Culture: What You Tolerate, You Teach

Eli Heard Everything

Eli was the priest at Shiloh and by any ordinary measure a good man. He blessed Hannah. He raised Samuel. He was the recognized spiritual authority of Israel at a time when, as the text says, word from the LORD was rare. And his sons, Hophni and Phinehas, ran the sanctuary as a racket. They took meat from the sacrifices by force before the fat was burned, which is to say they stole from worshipers and from God in the same motion, and 1 Samuel 2:22 says Eli heard about all of it — including what they did with the women who served at the doorway of the tent of meeting.

So he spoke to them. That is the part people miss. Eli was not silent. His words in 1 Samuel 2:23-25 are perfectly reasonable: why do you do such things, the report I hear is not good, if one man sins against another God will mediate, but if a man sins against the LORD who can intercede for him? That is a correct rebuke. It is theologically sound. It changed nothing, and the text tells you plainly why: they did not listen to their father.

Then comes the verdict, and it does not fall on the sons alone. God tells Samuel that He is about to judge Eli's house forever "for the iniquity which he knew, because his sons brought a curse on themselves and he did not rebuke them" (1 Samuel 3:13, NASB 1995). Read that against 2:23 and you have an apparent contradiction — he did rebuke them, we just read it — and this is one of those places where the argument rests on a word, so we go to the Hebrew.

The verb rendered "rebuke" in 3:13 is spelled kahah (כהה). Its far more common sense, attested across dozens of Old Testament occurrences, is to grow dim, faint, or weak — it is the word used of eyes losing their light and of a flame guttering out. Here, in a form that occurs nowhere else in the Hebrew Bible, the standard lexicons (BDB, HALOT) list a separate, rarer root under the same spelling carrying the sense of restraining or rebuking, and this is where the translations divide: many English versions render it "restrain," which fits the context better than "rebuke" precisely because Eli plainly did speak. Whether that rare root is truly a distinct word or an intensified extension of "grow dim" is disputed among lexicographers, and I am not going to manufacture certainty about it. This verse carries a second, better-attested textual issue that I want to name rather than hide: the Masoretes flagged "brought a curse on themselves" as one of eighteen places where they preserved a scribal softening of an original "cursed God," a reading the Septuagint still has. Nothing in this lesson's argument turns on that phrase, so I am not going to lean on it, but an honest treatment of this verse names it. What the narrator has already told you eleven verses earlier still stands on its own: in 1 Samuel 3:2, Eli's eyes had begun to grow dim — the same root, in its common sense. Commentators have long observed the wordplay, and whether or not it is deliberate, it is a precise description of the failure. Eli's sight failed and so did his grip. He saw enough to speak and not enough to stop it.

That is the whole lesson in one man. The standard was stated. The correction was delivered. The behavior continued, and continuing without consequence is what taught Israel what the priesthood at Shiloh actually was. By chapter 4 the ark is captured and both sons are dead, and Israel's estimate of its own worship has collapsed along with them. Rot that a leader speaks against but does not stop is not partially handled. It is fully authorized, with a disclaimer attached.

Lesson 7.8 asked what only you can do. Here is one of the answers, and it is not on anybody's job description: only you can set the line. Which brings us to a definition worth memorizing. Culture is not what you announce. Culture forms at The Tolerance Line — the worst behavior your people have watched you walk past — and everything above that line is aspiration.

Three Signals: Reward, Permit, Repeat

Operationally, a leader emits three signals continuously, and the organization assembles its culture out of them. The first is what you reward — who gets promoted, who gets the good assignment, whose name you say in front of the room, and, more revealingly, what the promoted people are actually like. The second is what you permit — what happens after somebody crosses a line, including nothing. The third is what you repeat — what you bring up in every meeting, what you measure, what you ask about first when you walk in.

Permission is the loudest of the three and it is the one leaders consistently underestimate, for a specific reason: rewards and repetitions are things you do, and you know you are doing them. Permission is inferred from inaction, which means you can issue it constantly without ever noticing. Your people, however, notice with great precision, because the cost of misreading you falls on them. When the top salesperson humiliates a junior colleague in a meeting and nothing happens, the organization has just learned its real exchange rate between results and dignity. You did not say anything. That was the message.

Edgar Schein gave the most widely used map of this territory, and I want you to hold it correctly. In Organizational Culture and Leadership, Schein describes culture at three levels: artifacts — the visible stuff, the office layout, the rituals, the language; espoused beliefs and values — what the organization says it believes, the values statement, the strategy; and basic underlying assumptions — the unexamined, taken-for-granted beliefs that actually drive behavior and that members cannot easily articulate because they no longer notice them. His crucial claim is that the top two levels routinely contradict the bottom one, and when they do, the bottom one wins.

Schein also catalogued what he called primary embedding mechanisms — how leaders actually transmit culture, whether they intend to or not. His list includes what leaders systematically pay attention to, measure, and control; how leaders react to critical incidents and crises; how they allocate resources; deliberate role modeling and coaching; how they allocate rewards and status; and how they recruit, select, promote, and remove people. Notice that four of the six are not communication at all. They are decisions, most of them made under pressure, and the crisis item matters most because that is when everyone is watching hardest and you have the least time to perform.

Now the honesty this academy requires. Schein's model is a practitioner-scholar framework built from consulting engagements and clinical observation, not a theory validated by controlled experiment, and it has been criticized for being difficult to falsify — the third level is defined as tacit and unarticulated, which makes disconfirming evidence hard to specify. Use it as a diagnostic lens that reliably shows leaders something they were not seeing. Do not cite it as a finding.

And while we are being honest, deal with the slogan. "Culture eats strategy for breakfast" is attributed to Peter Drucker in roughly every leadership deck ever built. People who have gone looking for it in his published work have not found it. Its business circulation is usually traced to its use inside Ford during the Mulally era, and the aphorism may well be older than that, but the Drucker attribution should be treated as undocumented. I am not asking you to abandon the idea — the underlying claim, that a strong culture will defeat a strategy it does not believe in, is defensible and observable. I am asking you to stop borrowing Drucker's authority for a sentence nobody can show he wrote. A leader who is careless with a small attribution will be careless with a large one, and your credibility is one instrument, not several.

A Little Leaven

Paul reaches for the same mechanism twice, in two different fights. To the Corinthians, confronting a church that was tolerating open sexual immorality and feeling rather sophisticated about it, he writes: "Your boasting is not good. Do you not know that a little leaven leavens the whole lump of dough?" (1 Corinthians 5:6, NASB 1995). To the Galatians, fighting an entirely different battle over circumcision and the gospel, he uses the identical line: "A little leaven leavens the whole lump of dough" (Galatians 5:9).

The Greek word is zymē (ζύμη), ordinary sourdough leaven — a small portion of fermented dough worked into a new batch. What matters is the mechanism the image encodes, and it is worth stating plainly because it is exactly wrong in the way most leaders think about tolerated behavior. Leaven does not sit in the corner of the dough being a small problem. It propagates. It changes the whole lump by contact and time, and it does so quietly, and there is no stage at which it is contained. That Paul uses the same image for a moral compromise in one letter and a doctrinal one in another tells you he regarded the dynamic as general. Small tolerated things do not stay small and do not stay local.

The empirical literature on norms says something structurally similar, from an entirely different starting point. Cialdini, Reno, and Kallgren's focus theory of normative conduct, developed through field experiments published in 1990, distinguishes descriptive norms — what people actually do here — from injunctive norms — what people here approve or disapprove of. Their littering studies found that people were substantially more likely to litter in an environment that was already littered, and that a single visible cue of what others do could shift behavior in either direction. This is a well-replicated line of work and it is directly applicable: your people are not reading your values statement to determine what is normal. They are reading the environment, and tolerated behavior is the environment.

You will notice how close this sits to broken-windows theory, and here I want to model the discipline rather than skip it. James Q. Wilson and George Kelling's 1982 Atlantic essay argued that visible disorder signals that no one is in charge and invites more serious crime. As a description of how signals shape behavior it is intuitive and it overlaps with the norms research. As an empirical claim about crime causation it is genuinely contested — reanalyses, including work by Bernard Harcourt and Jens Ludwig published in 2006, questioned whether the New York evidence supported the causal story, and the broader debate about what drove the crime decline of the 1990s remains unresolved among serious criminologists. So take the transferable principle, which the norms research supports on its own terms, and do not import the policing conclusion as settled. Signals shape behavior. What follows for police strategy is a separate argument requiring separate evidence, and we take it up in Module 9.

The practical translation is uncomfortable. If leaven is the right model, then the question is not whether a tolerated behavior is serious enough to address. The question is whether it is going to propagate, and small things propagate fastest because they are cheap to imitate. Nobody copies the spectacular ethical violation. Everybody copies the meeting that starts eight minutes late, the report that goes out without being checked, the senior person who is exempt from the rule everyone else follows.

The City Is a School

Aristotle saw this two millennia before organizational behavior existed as a field, and he saw it at the scale of a city rather than a company. His argument in Book II of the Nicomachean Ethics is that moral virtue comes about as a result of habit — we become just by doing just acts, temperate by doing temperate acts, brave by doing brave acts. The virtues are not in us by nature and not produced by instruction; they are produced by repeated action until the action becomes a settled disposition. Then he draws the political conclusion that most modern readers skip: this is why lawgivers make citizens good by forming habits in them, and why it makes no small difference whether we are habituated one way or another from youth. It makes, he says, all the difference.

Sit with the implication. A community is a formation program whether or not anyone designed it as one. Its laws, its praises, its punishments, and its ordinary daily requirements are its curriculum. A city that reliably praises one thing and reliably declines to punish another is teaching a course in what a human being ought to be, continuously, to everyone inside it. This is why Aristotle treats the constitution of a regime as an educational question and not merely an administrative one. And it is why a mayor is a moral educator whether he wants the title or not — not through his speeches, which people discount, but through what his government consistently rewards and consistently ignores.

Alasdair MacIntyre adds the part that should worry any executive. In After Virtue he distinguishes practices from institutions. A practice — medicine, teaching, policing done well, playing chess seriously — has internal goods, goods you can only obtain by pursuing excellence in that practice on its own terms. Institutions deal in external goods: money, power, status, survival. MacIntyre's point is that institutions are necessary — no practice survives long without one to sustain it — and that the institution's appetite for external goods is a standing threat to the practice it exists to protect. The ideals of the practice are always vulnerable to the acquisitiveness of the institution, and only the virtues of the people inside it hold the line.

That is the most useful description of institutional decay I know, and it explains something the org-chart account cannot. Agencies rarely fail because their people forgot how to do the work. They fail because the institution's need — for a clean number, an absence of scandal, a budget defended, a commissioner protected — quietly outranked the internal standard of the work itself, and the people who kept insisting on the standard became inconvenient. Nobody voted on this. It happened at the tolerance line, one exception at a time.

So here is the mid-lesson application, and I want you to actually do it rather than nod at it. Think of the practice at the center of your organization — the thing done well that justifies your existence. Now name the last time the institution's need won a contest against that standard, and name who noticed. Somebody noticed. In most organizations, the people who notice first are junior, because they are the ones who were told the standard most recently and have not yet learned which parts of it are decorative.

The First Firing and the First Exception

Take it to City Hall, where the mechanism is visible at industrial scale. A newly inaugurated mayor issues an inaugural address that nobody in the agencies will still be quoting in March. Two things he does in the first six months will still be governing behavior in his fourth year: the first person he removes, and the first exception he grants.

The first removal is the loudest cultural act available to a chief executive, because it is a demonstration rather than a claim. Whom you fire, for what, and how publicly tells 300,000 people what is actually disqualifying here. Fire someone for incompetence and competence becomes real; fire someone for disloyalty and loyalty becomes the currency; fire no one at all through a year of visible failure and you have taught the most consequential lesson of the three, which is that nothing is disqualifying. The first exception works the same way in the other direction. The first time a favored commissioner is allowed to skip the procurement review, or a political ally's agency is exempted from the performance regime everyone else is held to, you have not made one exception. You have published a rule about who the rules are for, and every subsequent enforcement action will be read against it.

New York has run this experiment twice with its police department, and the commission reports — not the press coverage — are the sources to read. The Knapp Commission, appointed by Mayor Lindsay in 1970 after Frank Serpico's allegations and reporting in 1972, described corruption that was widespread and organized, and gave the culture its vocabulary: "grass eaters," who accepted the payoffs that came their way, and "meat eaters," who aggressively pursued them. Knapp's most important finding was not about the meat eaters. It was that the sheer number of grass eaters made corruption respectable inside the department, and that a code of silence, reinforced by the department's own unwillingness to acknowledge the problem, kept it alive.

Twenty-two years later, Mayor Dinkins appointed the Mollen Commission, which reported in 1994 under former judge Milton Mollen. Its findings are the more damning of the two for our purposes. The character of the corruption had changed — no longer department-wide organized graft but crews of officers in particular precincts involved in drug-related corruption and, significantly, brutality. But Mollen's central institutional conclusion was about tolerance: the department's own anti-corruption apparatus had deteriorated, and commanders had learned that uncovering corruption was more damaging to a career than failing to find it. An institution had concluded that scandal was more dangerous than rot. That is a tolerance line, drawn by thousands of small decisions nobody would describe as corrupt, and it produced exactly what a tolerance line always produces. The commission's recommendation for permanent external monitoring led to the creation of the Commission to Combat Police Corruption in 1995.

So take the diagnostic and use it on yourself before somebody uses it on you. What is your tolerance line — the worst thing you have visibly walked past in the last ninety days? Who is the exception in your organization, the person everyone knows the rule does not apply to, and what have you taught by leaving them there? And what does your promotion record say about what this place actually rewards, as opposed to what your values statement says? Answer those three honestly and you will know your culture better than any survey will tell you.

One more thing before we close this out. Everything in this lesson is dangerous in the hands of a leader who wants a pretext for severity. Enforcing a standard is not the same as being harsh, and the leader who discovers this material and immediately becomes a disciplinarian has missed it. The point is consistency, not force: the same standard applied to the person you like and the person you need, publicly enough that people can see it is the same. That requires something we have not fully addressed yet — an ambition that is aimed at the institution rather than at your own standing inside it. That is Lesson 7.10, and it is where Module 7 ends.