In 1968, your odds of dying on a commercial flight were roughly 1 in 350,000. By 2022, that number was 1 in 13.7 million. Aviation didn't get there by building better planes. It got there by building a system that treats every failure as information - and that makes it structurally impossible for a crash to be documented and then forgotten.

This episode traces that system from Tenerife in 1977 - where a flight engineer asked the right question and was overruled in four words by the most senior pilot at KLM - through Portland in 1978, where a first officer watched the fuel run out because the culture taught him to hint rather than assert. It follows Crew Resource Management from its origin to its proof case: United Flight 232 in 1989, where Captain Al Haynes invited a passenger into the cockpit and asked, "Why would I know more than the other three?" 185 people survived an unsurvivable situation.

Then it follows the NTSB - the independent investigation architecture that turned crash data into mandatory public lessons for fifty years, producing the best safety record of any transportation industry in history.

And then it follows the Boeing 737 MAX. 346 people dead. A flight control system erased from the pilot manual before any pilot could object to it. The safety architecture selectively dismantled for a commercial incentive.

The distance between Tenerife and Sioux City is what institutional design can accomplish. The distance between the NTSB's fifty-year record and the 737 MAX is what happens when that design is selectively abandoned.

Show Notes:

The 1 in 350,000 figure (1968–1977) and 1 in 13.7 million figure (2018-2022) are from the MIT/Barnett study and represent global commercial aviation passenger boarding fatality risk. These are the most rigorous publicly available estimates.

The 96% self-certification figure is sourced from the PMC engineering ethics paper citing Kitroeff et al. 2019 (New York Times reporting). Boeing contests aspects of how this figure was framed publicly; the DOT Inspector General report documents the delegation structure without using this precise percentage.

Sources Referenced:

• MIT News: "Study: Flying Keeps Getting Safer" (2024) - source for the 1-in-13.7-million figure and the decade-doubling safety improvement trend; MIT professor Arnold Barnett
• Cirium: "Flying Safer Than Ever: The Evolution of Aviation Safety" - source for the one-seventeenth figure and the historical passenger fatality rate context
• Incident Prevention: "Lessons Learned from the Tenerife Airport Disaster" - source for van Zanten as head of KLM safety and the first officer's objection to the fuel load
• United Airlines Flight 232 / Haynes quote - from Haynes's widely published public statements and congressional testimony; verify exact wording against primary source
• NTSB: History of the National Transportation Safety Board - for founding, independence, and 15,500 recommendations figures
• Congress.gov / CRS Report R44587: "The National Transportation Safety Board: Background and Possible Issues" - for the 82% implementation rate and the 90-day response requirement
• PMC: "The Boeing 737 MAX: Lessons for Engineering Ethics" (PMC7351545) - for the 96% self-certification figure, the MCAS failure analysis, and the organizational context
• DOT Inspector General Report AV2021020: "FAA's Certification and Oversight of the 737 MAX" - primary source for the internal-document-only designation and the FAA's incomplete understanding of MCAS at certification
• Seattle Times: "The Inside Story of MCAS" (Dominic Gates, June 2019) - source for the pressure to avoid simulator training and the removal of MCAS from the flight manual

www.tenthman.ai

On December 14th, 1799, George Washington's physicians did what medicine had done for two thousand years. They bled him. Three times. He was dead by evening. They were not reckless. They were following the most credentialed, most consensus-supported medical protocol in the Western world.

This episode is about what it took to dismantle that consensus, and what medicine had to build in its place. It covers James Lind's 1747 controlled trial aboard HMS Salisbury, the physician who proved that doctors were the disease vector and was driven to an asylum for it, the 1784 French royal commission that discovered the placebo effect while investigating a magnetic fraud, and Austin Bradford Hill's sealed envelopes - the design that finally removed personal responsibility from the clinician.

Episode 3 of The Tenth Man, a podcast about the specific mechanisms different domains have built to make dissent structurally impossible to ignore.

Show Notes:

The George Washington death account is historically well-documented; the three-to-four-liter blood removal figure is consistent across historical accounts but is a reconstruction based on recorded "bleedings" of specific volumes, not a single contemporaneous measurement. State it as the scholarly consensus estimate.

The Semmelweis mortality figures (10% and 30% peaks in the physicians' ward, under 2% in the midwives' ward) are from the historical record documented by Semmelweis himself and subsequent historians; the specific monthly variation has some range across sources. The orders of magnitude are not disputed.

The "removed personal responsibility from the clinician" phrase is from the 1948 BMJ editorial accompanying the streptomycin trial report. It is presented in the episode in its original context - as a description of the design's intention, not a criticism of it.

Sources Referenced:

• The Blood Project: "Letting Blood: The Rise, Reign, and Fall of Medicine's Oldest Therapy" (2025) - narrative history of bloodletting with strong sourcing on the cultural and professional dimensions of its persistence
• Britannica: "Bloodletting" - for George Washington and the Marie-Antoinette references; the 1942 Osler textbook reference
• British Columbia Medical Journal: "The History of Bloodletting" - for Galen context and the quote on the social, economic and intellectual persistence of the practice
• James Lind Library, "James Lind and Scurvy: 1747 to 1795" - peer-reviewed medical history; primary source for Lind's trial design, the Salisbury context, and the "controlled empiricism" characterization of Lind's methodology
• PMC: Lind and Scurvy (1276007) - for the 80-of-350 figure and the description of Lind's result as "convincing, particularly because the differences shown were so dramatic"
• Amusing Planet: "James Lind and the First Clinical Trial" (2025) - for da Gama and Magellan casualty figures and the forty-year adoption gap
• Wellcome Collection: "The Father of Handwashing" - for the street-birth detail and the broader narrative of Semmelweis's career and rejection
• Linda Hall Library: Ignaz Semmelweis - for the complexity flag: publication delay, monograph length and tone, inflexibility with contradictory cases
• The Lancet: "Placebo controls, exorcisms, and the devil" (2009) - for the Franklin/Lavoisier commission framing and the historical significance of the 1784 experiments
• PubMed: "The origins of modern clinical research" (12461388) - for the 1784 commission as the first published use of intentional subject ignorance and sham intervention
• PMC: "The MRC Randomized Trial of Streptomycin and Its Legacy" (PMC1592068) - for the clinical context, streptomycin shortage, and trial design details

www.tenthman.ai

On the night of January 27th, 1986, a group of engineers in Utah tried to stop the Challenger launch. They had the data. They had a documented history of a known design flaw. They had a teleconference with NASA that lasted hours.

They did not stop the launch.

This episode is about a harder problem than suppressed dissent: what happens when the dissent is present, documented, voiced, on the record, and institutionally weightless. It covers the O-ring warnings that went back a full year before the disaster, the night-before teleconference and what happened when Morton Thiokol engineers were removed from the room, Richard Feynman's minority report on the Rogers Commission, and the sociologist who spent years in the National Archives and came back with a finding that changed the story.

Episode 2 of The Tenth Man, a podcast about the specific mechanisms different domains have built to make dissent structurally impossible to ignore.

Show Notes:

The "take off your engineering hat" exchange is documented in Rogers Commission testimony by both Kilminster and Boisjoly and is cited across multiple independent sources. It can be stated as fact. The Feynman risk estimate discrepancy (1-in-100,000 vs. engineer estimates of 1-in-50 to 1-in-100) is documented in Appendix F of the Rogers Commission Report and is primary source material.

Vaughan's normalization of deviance thesis is academic and interpretive - it is the most persuasive scholarly account of the organizational failure, but it represents a revision of the Rogers Commission's more blame-focused framing. The episode presents it as such, not as settled fact.

Sources Referenced:

• Rogers Commission Report (June 6, 1986) - Presidential Commission on the Space Shuttle Challenger Accident. Full text publicly available via NASA History Division (nasa.gov/history/rogersrep). Primary source for teleconference testimony, commission findings, and nine recommendations.
• Richard Feynman, Appendix F: "Personal observations on the reliability of the Shuttle" - included in the Rogers Commission Report. Primary source for the 1-in-100,000 vs. engineer estimates discrepancy and the "reality must take precedence" conclusion.
• Roger Boisjoly testimony, Rogers Commission (1986) - primary source for the teleconference account, the "unethical decision-making forum" characterization, and the caucus sequence. Also available via NASA History.
• Roger Boisjoly, "O-Ring Erosion/Potential Failure Criticality" memo, July 31, 1985 - available via National Archives and cited across multiple secondary sources. Primary documentation of the pre-disaster written warning.
• Diane Vaughan, The Challenger Launch Decision: Risky Technology, Culture, and Deviance at NASA (University of Chicago Press, 1996; enlarged edition with Columbia preface). Primary academic source for normalization of deviance theory and the revisionist organizational account. Vaughan is Professor of Sociology at Columbia University.
• Columbia Magazine, Winter 2025–26: "This Is Not Normal" - interview with Diane Vaughan on the 40th anniversary of the Challenger disaster. Source for Vaughan quotes including the NASA luncheon account and the "completely different" archival finding.
• Columbia Accident Investigation Board (CAIB) Report, 2003 - primary source for the "causes of the institutional failure responsible for Challenger have not been fixed" finding. Publicly available via NASA.
• Online Ethics Center, Texas A&M: "The Space Shuttle Challenger Disaster" - engineering ethics case study. Source for the Mason/Lund "management hat" exchange and the teleconference timeline.

www.tenthman.ai