Impact of Freedom of Information on Whitehall,” Public Administration 89, no. 4 (2011): 1664–1681, https://doi.org/10.1111/j.1467-9299.2011.01930.x; House of Commons Justice Committee, Post-legislative Scrutiny of the Freedom of Information Act 2000, First Report of Session 2012–13, HC 96-I (London: House of Commons, 2012), paras. 183–200. The two-year Whitehall study found that the feared general chilling effect on candor and recordkeeping had not broadly materialized. The Justice Committee nevertheless recorded a minority of interviewees describing oral briefing, attenuated minutes, informal meetings, or other changes partly associated with expected disclosure, while emphasizing that the evidence was mixed and that other changes in government working practices also affected recordkeeping. The sources therefore support a bounded possibility of record-shifting, not a general claim that freedomof-information law destroys official records.

Chapter 8 — A Civilization That Can Remember

“On January 28, 1986, the Space Shuttle Challenger broke apart shortly after launch…” — Presidential Commission on the Space Shuttle Challenger Accident, Report to the President (Washington, DC, June 6, 1986), especially chap. V and the Commission’s recommendations; NASA History Office, “Challenger STS-51L Accident.” The Commission identified the booster-joint failure and found the launch decision process flawed, including that senior decision-makers lacked material information about recent O-ring concerns and the contractor engineers’ launch objection. The case is used to distinguish available records from information that actually reached a consequential decision, not to claim that every relevant fact was known uniformly across NASA.

“NASA did not put the report on a shelf and do nothing…” — National Aeronautics and Space Administration, Implementation of the Recommendations of the Presidential Commission on the Space Shuttle Challenger Accident (Washington, DC, 1986–1987), Executive Summary and recommendations I–IX. NASA documented redesign, management, safety, hazard-review, maintenance, communications, and flight-rate actions in response to the Commission. The later Columbia findings therefore should not be read as evidence that no post-Challenger reform occurred.

“Seventeen years later, Columbia was lost during re-entry…” — Columbia Accident Investigation Board, Report, vol. 1 (Washington, DC: Government Printing Office, August 2003), especially chaps. 7–8; NASA History Office, “Columbia Accident Investigation Board Synopsis,” January 16, 2004. The Board identified a different physical failure from Challenger while locating organizational causes in Shuttle history and culture, discussing “Echoes of Challenger,” weaknesses in safety independence and communication, inadequate trend monitoring, and the effects of schedule and resource pressures. Its historical comparison supports the chapter’s claim that institutional lessons can weaken or lose operational relevance; it does not establish that Columbia was a simple repetition of Challenger or that every actor possessed the later Board’s integrated view.

“Current information systems already contain pieces of a different possibility…” — World Wide Web Consortium, PROV-DM: The PROV Data Model, W3C Recommendation (April 30, 2013); Herbert Van de Sompel, Michael L. Nelson, and Robert Sanderson, HTTP Framework for Time-Based Access to Resource States—Memento, RFC 7089 (December 2013). PROV-DM represents entities, activities, agents, derivations, and temporal information; Memento supports access to prior states of changing web resources. They are cited as components for versioned provenance and temporal retrieval, not as a complete politicalmemory architecture.

“Computational methods can identify events, dates, and temporal relations across documents…” — James Pustejovsky et al., “TimeML: Robust Specification of Event and Temporal Expressions in Text,” IWCS-5 (2003); Egoitz Laparra, Itziar Aldabe, and German Rigau, “Document Level Time-anchoring for TimeLine Extraction,” in Proceedings of ACL-IJCNLP 2015, 358–364. These works support structured temporal annotation and timeline extraction. They do not establish that an automated system can determine political causation, relevance, or responsibility from chronology alone.

“Machine scale could make that mistake faster…” — Bhuwan Dhingra et al., “Time-Aware Language Models as Temporal Knowledge Bases,” Transactions of the Association for Computational Linguistics 10 (2022): 257–273, https://doi.org/10.1162/tacl_a_00459; Zheng Chu et al., “TIMEBENCH: A Comprehensive Evaluation of Temporal Reasoning Abilities in Large Language Models,” in Proceedings of ACL 2024, 1204–1228. These studies document continuing limits in timesensitive knowledge and temporal reasoning. They support the requirement for case-specific, time-stamped, source-linked records rather than treating model memory as authoritative political chronology.

“One Canadian record made the distinction concrete for me…” — Liberal Party of Canada, “Planting Two Billion Trees and Using the Power of Nature to Fight Climate Change” (2019); Department of Finance Canada, Supporting Canadians and Fighting COVID-19: Fall Economic Statement 2020 (Ottawa: Department of Finance Canada, November 30, 2020), chap. 3.3.2.3; Commissioner of the Environment and Sustainable Development, Forests and Climate Change, Report 1 of the 2023 Reports 1 to 5 (Ottawa: Office of the Auditor General of Canada, April 2023), especially paras. 1.25–1.43; Natural Resources Canada, “Our Action: How We Will Get There Together,” updated January 12, 2026, accessed August 13, 2026; Department of Finance Canada, Budget 2025: Canada Strong (Ottawa: Department of Finance Canada, November 4, 2025), chap. 5, “Wrapping Up the Tree Planting Program”; House of Commons Standing Committee on Public Accounts, Forests and Climate Change, Report 4, 45th Parliament, 1st Session, presented December 8, 2025, recommendations 1–2; Government of Canada, Government Response to the Recommendations in the “Forests and Climate Change” Report of the Standing Committee on Public Accounts, presented April 13, 2026, sections 1.2–1.3 and 2.2. These sources establish the original commitment, subsequent funding and program implementation, the 2023 audit findings, reported planting through June 2025, the later wind-down, surviving agreements, changed emissions expectations, and revised monitoring. They do not establish motive or whether the revision was politically or morally sufficient.

Chapter 9 — Learning Across Difference

“Sweden’s Parliamentary Ombudsman began in 1809…” — The Swedish Parliamentary Ombudsmen (Riksdagens Ombudsmän), “History,” updated June 7, 2024, https://www.jo.se/en/about-jo/history/. The official history links the 1809 office to constitutional control after autocratic rule, describes its independence from the executive, and traces the long shift from a primarily prosecutorial role toward criticism, advice, consultation, and prevention. The history establishes the institutional lineage used here; it does not establish that every later ombudsman institution reproduces the Swedish design.

“A Swedish Ombudsman cannot simply replace an ordinary appeal…” — The Swedish Parliamentary Ombudsmen, “How to Complain,” “When a Complaint Is Submitted,” and “Decisions,” https://www.jo.se/en/ how-to-complain/ and https://www.jo.se/en/decisions/, both accessed August 22, 2026. The office states that it cannot alter judgments or ordinary decisions that should be appealed through the proper route, that complaints may be screened for jurisdictional and other reasons, and that its decisions are not legally binding. It also states that authorities in practice almost always follow Ombudsman criticism and often use it to improve internal rules and procedures. These are characteristics of the Swedish arrangement, not universal conditions for administrative correction.

“New Zealand makes the next step visible…” — Office of the Ombudsman, New Zealand, “Past, Present and Future,” https:// www.ombudsman.parliament.nz/about-ombudsman/past-present-andfuture/, accessed August 22, 2026. New Zealand identifies its first Ombudsman as appointed in 1962, the fourth such office in the world and the first outside Scandinavia, and documents later extensions to education and hospital boards, local government, official-information review, protected-disclosure functions, and detention inspection under OPCAT. The source establishes institutional mutation and expansion of jurisdiction; it does not show that every added function improved performance or that the New Zealand model should be copied elsewhere.

“The Council of Europe’s Venice Principles are revealing for this reason…” — European Commission for Democracy through Law (Venice Commission), Principles on the Protection and Promotion of the Ombudsman Institution (The Venice Principles), CDL- AD(2019)005 (2019), especially principles 1–4, 8, 9, 11–16, and 21–25; Council of Europe, “25 ‘Venice Principles’ — Democratic ABCs for Ombudsman Institutions,” March 19, 2019. The principles explicitly recognize that there is no standardized ombudsman model and identify independence, legal foundation, resources, access, mandate, cooperation, and protection from interference as conditions of effective operation. They are normative standards derived from comparative experience, not empirical proof that any office satisfying them will perform well.

“South Africa then prevented me from turning those conditions into a universal blueprint…” — Constitution of the Republic of South Africa, 1996, secs. 181–182, South African Government, https://www.gov.za/ documents/constitution/constitution-republic-south-africa-1996chapter-9-state-institutions. The Constitution establishes the Public Protector as an independent institution supporting constitutional democracy, requires accessibility, and authorizes investigation, reporting, and appropriate remedial action. The constitutional text establishes formal authority and independence requirements; it does not by itself establish the effectiveness of every Public Protector investigation or remedy.

“In South Africa, the Constitutional Court held in the 2016 Economic Freedom Fighters case…” — Economic Freedom Fighters v Speaker of the National Assembly and Others; Democratic Alliance v Speaker of the National Assembly and Others [2016] ZACC 11, especially paras. 52–76. The Constitutional Court held, in the circumstances before it, that the Public Protector’s remedial action could not simply be ignored and that those affected had to comply unless the action was properly set aside through legal process. The case is used to show a materially stronger consequence attached to one ombudsman-family mechanism; it does not mean the Public Protector is a court or that all of its findings are immune from judicial review.

“The International Ombudsman Institute alone describes a network…” — International Ombudsman Institute, “About the IOI,” https:// www.theioi.org/the-i-o-i/about-the-ioi/, accessed August 22, 2026. The IOI describes cooperation among around 200 independent ombudsman institutions from more than 100 countries, organized across six regional chapters. The count establishes global institutional diversity and scale; membership does not establish equivalent mandate, independence, authority, or effectiveness across the institutions.

“This is the point where machine scale changes the comparison…” — Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https://doi.org/10.1162/tacl_a_00288; Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474; Hannah Rashkin et al., “Measuring Attribution in Natural Language Generation Models,” Computational Linguistics 49, no. 4 (2023): 777–840, https://doi.org/10.1162/coli_a_00486. These works establish technical foundations for cross-lingual semantic comparison, retrieval over external records, and evaluation of source attribution. They do not establish that a model can autonomously determine institutional equivalence, causal effectiveness, legal standing, or what a society should borrow.

Chapter 10 — Cross-Pollination, Not Transplantation

“In 1946 the government created the Commissariat général du Plan…” — Daniel Agacinski, Julien Bueb, Anne Faure, Julien Fosse, Hélène Garner, Boris Le Hir, Gautier Maigne, and Mathilde Viennot, “La planification : idée d’hier ou piste pour demain ?,” France Stratégie, June 17, 2020, 1–4, https://www.strategie.gouv.fr/files/2025-01/fs-2020point-de-vue-planification-juin_2.pdf. France Stratégie dates the Commissariat’s creation to 1946 under Jean Monnet, describes the first Plan’s reconstruction and modernization aims, and traces direct inspiration from wartime production-coordination boards in which Monnet participated. It explicitly distinguishes the French arrangement from Soviet Gosplan. The source establishes institutional lineage and stated design; it does not establish that planning alone caused France’s postwar growth.

“Its commissions brought together administrators, employers, trade unions, experts, and sectoral participants…” — France Stratégie, “La planification,” 3–5; OECD, “Planning for Growth: The French Experiment,” OECD Observer, no. 4 (June 1963), especially 10–15, https://www.oecd.org/content/dam/oecd/en/publications/ reports/1963/06/oecd-observer-volume-1963-issue-3_g1g3413b/ observer-v1963-4-en.pdf. France Stratégie describes consultation with social partners and an ‘économie concertée’; the OECD account describes tripartite planning commissions, parliamentary approval, and commissions that did not impose firm-level production targets. These sources support the distinction between long-horizon coordination and a command economy while also showing substantial state influence through budgets, nationalized industries, and public policy.

“As the economy became more open and complex, later plans became more indicative…” — France Stratégie, “La planification,” 3–7. The retrospective notes that later plans had increasingly indicative reach and less reliable forecasts as their scope widened and international dependence increased; it also records the transformation of the Commissariat into the Centre d’analyse stratégique in 2006 and France Stratégie as its institutional descendant from 2013. These observations establish mutation and declining fit of the earlier planning form, not a verdict that the institution failed or should be revived.

“In 1989 the National People’s Congress adopted an Administrative Litigation Law…” — National People’s Congress of the People’s Republic of China, Administrative Litigation Law, adopted April 4, 1989, effective October 1, 1990, arts. 1–5, https://www.npc.gov.cn/ zgrdw/npc/lfzt/2014/2013-12/20/content_1817953.htm. Article 2 gave citizens, legal persons, and other organizations a statutory right to sue over specific administrative acts alleged to infringe lawful rights and interests; the law assigned people’s courts jurisdiction over administrative cases. The text establishes the formal correction route, not its effectiveness or independence in every case.

“The amended law therefore made the right to sue more explicit…” — National People’s Congress Standing Committee, Decision on Amending the Administrative Litigation Law, November 1, 2014, effective May 1, 2015, https://www.npc.gov.cn/zgrdw/npc/ xinwen/2014-11/02/content_1884662.htm; National People’s Congress, “Explanation of the Draft Amendment to the Administrative Litigation Law,” December 31, 2013, https://www.npc.gov.cn/zgrdw/npc/ lfzt/2014/2013-12/31/content_1822189.htm. The official explanation identifies difficulty in case filing as a major reform problem; the enacted amendment required courts to protect the right to sue, barred administrative interference with case acceptance, expanded reviewable matters, and required responsible officials or appropriate representatives to appear. These are formal reforms, not proof that interference or filing barriers disappeared.

“China’s Supreme People’s Court reported 241,000 first-instance administrative cases accepted in 2015…” — Supreme People’s Court of the People’s Republic of China, The Work of China’s Courts 2015, section 4, https:// english.court.gov.cn/pdf/TheWorkofChina%27sCourts2015.pdf; Supreme People’s Court, Chinese Court Judicial Reform (2013–2018), 150–151, https://english.court.gov.cn/specials/ ChineseCourtJudicialReform2013-2018.pdf. The 2015 report states that courts accepted 241,000 first-instance administrative cases, 59.2 percent more than in 2014, and notes experiments with trans-regional jurisdiction; the reform report states that such jurisdictional changes were intended to reduce local administrative interference. Official court reports describe institutional objectives and case volumes; they do not independently establish judicial effectiveness or causal effects of the reforms.

“Almost all of the growth arrived in that first year…” — Yuxia Zhang and John Zhuang Liu, “The Rise of the Chinese Judiciary and Its Limits: Administrative Litigation in the Reform Period,” The China Quarterly 262 (June 2025): 330–351, https:// doi.org/10.1017/S0305741025000311. Using 1.64 million administrative judicial documents from 2014–2020, the authors estimate a 79 percent rise in accepted cases, an increase in plaintiffs’ win rate from 33.2 to 42.2 percent, and chief-government-official attendance in more than half of trials. They also identify fundamental limits: little judicial engagement with citizens’ political rights and reluctance to conduct substantive review in political cases. The study supports a bounded claim of strengthened administrative litigation, not a claim of general judicial independence or regime democratization.

“China’s Constitution, meanwhile, states that leadership by the Communist Party of China is the defining feature…” — Constitution of the People’s Republic of China, current official English text, art. 1, National People’s Congress, https://www.npc.gov.cn/englishnpc/ constitution2019/201911/t20191120_384295.html. Article 1 states that leadership by the Communist Party of China is the defining feature of socialism with Chinese characteristics. It is cited to establish that the administrative-litigation reforms occurred without replacement of the national party architecture, not to characterize every feature of Chinese political practice.

“This is where machine intelligence can change design without becoming the designer…” — Mikel Artetxe and Holger Schwenk, “Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond,” Transactions of the Association for Computational Linguistics 7 (2019): 597–610, https://doi.org/10.1162/ tacl_a_00288; Patrick Lewis et al., “Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks,” Advances in Neural Information Processing Systems 33 (2020): 9459–9474; Runtao Ren, Jian Ma, and Zhimin Zheng, “Large Language Model for Interpreting Research Policy Using Adaptive Two-Stage Retrieval Augmented Fine-Tuning Method,” Expert Systems with Applications 278 (2025): 127330, https:// doi.org/10.1016/j.eswa.2025.127330. These works support multilingual semantic retrieval, retrieval over external records, and source-grounded policy interpretation. They do not establish that AI can determine institutional equivalence, generate a politically legitimate reform, identify every local condition, or decide which candidate design a society should adopt. The chapter’s design-space application is therefore a conditional extension of demonstrated retrieval and generation capabilities, with human and institutional judgment retained.

Chapter 11 — Stress Before Catastrophe

“For the 2026 supervisory stress test…” — Board of Governors of the Federal Reserve System, “2026 Stress Test Scenarios,” February 4, 2026, https://www.federalreserve.gov/publications/2026-stress-testscenarios.htm; Board of Governors of the Federal Reserve System, “Federal Reserve Board Finalizes Hypothetical Scenarios for Its Annual Stress Test,” February 4, 2026, https://www.federalreserve.gov/ newsevents/pressreleases/bcreg20260204a.htm. The Federal Reserve describes the severely adverse scenario as hypothetical rather than a forecast and specifies, among other variables, a 10 percent peak unemployment rate, an approximately 58 percent equity-price decline, a 30 percent house-price decline, and a 39 percent commercial-real-estateprice decline. The scenario is used here as an example of structured resilience testing, not as an analogy that makes political institutions financially measurable.

“In 2009, federal supervisors simultaneously examined nineteen…” — Board of Governors of the Federal Reserve System, “The Supervisory Capital Assessment Program: Design and Implementation,” April 24, 2009; Board of Governors of the Federal Reserve System, “Federal Reserve, OCC, and FDIC Release Results of the Supervisory Capital Assessment Program,” May 7, 2009, https://www.federalreserve.gov/ newsevents/pressreleases/bcreg20090507a.htm; Board of Governors of the Federal Reserve System, 96th Annual Report, 2009, 76, 100. The exercise assessed nineteen of the largest U.S. bank holding companies under a more adverse economic environment than generally anticipated and identified ten as needing additional capital buffers. The episode establishes the institutional use of forward-looking stress assessment after the financial crisis; it does not establish that stress testing prevented later financial instability.

“Exploratory-modeling research developed this logic…” — Steven C. Bankes, “Exploratory Modeling for Policy Analysis,” Operations Research 41, no. 3 (1993): 435–449, https://doi.org/10.1287/ opre.41.3.435; Jan H. Kwakkel and Erik Pruyt, “Exploratory Modeling and Analysis: An Approach for Model-Based Foresight under Deep Uncertainty,” Technological Forecasting and Social Change 80, no. 3 (2013): 419–431, https://doi.org/10.1016/j.techfore.2012.10.005. These works support exploratory use of multiple models, assumptions, and futures under deep uncertainty rather than reliance on one predictive run. They do not establish that political breakdown can be forecast accurately or that computational exploration supplies legitimate political choice.

“Its annual Tabletop the Vote exercise…” — U.S. Cybersecurity and Infrastructure Security Agency, “Election Security Training,” https:// www.cisa.gov/topics/election-security/election-security-training, accessed August 22, 2026; U.S. Cybersecurity and Infrastructure Security Agency, “Elections Cyber Tabletop in a Box,” https://www.cisa.gov/ stopransomware/elections-cyber-tabletop-box, accessed August 22, 2026. CISA describes Tabletop the Vote as an annual national election-security exercise bringing election officials, federal partners, and vendors together to identify best practices and areas for improvement, and describes customizable tabletop packages built from exercise objectives, scenarios, and discussion questions. The sources establish a real process-resilience exercise; they do not show that every exercise discovers every vulnerability or that an exercise result predicts an election incident.

“NIST’s guidance on red-teaming…” — National Institute of Standards and Technology, Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile, NIST AI 600-1 (Gaithersburg, MD: NIST, July 2024), especially MEASURE 1.3 and suggested actions MS-1.1-008 and MS-1.3-001–003, https://doi.org/10.6028/ NIST.AI.600-1. NIST recommends structured human-feedback exercises and red-teaming with relevant domain experts, users, independent assessors, and affected communities as appropriate to context, and explicitly recognizes risks that cannot be measured quantitatively. The chapter adapts this adversarial-testing discipline to political institutional design; NIST does not claim that generative AI can predict political collapse or determine which institutional reform should be adopted.

“If small changes in the corpus, prompt, model, or assumed baseline…” — Andrea Saltelli et al., “Five Ways to Ensure That Models Serve Society: A Manifesto,” Nature 582, no. 7813 (2020): 482–484, https:// doi.org/10.1038/d41586-020-01812-9. The authors argue for transparent assumptions, sensitivity to uncertainty, plural framing, and avoidance of false precision in model-supported public decisions. The article supports the chapter’s insistence that model fragility and alternative assumptions remain visible; it does not prescribe a specific AI architecture for political stress sensing.

Chapter 12 — The Gardener Must Remain Human

“In 2020, the District Court of The Hague examined a Dutch system called SyRI…” — District Court of The Hague, judgment of February 5, 2020, ECLI:NL:RBDHA:2020:865 (English translation ECLI:NL:RBDHA:2020:1878); Rechtspraak, “SyRI Legislation in Breach of European Convention on Human Rights,” February 13, 2020. The court held that the legislation failed the fair balance required by Article 8 of the European Convention on Human Rights and emphasized insufficient transparency and verifiability. The ruling concerns the Dutch SyRI statutory scheme before the court; it does not establish that every form of public-sector data matching or AI-assisted fraud detection is unlawful or incompatible with Article 8.

“On January 21, 2024, a robocall using an artificial voice resembling President Joe Biden…” — Federal Communications Commission, Forfeiture Order, Steve Kramer, FCC 24-104 (adopted September 26, 2024; released September 30, 2024), paras. 1–3, 6–12. The Commission imposed a $6 million forfeiture and documented 9,581 initiated calls using misleading caller-identification information. The order does not establish persuasive effect.

“In a preregistered study involving nine hundred participants…” — Francesco Salvi et al., “On the Conversational Persuasiveness of GPT-4,” Nature Human Behaviour (2025): 1645–1653, https:// doi.org/10.1038/s41562-025-02194-6. In the controlled study (N = 900), personalized GPT-4 was more persuasive than human opponents in roughly two-thirds of the comparisons in which one side was more persuasive than the other. The study does not estimate real-world electoral conversion.

“European rules for online political advertising…” — Regulation (EU) 2024/900 of 13 March 2024 on the transparency and targeting of political advertising, arts. 18–19, applicable from October 10, 2025; European Data Protection Board, Guidelines 8/2020 on the Targeting of Social Media Users, version 2.0 (April 13, 2021), especially paras. 10–18, 43–48, 80–88, and 118–121. This is a jurisdiction-specific framework, not a universal design for political persuasion.

Conclusion — You Do Not Need Our Four Hundred Years

The Conclusion introduces no new factual evidence about Earth. The cases it revisits are documented in the notes to the chapters that first examine them.