In the middle of an actual, like, actual workplace coup with secret meetings, purges, puppetmasters, frame-ups, leak investigations, and (rumored) slack DM surveillance, I was listening to the bombshell podcast and heard about this book about coup-proofing, and how it ruins teams. I picked up a copy, and this year I read it.

The book (Caitlin Talmadge / The Dictator’s Army) is about how coup fears hamstrung management in two wars: Saddam Hussein’s 8-year failed invasion of Iran, and South Vietnam’s final defense against the North. This article is about private sector leadership lessons I took from the book.

Writing about private sector leadership on the eve of a plausible military coup seems like peeing into the ocean, but it’s this or doomscrolling and holding in your pee can kill you.

Most management books are aspirational and map out success. This one is about repeated failure, despite great funding, first-mover advantage and better equipment, through repeated mistakes, and in the public eye. Nothing that I would know anything about cough.

Startups sub 30 people have a hard time accessing experienced hires who combine domain skill and the ability to work on a team. There are real reasons for this, and some boil down to something like Saddam’s coup fears. But if our society plans to continue airlifting cash into small companies, onboarding good middle managers earlier can make more of these bets pay off.

If you prefer a soundtrack, play this masterclass on management (from GI Joe’s Cobra Commander) in the background. It sometimes gets me through the week.

  1. Despots meddle
  2. Managing up
  3. Demotions, temporary and permanent
  4. Combined arms
  5. Some people out there don’t have maps
  6. The butcher’s bill
  7. Professional silence

Despots meddle

Despotism is a legitimate mode of government in dealing with barbarians, provided the end be their improvement, and the means justified by actually effecting that end.

Liberty, as a principle, has no application to any state of things anterior to the time when mankind have become capable of being improved by free and equal discussion. – JS Mill

By inference, if you’re not hiring barbarians and / or they don’t need improvement, or your team is capable of discussion, consider a lighter-weight leadership style.

‘Despot’ also means bronze age feudal landowners or the heir apparent of Byzantium, but we haven’t had those for a while. Adaptive reuse of the term in revolutionary France gave the term its modern meaning: an absolute ruler who governs without restraint by the law. See also ‘frondeur’.

When growing teams take on more complex goals but don’t update their management, they encounter pathologies of delegation – senior leadership who isn’t sure how to solve problems using middle managers, who give them the wrong length leash. This is the 10 person to 30 person growth spurt. Good companies outgrow it this if they survive.

Here’s what that looked like in Iraq:

decision-making at the senior levels of the Iraqi military revolved more around the fear of what Saddam might do than what the Iranians might do.1

by the middle of the war, Saddam was making a serious effort to inform himself about military matters in order to understand what was happening. However, Saddam was the only decision maker,2

Technocratic middle managers were spending most of their energy undoing their leader’s bad decisions, rather than fighting and winning. There’s a spectrum between ‘own worst enemy’ and ‘annoying bandwidth bottleneck’, but new managers generally are on it somewhere.

Over-centralization of authority sends a message that people shouldn’t act on their own, which paralyzes your org. Note that I’m not arguing against hierarchical authority or a clear chain of command, but rather against authority all concentrating at the top, making it impossible for middle managers to do their jobs. In Vietnam:

few officers who had come up in the Diem system were willing to do anything “in the absence of detailed orders.”3

Centralization also hamstrings the flexibility of a large team to adapt to new information. Top leaders, often lacking domain skill or up-to-date information, are best at ‘do nothing without asking me’ orders that trap middle managers in bad positions. In one case:

Because Saddam had decreed that units were not allowed to cede any ground, the Iraqi defense was poorly coordinated and ineffective, eventually inducing panic.4 … Iraqi officers in the field lacked the authority to redeploy forward troops or to call on reserves.5

And because the dictators aren’t an expert in any part of the organization, lack the attention to focus on everything at once, and aren’t accountable to anyone, they have zero memory. In one case in Vietnam, President Thieu’s personal intervention not only disrupted the existing strategy, but he had forgotten by the next day that he’d given the order.

sudden change in orders from an uninformed Thieu was a disaster, resulting in neither a robust defense of the city nor a clean withdrawal.6 … Amazingly … Hue’s abandonment apparently came as a surprise to President Thieu7

Unaccountable, illogical plans from inexpert leaders create hills to die on, in some cases literally. They also come through unusual channels (because hands-on leaders prioritize fast action). Unusual channels = coordination and preparation impossible for middle managers. In one case in South Vietnam, a general:

personally interceded at all levels, sometimes going so far as to personally issue orders by radio to individual brigade commanders without notifying the division operations center. Both Giai and his division advisers often learned of new orders only as they were being carried out.8

By comparison, the North Vietnamese had good delegation and sane command structures in place. From an internal document, they valued:

“unanimity of thought from the top to the bottom regarding the opportunity we faced and the tactics to be employed” but significant room for commanders to execute their plans9

After the American revolution, George III was having a conversation with his American portraitist Benjamin West; he heard Washington was going to resign his commission and ‘return to his farm’. The King was like ‘ha pull the other one’, recognizing that giving up power is very hard, and said something like ‘if he does it I take back everything I said about the guy.’ The point is that accepting the limits to authority is hard, but also critical for good leaders.

Managing up

One of the lessons of 2020 is probably that sometimes the only person in the room who can stop a bad outcome is a middle manager, and often they can only do it by managing up, not through direct action.

Middle managers are frequently in situations where their domain skills are better than senior leadership – either because they’re technocrats with domain skills, or because they’re dealing with the problem day-to-day and have better information, or both.

We still get shouted down a lot, resulting in waste: of money, of time, of resources, of opportunities, of morale. In a war it can be lives.

We speak out a few times but eventually converge to inaction as your wtfs trail off. We give up because we’re tired, but that’s not all of it: We stop arguing because when a leader experiences argument as dissent, their reactions are unpredictable and poison the team.

The way around this way of thinking is to (1) make very sure that you’re very right when you pick a fight with your boss, (2) fight for better delegation, not just a correct decision this one time. And (3) remember that if you succeed, things get permanently better; these are hard fights but companies need them to grow.

Middle managers are the only people who have the authority and connections to fight these fights. That’s why they leak, that’s why they manage up. The organization will fail if we don’t. Employees depend on companies to put food on the table; you do have a moral obligation to help your employer not fail (if you can) (barring circumstances like if you’ve looked at these caps recently).

I saw a JAMA article about ‘professional silence’ by doctors. Silence is not golden when an important argument is being suppressed.

Senior technocrats have a surprising amount of power when they align their story and band together. For Saddam, years into the war:

Iraqi generals blamed Saddam’s military organizational practices for both sets of losses, raising a serious danger of overthrow from within10

Because his leadership style was in conflict with their technical skills. He punished losses in ways that deterred ground commanders from requesting backup, creating a losing cycle:

Saddam put great pressure on Iraqi commanders on the ground to avoid losses, which led them not to report failures. Withholding losses from reports and thus not receiving reinforcements or other support left commanders in impossible combat conditions. However, this was better than reporting their failures and suffering execution.11

The command-and-control system was incapable of transmitting the true tactical situation. Senior Iraqi officers later noted that they often got more timely information from the media than they did from their own commanders at the front.12

In Vietnam, General Phu, a technocrat middle manager who had been brought in to fix the South Vietnamese army’s problems, tried to use his authority to salvage a bad campaign and was talked down.

Phu, who previously had enjoyed an excellent reputation as commander of the 1st Division, expressed grave reservations about the operation to those around him, complaining that he had done it only because of Thieu’s order13

President Thieu’s realtime intervention, leading to a bad loss, demoralized him and he was never able to lead again:

Thieu’s absurd order to withdraw immediately from the Central Highlands essentially caused another corps commander, General Phu, to give up.14

When the Brett Crozier stuff was breaking (Navy captain who brought his ship to port to deal with coronavirus), I saw a tweet from a navy dad about Crozier as a middle manager:

My USNA grad daughter, former SWO & now a primary care MD just said to me “Crozier knew exactly what he was doing. You don’t get that command w/o knowing the chain of command. He was desperate & knew what he had to do to save his crew.”15

Good systems have disobedience built in. In service animals, it’s called intelligent disobedience and it keeps people from confusing a staircase with a cliff. That article also talks about ‘Blink think choice voice’, a safety mnemonic to guide children through disobeying a creepy adult. Young workplaces are in my experience bad at enabling disobedience, but also sorely need it. Startups can be paradoxically more authoritarian than well-run large companies.

Channels for disobedience are useless without high quality people who have both domain skills and enough court sense to act selflessly some of the time. More on high quality people in the next section.

Demotions, temporary and permanent

Serious organizations are serious about promotion, and will seek to pack the org chart with competent people who can deal with authority and responsibility. At small companies experiencing management scuffles, promotion and hiring can be more about filling power vacuums and isolating enemies. I once watched a manager hire his worst frat buddy to run a division in order to de-power his cofounder, and for the next six months the cofounder had to do the frat buddy’s homework.

Autocrats promote badly because when your goals are selfish or illogical, the only form of ‘competence’ you recognize is loyalty. Here’s how the Iraqi military approached that in the 80s:

Before the executions were stopped, anyone who carried out an execution was promoted by Saddam.16

(And Saddam himself often punished major failure with execution).

He spared one of his favorites, Major General Tala al-Duri, by removing him from command of the 9th Armored Division just three days before that division collapsed during Operation Ramadan in 1982; Saddam then ordered al-Duri’s successor to be shot as a result of the division’s defeat.17

In reference to al-Duri’s effort to return to active duty in the late 1990s, Hamdani said “[al-Duri] had caused the death of a third of the Iraqi Army and now wanted to return to finish the job.”18

The Iranians had their own version of this: the invasion of a swampy peninsula called Al-Faw was their only successful offensive campaign, and could have been the turning point of the war, but they didn’t know they’d won. Both Iran & Iraq thought they lost the battle. Iran, whose revolutionary religious leadership had been conducting a years-long purge of their legacy military, used the apparent loss as an opportunity to clean house of their few remaining professionals.

Part of the problem: some important middle managers in the Al-Faw operation died in a plane crash on their way back to Tehran19. If aviation had been safer back then, the course of the war and the history of the region might taken a different path.

I don’t mean to canonize war or victory; but merely in terms of lives spent, ending the war early would have been a win. Just improving leadership would have been a win – the Iranians’ main tactic before and after Al-Faw was to overwhelm Iraqi positions with their untrained, martyrdom-motivated revolutionary infantry (more on this in the ‘ideology’ section below).

Saddam was an early practitioner of my favorite hiring trick, the fine art of ‘culture fit’. His “belief that Bedouin tribal courage and loyalty were the most important attributes of successful military leadership”20 led to situations like this one:

Saddam willingly tolerated the opinions and actions of his son-in-law Hussein Kamel, whom one of the generals interviewed described as a semiliterate.21

I’ve been a lackey. In my first two jobs I had direct routes to senior management and was responsible for fast-tracking pet projects. Which kind of made me a pet. This is fun if you’re young and don’t know what’s going on, but in retrospect it was bad for the companies I was at (senior people quit because my unearned authority made their jobs harder).

Here’s the inverse of this, from a more professional leader, Leslie Groves of the manhattan project:

It was important that the officers whom we selected command the respect of persons already in the project … the scientists were most critical of anyone whose mental alertness did not equal or excel theirs. Slowness of comprehension or inability to keep all the pertinent facts in mind, once they were explained, was fatal22

The combination of bad people and bad practices in the org meant that the Iraqi army couldn’t make use of serious resource advantages:

Despite having a huge initial advantage in numbers (as much as 6:1 in some places) and a sustained edge in technology, the Iraqis displayed serious problems turning these advantages into fighting power. In addition to the inaccurate and virtually irrelevant air strikes conducted at the outset, Iraqi artillery units repeatedly made simple fusing errors that limited their weapons’ impact,23

Don’t list your advantages on paper and assume you can hire just anyone to use them. Talmadge cites the same weaknesses in the South Vietnamese army: training focused more on political indoctrination than their weapons meant that units couldn’t ‘combine arms’ (i.e. coordinate, more on that below). The south was generally worse than the north at basic weapon competencies.

Big companies have this same training gap – you spend your whole first week on ‘training’ that was put together by HR, is mostly political indoctrination, and had nothing to do with your real job. All these armies had ‘HR’ functions, in the form of political operatives embedded in combat units with unclear authority and chain of command. When senior leaders don’t take responsibility for education and promotion, ‘glue’ silos like HR are tempted to seize power by colonizing the vacuum.

I heard an interview with James Stavridis, a senior leader in the US Navy, who observed that promotion structures in his department have “pretty good filters along the way”24 and get you “literally 1 in 1000 from the beginning tranche of people who come in”25.

The host was asking about the problem of “esteemed institutions that still allow bad leaders to emerge”26. I’m not sure Stavridis had a satisfying answer, but my point is that the Navy thinks carefully about recruitment and advancement as a tool for building cultures, and that organizations are built of people, and big teams get the behavior that middle managers allow.

The deep state isn’t always a bad thing. Startups could afford a little deep state action.

Combined arms

Here’s how fancy things got in the coup I witnessed: One time an internal leader was purged, and I had an informal exit convo with them in a conference room. It ran long, and was pretty honest i.e. dark with lots of historical hypotheticals, and afterwards a senior leader brought me into the same room and was like, hey: A different senior leader was in the conference room next door, and overheard your convo, because the walls are pretty thin, and is there anything I need to know.

I was flatfooted (roles reversed, I would have handled this by doing nothing) and said the first thing that popped into my head, which was: That was an exit interview. It was frank because it’s important that we’re realistic about our weaknesses.

Imagine how shocked I was to find out the North Vietnamese had a policy of “candid discussion of weaknesses”27 to enable honest tactical communication in their politicized communist regime. The Iraqis took 8 years to get to that point, for coup reasons and to avoid provoking Saddam, and it cost them:

Hussein discouraged inter-service cooperation so that the Air Force and the Army could not coordinate a coup.28

By the time information from ground commanders on the battlefield filtered through the various political circles to the Air Force, it would be too late for them to perform in an effective manner.29

[a ground commander] requested air support after he came under attack, but … his request had to be routed through Baghdad due to restrictions on army–air force communication,30

Classic management bandwidth bottleneck. Classic failure to delegate and decentralize.

The information barriers weren’t restricted to coordinated military action. Even upwards-only information updates were perceived as dangerous.

when front-line commanders or staff officers managed to get reasonable accounts of what was happening up the chain of command to Saddam, the military courtiers were outraged.31

The obstacles to horizontal vs vertical communication may seem like different problems. Based on my up-close coup, I think they’re not – they’re both failures of delegation, or failures of imagination around how middle managers can work together to achieve the company’s strategy. Which often means the strategy is bad or senior leaders don’t understand it. In my experience, when leadership believes that the rank & file aren’t executing on their strategy, the fish is rotting from the head down. (A fun chinese phrase I learned this year).

As a result:

Virtually all attempts to combine arms – for example, by relying on artillery from across the Shatt to support infantry in Karbala-4, or by linking up the two major prongs of the assault on Basra in Karbala-5 – failed.32

Coordination requires information to flow freely. It’s telling that these communication failures lead to failures of coordination. Bad coordination is the calling card of bad teams.

It’s rare to see pure anger in the office now I think (in my particular filter bubble at least). I saw it twice at my short stint in a big bank: once for trying to talk to someone in a different silo of the bank, and once for honestly answering a numeric question from my boss’s boss’s boss when my boss’s boss was out of town.

This kind of siege mentality erodes the value provided by working in teams. When you start to control information it becomes your whole job.

Some people out there don’t have maps

Obedience has a weird place in modern culture. I think our main metaphors for management are still Taylorism (the guy who brought math to the assembly line) and the popular portrayal of military orders as ‘do or die’. And possibly the portrayal of ‘respect for authority’ in The West Wing, the TV show that taught my generation how to lick boots (‘good evening mr president’).

If these were ever right, they aren’t now, but a million mini despots still think obedience matters at work.

One way to look at autocrats is as leaders who don’t know their main role is communication. Or that if they’re the ‘decider’, that’s a part time job and it comes after hearing all the options to decide between. At best you’re a tie breaker.

Intolerance of argument = intolerance for information, and for plans, and is deadly. It also means a lack of understanding of what arguments looks or sounds like. Corollary: if the senior person in a room can’t run a meeting, it doesn’t matter who else can.

How bad was the Iraqi invasion prep in 1980:

Saddam held a number of differing, and in some cases conflicting, views throughout the immediate pre-war period as to what he hoped to achieve33

how military force was going to accomplish such a victory without clear operational or tactical goals was a mystery that Saddam and those surrounding him were incapable of examining.34

participants were unable to agree on any coherent strategic aims.35

Many stakeholders didn’t know what was coming. During my personal coup, possibly because of leak fears, a senior manager ‘launched’ (announced) a product change without telling anyone else on the team. The morale hit was palpable and instantaneous, despite this change being one that insiders and customers had fought for.

IIRC the change never saw the light of day, and was panned by tech press as ex employees or leakers complained about the surprise announcement. Here’s how that went in 1980 Iraq:

Saddam’s representatives made clear that war with Iran could occur in the immediate future. Nevertheless, most senior officers took the warning as indicating that a conflict might occur within a two-year period, during which time they could prepare their forces. At the time, however, none of the interviewed generals believed war would occur within two months.36 … The navy’s senior officers were not informed of the decision to invade until September 20, 2 days before the invasion.37

“Our troops were just lined up on the border and told to drive into Iran. They had an objective, but no idea how to get there or what they were doing, or how their mission fit the plan, or who would be supporting them.”38

The icing on the cake was that some people out there don’t have maps (and as I re-watch this I notice that Upton even talks about ‘help the iraq’):

Adding to the difficulties that were to plague the Iraqis in their advance was the fact that they possessed no up-to-date maps and, in some cases, no maps at all of the territory through which their forces would advance. … 9 Iraqi divisions, as General Makki put it, drove down the 9 separate roads across the Iran–Iraq border. … a number of formations got lost39

If you clicked the cobra commander video I linked in the intro, you saw him berating his technocrat middle manager Destro for offering advice. ‘Surely you will not concede failure so soon / I will concede your failure’. This is a failed relationship and the emotional tone is not far off from the four horsemen of divorce (criticism, contempt, defensiveness, stonewalling).

Saddam was the same way – he had no tolerance for professionals until, years into the war, all of his other advisers had failed at great cost. He had no desire to convince or listen and valued the strength of his personalist system over victory.

Observations from frustrated middle managers who had to work with him:

He asked if it were possible to make luck and coincidence a ‘principle of war.’ … I interrupted to say that it was impossible … I noticed Khairallah sink in his seat and Hussein Kamel adjust his position … their body language said to me ‘Be careful’40

At a political level, he was an excellent tactical player; however, at the strategic level, 99 percent of his concepts were wrong.41

This is the frustration of an expert with good knowledge of the situation who is forced to explain their craft to a child, and then execute the bananas plan the child comes up with, and promise unrealistic results in exchange for the rubber stamp.

In real strategy planning, everybody in the room has to understand that this is a risk reward calculation with contingencies and limited resources. If the purpose of the meeting is to get subordinates to commit to a 100% successful, short-deadline plan, you’ll never come up with a viable strategy. Strategies either embrace uncertainty or they’re at the mercy of it. ‘Eat not from the tree of knowledge lest ye die’ is a bad rule to give your experts.

The alternative to good strategy is workplaces that only allow ‘heroic’ nights & weekends action, because all work time converges to fluffing activities; where the organized framework for decision-making is immune to argument, resistant to the work of identifying problems, and can’t abandon whims for plans even in the face of repeated failures.

The butcher’s bill

My understanding of ideology is totally different after reading this book. Here’s a zany observation:

When the Islamic Revolution flared up in Iran … Iraq had a secular regime … Khomeini thought a second victory could be achieved in Iraq.42

The [Iraqi] strategic plan for war was based on the assumption that Iraq forces could achieve victory in four to six weeks … Iranian forces would be forced to concentrate and engage in battle. The counter-revolutionary forces would then rise up and seize control of Tehran and bust Khomeini’s regime43

Note: both Iran and Iraq were betting that their powerful ideology would inspire the partisans to rise up in the cities of their enemy. They forgot the first rule of scams, which is never to believe your own con. Or in Talmadge’s language, they confused internal and external threats and applied the wrong tools.

Startups use ideological recruitment too. ‘I really think we can change the world’ and ‘you’re being compensated in equity that will evaporate if you deign to leave us’ are statements that frequently go hand in hand.

IMO these claims both translate to ‘I like being in charge’, and are a personality test – if you don’t think it’s ludicrous to be paid less to take more risk, you probably won’t argue against other illogical plans and alert the adults on our board that we need one of those toddler walking ropes.

The Iranians’ main tactic was a human wave, costly in lives, that overwhelmed Iraqi minefields and positions, as much by absorbing damage as by firing their weapons. They generally failed to follow up with other kinds of forces once their infantry had cleared the way. It was only effective in that it allowed the war to continue going on, and it relied on ideological recruiting:

Of crucial significance here was the Shi’i emphasis on martyrdom as a righteous act – a theme emphasized repeatedly to Iranian soldiers in their training and indoctrination.44 … Only after the revolution had taken hold in Iran were the human wave attacks possible.45

It also couldn’t last. Bad arguments can’t stand up to continued public failure:

Eventually, the butcher’s bill broke the religious mania Khomeini had created,46

massive casualties and strategic futility of the human wave attacks had become obvious to the Iranian populace.47

I don’t mean to make light of real sacrifices in war by comparing it to soft-hands private sector work. But bad management is bad management. It repeats its mistakes and it starves out competence in your organization:

an army cannot propagandize its way to competence in basic tactics and complex operations48

[Iran’s military] never brought an end to the feuding and lack of coordination between the Pasdaran and regular force, and it continued to reward loyalty and belief, rather than professionalism49

I once asked my favorite recruiter how to spin a bad placement, and his advice was something like: “it’s not just about spin. You solved problems there by being a martyr. If you don’t find a more sustainable way to contribute, only other martyr shops will want you.”

Absent authority and good planning, people ship work through sacrificing their free time and risking their social standing. This is brave but not particularly effective from the org’s perspective.

Professional silence

I thought a lot about Hanlon’s razor as I was reading this book. The book deals with bad leaders, but ‘incompetence’ and ‘malice’ are both synonyms for ‘bad’. Saddam fought his way to the top; I’m not sure I’d call him ‘competent’ or ‘expert’, but certainly talented in his chosen profession.

He also kind of embraced evil. This sounds dark, right:

“A leader is not made in a factory in Europe,” he observed. “Leadership is developed gradually. This was done underground.”50

He also kind of sucked.

I thought about times in my career when my ‘professional silence’ was a problem; I talked to an ex-coworker after we had both left who described borderline behavior by a department head. I was like ‘nobody listened to that person, their power wasn’t real’. My friend shook their head and said ‘I wish somebody had told me.’

Gossip is good. Secrecy siloes suffering, letting everyone have the same problem without knowing it, and all going quietly crazy. Don’t let leaders lie in public; it’s a middle manager’s job to call them out. Rock the boat.

Our arguments aren’t always good. Our plans aren’t always vetted. We should fix that.

Saddam’s problem with professionals struck me. One of the legacies of a hard year in 2020 will be a global conversation about managing experts. Do leaders need to be experts themselves? (The vice president of Taiwan is an epidemiologist, that’s pretty cool). If not, how do we evaluate a leader’s ability to work with experts?

As a domain expert, what can I do to better manage up? Experts have an obligation to convince, but not everyone can be convinced. How can I spot the ones who can’t?

Last thought on Hanlon’s razor: corruption feeds on silence. I heard a Tara Westover interview about how abuse comes hand in hand with justification and silencing. I’m not sure if silencing people is incompetent or malicious but I do think it’s usually bad. But middle managers are hard to silence. Don’t forget to use that once in a while.

Notes

  1. Woods et al, Saddam’s Generals, kindle L323. For the record this is the worst-formatted e-book I’ve ever seen. It’s also insanely cheap (like $3 I think) and educational whether you finish it or just read the intro. ↩

  2. Woods, L359 ↩

  3. Talmadge, The Dictator’s Army, kindle L1,301 ↩

  4. Talmadge, L4,007 ↩

  5. Talmadge, L4,063 ↩

  6. Talmadge, L2,383 ↩

  7. Talmadge, L2,664 ↩

  8. Talmadge, L2,360 ↩

  9. Talmadge, L2,745 ↩

  10. Talmadge, L4,331 ↩

  11. Talmadge, L3,426 ↩

  12. Talmadge, L3,432 ↩

  13. Talmadge, L2,569 ↩

  14. Talmadge, L2,711 ↩

  15. this got taken down so I’m not linking to the deleted tweet ↩

  16. Woods, L1,017 ↩

  17. Woods, L379 ↩

  18. Woods, L384 ↩

  19. Talmadge, L3,989 ↩

  20. Woods, L341 ↩

  21. Woods, L364 ↩

  22. Leslie Groves, Now It Can Be Told, p 373. If you’re looking for a 5000 word article about this book somebody wrote one. It’s basically the same outline as this one with opposite conclusions. ↩

  23. Talmadge, L3,890 ↩

  24. James Stavridis, Preet Bharara podcast, 20:45 ↩

  25. Stavridis, 21:00 ↩

  26. Stavridis, 48:40 ↩

  27. Talmadge, L1,542 ↩

  28. Talmadge, L4,199 ↩

  29. Talmadge, L4,200 ↩

  30. Talmadge, L4,326 ↩

  31. Woods, L354 ↩

  32. Talmadge, L4,463 ↩

  33. Woods, L237 ↩

  34. Woods, L256 ↩

  35. Woods, L266 ↩

  36. Woods, L220 ↩

  37. Woods, L226 ↩

  38. Talmadge, L3,906 ↩

  39. Woods, L261 ↩

  40. Woods, L1,193 ↩

  41. Woods, L333 ↩

  42. Woods, L1,038 ↩

  43. Woods, L1,084 ↩

  44. Talmadge, L4,690 ↩

  45. Talmadge, L4,701 ↩

  46. Woods, L454 ↩

  47. Talmadge, L4,484 ↩

  48. Talmadge, L2,777 ↩

  49. Talmadge, L4,234 ↩

  50. Talmadge, L3,135 ↩