Response

Do You Have a Minute for a Call?

res.p.000001A BDR once sent me a Teams message while I was in the middle of a busy day.

res.p.000002“Do you have a minute for a call?”

res.p.000003I had hired an outside consultant to help us move faster. He was experienced, expensive, and good at what he did. One of the things he had been working on was a reconnect campaign aimed at people who had attended one of our webinars and then gone quiet. The campaign had been reviewed by people inside the company. It was nearly ready to go.

res.p.000004The BDR had a problem with it.

res.p.000005She spent her days much closer to the people we were trying to reach than I did. She heard the questions they actually asked. She knew which parts of our value proposition made people lean forward and which required explanation. The campaign was trying to sell efficiency. What she was hearing in the market was something simpler: people wanted simplicity.

res.p.000006Those ideas are close enough to survive a conference room. They are not the same thing when you put them in front of a customer.

res.p.000007The consultant had been listening to senior internal voices with considerable knowledge of the business and considerably less daily exposure to the conversations happening at its edge. I had been too busy to oversee his work closely, which was one of the reasons I had hired him in the first place. The system was behaving reasonably. I had delegated to someone capable. He had gathered input from experienced leaders. A campaign had emerged from the process.

res.p.000008And the BDR thought we were about to say the wrong thing.

res.p.000009I was grateful she called.

res.p.000010There was no investigation into whether she had enough evidence to justify interrupting the work. I did not need her to prepare a deck explaining why the expert was wrong. I called him. We talked through what she was hearing. He understood the problem immediately and was grateful for the correction. We recalibrated the campaign and changed it before the outreach began.

res.p.000011Nothing heroic happened. Nobody saved the company. An expensive consultant did not become less expert because someone closer to the market knew something he did not. I did not become a better leader because I happened to answer Teams that afternoon. We caught a mismatch while it was still cheap.

res.p.000012Had she stayed quiet, the campaign probably would have launched. Then the cost would have changed shape.

res.p.000013Messages would have gone out. Prospects would have received language that did not quite resemble the problem they were living with. Some would have ignored it. Others might have engaged for the wrong reason and discovered the difference later. We would have gathered performance data and begun deciding what the data meant. Perhaps we would have changed the subject line, adjusted the sequence, questioned the audience or blamed the age of the leads. Enough weak response might eventually have forced us back toward the message itself.

res.p.000014By then we would have learned, slowly and expensively, something one person already knew before we sent the first email.

res.p.000015Worse, we would have told the market that we did not understand it.

res.p.000016That is the territory of this half of the book.

res.p.000017Signal was concerned with our availability to reality: what interferes with perception, what makes correction difficult, how fear and desire recruit intelligence, why other people must remain capable of telling us something we did not generate ourselves. But reception is only part of the problem. Reality reaches us constantly. What happens next determines how much work attaches itself to the difference between what we have learned and what we continue doing.

res.p.000018A piece of bad news can change a plan while the plan is still cheap, or it can spend six months acquiring dependencies. An employee can tell you that a commitment is impossible before you make it, or a team can demonstrate the impossibility through missed deadlines after everyone has organized around the promise. A customer can tell you what they need while the message is still being written, or implementation can explain the mismatch after the contract is signed. A number can contradict your theory before the theory becomes strategy, budget and reputation.

res.p.000019The information may be identical. The response changes its cost.

res.p.000020Correctability does not make life frictionless. Honest conversations still end badly; clear perception does not route us around grief, scarcity, conflict, obligation, or the stubborn independence of other people. Much of what is difficult remains difficult because reality contains limits, irreversible choices, and things worth caring about.

res.p.000021But we add difficulty of our own.

res.p.000022We make a promise and then construct labor around the gap between the promise and our capacity. We protect a decision after the evidence has changed because the decision has become attached to our competence. We ask people for candor and teach them, through our reactions, how much candor we can actually tolerate. We build reporting systems to recover information that could once have traveled through conversation. We optimize work built on a premise nobody has checked recently. Then, because all of this activity is real, it becomes difficult to distinguish the work required by the problem from the work required by our relationship to the problem.

res.p.000023The distinction matters because those two kinds of difficulty behave differently.

res.p.000024Reality does not disappear when we become better aligned with it. Often it becomes harder to evade. The BDR’s call created work that afternoon. We had to stop, reconsider something we thought was nearly finished, tell a consultant that his work needed to change and make those changes before launch. Continuing would have been easier in the next ten minutes.

res.p.000025Response has a horizon.

res.p.000026The cheapest action now may create an expensive Tuesday three months from now. The awkward conversation today may prevent a whole architecture of avoidance. The clean no may force a prioritization somebody hoped not to make. The corrected number may ruin a beautiful explanation. Sometimes doing less work requires being willing to interrupt work that already feels productive.

res.p.000027And sometimes the right response does not make anything cheaper at all.

res.p.000028There are jobs I have declined because I did not believe in what the company was successfully selling. I work in an industry nobody has ever accused of being glamorous because I think preserving information over long periods of time serves a genuine social good, and I want the company doing that work best to win. Another choice might have produced more money, more status or a more interesting answer at a party. Alignment cannot mean that the choice I believe is right must eventually reveal itself as the secretly more profitable one. That would turn reality into a morality play written around my incentives.

res.p.000029Signal does not promise that.

res.p.000030Response cannot either.

res.p.000031What it can do is ask a more practical question. Once reality has reached us—through data, consequence, another person, our own recognition, or the uncomfortable failure of something we expected to work—how much additional difficulty will we create before we allow what we know to change what we do?

res.p.000032The BDR needed less than a minute to ask for the call.

res.p.000033That was enough.

Chapter 1 — The Work Created by the Gap

res.p.000034For a while, I had an attribution problem that looked like a mathematics problem.

res.p.000035Our marketing systems were overstating return on investment. Not everywhere and not in a way that made the entire reporting environment useless, but enough that some channels were receiving more credit than anyone looking closely at the customer journey could really defend. The configuration had produced a number. The number was precise. The precision was the problem.

res.p.000036Finance noticed.

res.p.000037Once people stop believing a number, they do not simply stop using it. Usually they begin asking the number to prove itself. More cuts of the data appear. More qualifications get added. People who understand the system explain why the output should not be interpreted literally. Someone creates a more sophisticated view for people capable of handling the caveats. Then another view is needed for people who might take the first one too seriously.

res.p.000038We were trying to make the model more trustworthy by making it harder for an uninitiated person to misunderstand.

res.p.000039There was logic to this. Attribution really is difficult. A customer may encounter a paid advertisement, return later through search, attend a webinar, read several emails, speak with sales and finally convert through a channel that deserves almost none of the causal credit implied by being last. Different attribution models answer different questions. Anyone promising a perfectly objective division of revenue among those interactions is usually concealing a judgment somewhere inside the mathematics.

res.p.000040But our problem had become stranger than that.

res.p.000041The more obviously the original configuration overstated some forms of contribution, the less comfortable we became exposing the model without explanation. The less comfortable we became exposing it, the more expertise was required to interpret it. The more expertise interpretation required, the easier it became for everyone outside the small group of experts to distrust the thing as a black box.

res.p.000042We were generating sophistication partly to compensate for a representation we already knew was too confident.

res.p.000043Eventually, the answer was not another attribution model.

res.p.000044Finance and marketing sat down and admitted what the system could not tell us cleanly. Some channels were ambiguous. We knew they contributed. We did not believe the existing model had earned the right to claim all the value it assigned them. So we negotiated a discount.

res.p.000045Not a metaphysical solution to attribution. Not the number hidden somewhere in nature waiting for sufficiently advanced analytics to reveal it. A mutually tolerable adjustment to the channels where the evidence did not support the precision of the reporting.

res.p.000046Then we automated it.

res.p.000047The result was deliberately less ambitious than the machinery we had been building around the problem. We had a representation everyone understood. Finance knew where judgment entered the model. Marketing did not have to defend numbers we privately considered too generous. The adjustment happened consistently instead of being rediscovered every time somebody opened the report.

res.p.000048There was still uncertainty. We had simply stopped asking the reporting system to disguise it.

res.p.000049That distinction shows up everywhere.

res.p.000050A representation is useful because reality is too large to carry around whole. A forecast compresses thousands of uncertain events into a number. An organizational chart compresses a network of relationships into boxes and lines. A campaign brief compresses a market into an audience, a problem and a message. A job description compresses a shifting body of work into responsibilities. A calendar turns values, promises and constraints into blocks of time. A bank balance reduces an enormous range of future possibilities to a number we can act against.

res.p.000051We need these things. Human beings could not function without representations.

res.p.000052Trouble begins when the representation and the thing represented drift apart while the rest of the system continues behaving as though they still correspond.

res.p.000053Then work appears.

res.p.000054If the forecast claims more certainty than the evidence supports, people spend time explaining variance against a number that was never as solid as its formatting suggested. If an organizational chart assigns authority somewhere other than the place decisions are actually made, employees learn both structures and quietly translate between them. If a campaign describes a customer problem customers do not recognize, marketing optimizes response while sales explains the mismatch later. If a job description says someone owns an outcome but every meaningful decision requires approval elsewhere, meetings accumulate around the missing authority.

res.p.000055None of that work is imaginary. The analyst really does need to reconcile the figures. The employee really does need to know who can approve the decision. Sales really does have to handle the customer who arrived with the wrong expectation. The meeting really may be necessary by the time it appears on the calendar.

res.p.000056The interesting question is why it became necessary.

res.p.000057That question is easy to lose because compensatory work often looks respectable. A dashboard looks like management. An approval chain looks like governance. Another layer in an attribution model looks like analytical rigor. A recurring meeting looks like coordination. Sometimes those things are exactly what they appear to be. Financial controls exist because money can be mishandled. Hospitals use redundant checks because human error can kill people. Teams communicate because information really is distributed. A second pair of eyes can catch something the first missed.

res.p.000058The goal is not to build a world with no resistance, redundancy or verification. That would be an extraordinarily efficient way to become fragile.

res.p.000059The useful distinction is between work demanded by reality and work demanded by the gap between reality and the way the system represents it.

res.p.000060The attribution problem taught me something about that distinction because every person involved was behaving reasonably. The original model was not created by villains. Finance was right not to trust numbers it thought were overstated. The people complicating the model were trying to make it more accurate. The people restricting access to the complicated model were trying to prevent incorrect interpretation.

res.p.000061Each response made sense locally.

res.p.000062Together they were producing a system in which an imperfect number generated distrust, distrust generated complexity, and complexity generated another reason for distrust.

res.p.000063That kind of loop can run for years.

res.p.000064Organizations are full of people competently maintaining arrangements nobody would design from scratch. There is the report assembled every Thursday because two systems disagree. The manager who has to explain what a policy “really means.” The unofficial spreadsheet beside the official platform. The employee everybody asks before following the documented process. The executive dashboard that requires a private conversation to interpret correctly. The meeting whose agenda makes little sense until somebody explains the political history that created it.

res.p.000065Often these people are doing excellent work.

res.p.000066That is what makes the condition durable.

res.p.000067If the person reconciling the numbers is good enough, the numbers reconcile. If the manager translating the policy is perceptive enough, the team functions. If the employee maintaining the spreadsheet is conscientious enough, the official system continues to look adequate. Competence can absorb the consequences of a mismatch so effectively that the organization experiences the compensation as proof that the underlying design works.

res.p.000068Response begins by refusing that inference.

res.p.000069The fact that we successfully dealt with a problem does not tell us whether the problem needed to exist in that form. The fact that a process produced an acceptable outcome does not tell us what people had to carry to produce it. The fact that a representation is useful does not entitle it to more certainty than the reality behind it can support.

res.p.000070In the attribution case, we did not finally discover the perfect number.

res.p.000071We made the uncertainty visible enough to respond to it honestly.

res.p.000072That was cheaper than continuing to build machinery around pretending it was not there.

res.p.000073And it suggests the first question worth carrying through the rest of this book:

res.p.000074What work would disappear if the way we described the situation corresponded more closely to the situation itself?

res.p.000075Not all of it.

res.p.000076Enough to look.

Chapter 2 — A Story Starts Charging Rent

res.p.000077A sales vice president once asked me why her territory had received fewer leads than comparable parts of the business over the previous three months. I had an answer surprisingly quickly. We had recently made changes to our paid advertising, I had been involved in those changes, and I was already carrying some anxiety about their unintended consequences. Seasonality was also plausible; different markets do not move in perfect synchrony, and three months is long enough for a temporary imbalance to begin looking meaningful. I put together a series of hypotheses around those possibilities and shared them with her. None of them was foolish. I knew the campaign changes had occurred, knew they could affect volume, and knew seasonality existed. I had enough experience to construct a coherent explanation from real facts, and enough personal investment in the recent changes for one class of explanation to become unusually available.

res.p.000078Then I asked my digital team to dig into the data. Lead volume had not actually declined relative to the other territories. Qualification had.

res.p.000079The phenomenon I had explained was not the phenomenon occurring. My team had not discovered some obscure analytical nuance I could never have known; they had checked the premise. I had moved quickly enough from a question to an explanation that I skipped the small indignity of asking whether the question accurately described reality. My anxiety supplied the direction and experience supplied the supporting material. Had the team confirmed my explanation, none of this would have been visible. I might even have looked perceptive: recent advertising changes, seasonality, territory performance, a senior marketer rapidly identifying likely causes in a complicated system. Instead, reality disagreed early enough that we could simply change the problem. We were no longer asking why marketing had produced fewer leads in that territory. We were asking why a similar volume of leads was producing fewer qualified outcomes. The first question pointed toward advertising, audience generation and seasonality; the second pointed elsewhere. Different data mattered, different people needed to be involved, and different interventions became plausible.

res.p.000080I think about that episode whenever people talk about the cost of being wrong as though error itself were the expensive part. Error is ordinary. Any person or organization making decisions under uncertainty will produce it. What becomes expensive is allowing other things to organize themselves around the error. A hypothesis can remain cheap for a long time if everyone remembers that it is a hypothesis. We can test it, discard it, modify it or discover later that part of it was right. Trouble begins when its grammatical status quietly changes. Maybe this is happening becomes this is what is happening, often without any explicit decision that the evidence has crossed that threshold. Once the story begins producing decisions, it begins accumulating dependents.

res.p.000081Had I not asked the team to check, I might reasonably have changed campaign allocation to restore the volume I believed we had lost. That change would have altered the data arriving the following month. Someone might have produced new creative for the territory; budget might have moved from another market; the sales vice president could have told her team that marketing was addressing the lead shortfall. A quarterly review might eventually have included a slide explaining what had happened and what we had done about it. Every person involved could have performed their part competently. We would simply have been getting better organized around something that was not true.

res.p.000082Capability makes this problem more interesting, not less. A weak organization may stumble over a bad premise quickly because it cannot execute much of anything. A strong one can move money, people, messages and attention with considerable speed. Give that machinery a good description of reality and capability becomes leverage. Give it a bad description and capability remains leverage; it simply points somewhere else. The same dynamic operates inside a single mind. Believe that your boss does not trust you and a short reply acquires meaning. You become more careful in meetings, the caution makes you look less confident, your boss asks more questions, and the questions become evidence for the story that produced the caution. Decide that a team lacks commitment and you begin watching missed deadlines and weak follow-through differently. You push harder, status conversations acquire an accusatory quality, people become defensive, and the defensiveness enters the file as additional proof that they do not take ownership. The explanation begins producing some of the world from which it gathers confirmation.

res.p.000083None of this requires stupidity. Intelligence can make the loop more durable because intelligence gives us more material with which to explain what we already expect to see. Complicated situations contain an abundance of true facts, usually enough to support several different stories. Asking whether there is evidence for a story is therefore less protective than it sounds. Of course there is evidence. The more useful question is what we decided the evidence needed to explain. In my conversation with the sales VP, I had evidence that advertising had changed, evidence that markets experience seasonality, and evidence that her results differed from what she expected. What I did not yet have was evidence that lead volume had declined relative to her peers. The dependent variable was missing from my explanation of the dependent variable. Once stated plainly, the mistake is almost embarrassingly obvious. Many useful corrections are.

res.p.000084Explanations nevertheless feel productive, particularly when understanding the system is part of your job. “I don’t know yet whether that is actually happening” can feel like a failure to contribute when everyone else is waiting for interpretation. Expertise creates pressure to reduce ambiguity, and leadership intensifies it. People bring you anomalies because they want the anomaly made legible. Experience is valuable precisely because it allows patterns to become recognizable earlier, but that speed creates its own discipline: sometimes the competent response is to leave the anomaly unresolved for another hour while someone checks whether the thing requiring explanation is actually there.

res.p.000085Organizations vary enormously in how well they tolerate that interval. In some, a person can say, “I think this is happening, but I haven’t checked the premise yet,” and everyone understands the distinction. Forecasts can contain assumptions. Projects can become yellow before someone has assembled an airtight case for red. A leader can ask a question without everyone assuming the question already contains the preferred answer. Where that tolerance is absent, uncertainty does not disappear. It changes clothes. It becomes a precise number built from weak assumptions, a confident status supported by private caveats, a strategy deck whose clean arrows conceal an argument still taking place in the data, or a project described as on track because one increasingly implausible route remains by which the date could technically survive. The uncertainty is still present, but now somebody also has to maintain the certainty.

res.p.000086That maintenance becomes real work. A project manager who knows a deadline is in danger may still report the project as on track without consciously deciding to deceive anyone. Perhaps the team could still make it. Perhaps the manager wants a few more days to understand the problem before escalating. Perhaps every red status becomes an executive event and they have learned to arrive with solutions rather than uncertainty. Once “on track” enters the organization as fact, however, the project acquires a second state. There is the project the team is actually doing and the project other people believe is happening. A scope reduction is no longer only a scope decision because it may expose the previous status. A new dependency problem requires both an operational response and a narrative explanation. People can find themselves working toward one date while speaking about another. The organization has created a second project whose deliverable is preservation of the first project’s description.

res.p.000087The multiplication is familiar elsewhere. A salesperson allows a customer to believe a feature exists because correcting the misunderstanding might endanger the deal, and implementation later inherits two products: the one the company actually sells and the one the customer believes they bought. A leader says disagreement is welcome and then punishes the first inconvenient disagreement, leaving employees to navigate two cultures: the one described in leadership meetings and the one useful for surviving them. Experienced employees become translators between official and actual systems. They know which deadline is real, which policy is ceremonial, which number needs checking, which executive needs to hear something privately before it can be said in the room. Some of this is ordinary institutional knowledge; no written process can fully describe a living organization. Some of it is accumulated expertise in maintaining contradictions the organization has stopped noticing.

res.p.000088You can often see the distinction when one of those experienced people leaves. A process everyone assumed was stable suddenly fails. The report no longer reconciles. A handoff begins producing errors. Approvals take twice as long. Leadership concludes, correctly, that the departing employee possessed unusual knowledge. Sometimes that knowledge consisted largely of knowing where the organization’s description of itself stopped being true, and knowing how to close the gap quietly enough that nobody else needed to notice it.

res.p.000089Deliberate lying is too narrow a frame for the problem. Conscious deception belongs here and can become enormously expensive, but the more common cases are built by people who believe themselves. I believed my hypotheses about lead volume because they made sense. Their plausibility did not make them true, and the protection against that kind of error cannot be endless self-doubt. If every experienced person had to approach every situation from complete epistemic innocence, experience would be useless. Pattern recognition and judgment are real capacities. Sometimes the first explanation is exactly right. The discipline is simpler: let reality answer before too many things depend on you having been right.

res.p.000090My team did. The correction cost almost nothing. Nobody had to rescue my reputation from the hypotheses because I had not yet made them carry it. We checked the premise, discovered a different problem and moved. The story never had time to start charging rent.

Chapter 3 — When the Decision Becomes You

res.p.000091A decision changes character when changing it begins to feel like changing the story you tell about yourself.

res.p.000092At first, most decisions are fairly ordinary objects. We make a recommendation, choose a direction, hire someone, back a project, predict an outcome. The choice may matter a great deal, but it remains something we have done. Then other things attach themselves to it. We explain the reasoning in a meeting. We persuade someone who disagrees. Money moves. People start working. Perhaps we speak with enough confidence that others stop reserving judgment and begin relying on ours. By the time contrary evidence arrives, the decision has accumulated social history. Abandoning it no longer means only that the plan should change. It may mean that someone who warned us was right, that the argument we made was weaker than we thought, or that other people acted on confidence we supplied.

res.p.000093The evidence itself has not become more complicated. Our relationship to it has.

res.p.000094Ordinary errors survive well past the point where they remain useful. The decision begins as a claim about the world and gradually becomes evidence about the person who made it. Once competence, authority or reputation depends on its survival, every new fact has to do two jobs: tell us something about the situation and negotiate with whatever the situation now seems to say about us. A strategy can be doubtful by March and indefensible by June, yet the cost of reversing it may feel higher in June precisely because we have spent three additional months building an identity around being the people who chose it. Financial sunk cost explains only part of that persistence. The psychological investment has its own ledger.

res.p.000095Competent people can become particularly good at maintaining it. We have experience, analogies, context and vocabulary. We know markets change, teams execute unevenly, experiments need time, customers misunderstand offers, competitors behave unexpectedly and outcomes contain luck. Every one of those propositions is true often enough to be dangerous when we need one of them to rescue a conclusion. Intelligence gives us more ways to find reality. It also gives us more ways to bargain with it.

res.p.000096The distinction I have found useful is whether I want an explanation to be true or need it to be true. Wanting is ordinary. Nobody approaches evidence without hopes, fears or preferences. Needing has a different texture. The range of acceptable explanations begins narrowing before the evidence has finished arriving. The humiliating interpretation receives a prosecutorial standard of proof while the flattering one gets admitted on circumstantial evidence. We experience ourselves as investigating when we have quietly become counsel for the defense.

res.p.000097I know another version of this problem from music.

res.p.000098Musical practice gives you thousands of encounters with error that are too immediate to sustain much mythology. You play the note sharp. The entrance is late. The rhythm drifts. A phrase that felt expressive turns out, on a recording, to be shapeless. You tense before a difficult passage and rush the easy one after it. The mistake arrives in sound, and the instrument is indifferent to the story you were telling yourself while you made it. You can explain why the note was sharp if the explanation helps you fix it. The room does not owe you a different frequency because your intentions were excellent.

res.p.000099The response is usually practical: slow the passage down, change the fingering, isolate two measures, subdivide the rhythm, record it, use a metronome, listen to the person beside you, try again. Over time a bad note learns to stay local. I can be a musician and still have played that passage badly. I can understand what the phrase requires and fail to produce it. I can spend weeks on a technical approach and discover that the approach itself is creating the problem. The failed attempt does not have to turn the previous weeks into fraud or the next attempt into a referendum on whether I deserve to make music.

res.p.000100Keeping the error local makes it much easier to respond precisely. If my intonation is unstable in four measures, I have somewhere to work. If the conclusion is that I am bad at music, the diagnosis has expanded until it contains almost no usable information. The first description points toward an adjustment. The second recruits the whole self into a problem that originally occupied a few seconds.

res.p.000101Music becomes even more interesting when other people enter the room. A note can agree perfectly with a tuner and be wrong inside a chord. A tempo that felt right alone may drag against an ensemble. A phrase that made sense in practice may become indulgent when another part enters underneath it. The standard no longer resides entirely inside the player. You have to listen outward while continuing to play.

res.p.000102That combination is harder than it sounds. Ensemble playing requires commitment. Tentative playing creates problems of its own; nobody wants to perform beside someone who is perpetually withholding the note until certainty arrives. You choose the pitch, the articulation, the tempo and the shape of the phrase, and you play them as music rather than as a provisional memo. At the same time, the room may answer. The conductor hears balance differently from the podium. The section hears an articulation problem the conductor cannot. A younger musician notices that everyone has normalized a tuning issue. Someone else breathes, moves or changes emphasis, and suddenly the phrase you had correctly understood a moment earlier requires something different.

res.p.000103You commit, and you remain permeable.

res.p.000104That is a form of decisiveness I trust much more than certainty. It allows a person to act before every variable is known without requiring the action to become permanent proof of judgment. Yesterday’s decision can have been responsible given yesterday’s information and still become obsolete today. Responsibility survives the revision. What disappears is the requirement that consistency mean reaching the same conclusion after the conditions have changed.

res.p.000105Many workplaces make this much harder than rehearsal rooms do. Correction arrives carrying status. A project going red can become a judgment on the project manager. A forecast miss can become a judgment on the analyst. Reversing an executive decision can be read as evidence that the original decision was foolish, which encourages everyone involved to spend another month discovering increasingly sophisticated reasons why the evidence has not quite crossed the threshold for reversal. When uncertainty threatens rank or competence, people learn to perform certainty instead. Plans become more detailed as confidence weakens. Retreat becomes sequencing. Failure becomes evolution. Sometimes those words accurately describe what happened. Sometimes they provide enough linguistic padding for a decision to change without anybody having to acknowledge that reality changed it.

res.p.000106A humane system should leave room for some face-saving. Public humiliation is an atrocious error-correction mechanism. If updating a belief requires ceremonial self-abasement, people will defend bad beliefs longer. Accountability has a different purpose. Someone who wasted money, harmed another person or ignored warnings may have consequences to face even after changing course. Correction does not erase what happened. It simply prevents the preservation of a fictional self from becoming an additional condition that must be satisfied before the correction can begin.

res.p.000107Leadership complicates all of this because the leader’s psychology becomes part of the information system. People who work with you learn how you receive bad news. They learn whether “What do you think?” means that you want an answer or that you want them to participate in arriving at yours. They learn whether disagreement is safest in the meeting, beforehand in private, or afterward once the decision can no longer change. They know whether you need context before a problem, whether uncertainty irritates you, whether being surprised produces curiosity or an interrogation about why you were not told sooner. None of this requires an explicit rule. Human beings are excellent students of consequence.

res.p.000108A leader who needs to be right therefore imposes a tax on information before the information reaches them. A peer may say directly that an idea is bad. Someone whose career you can affect has more calculations to perform. How committed are you to the proposal? Who else supports it? Are you actually asking for criticism? If they disagree and you proceed anyway, will they later be remembered as the person who was not on board? The factual content of the message may take five seconds to formulate while the political computation surrounding its delivery takes an afternoon.

res.p.000109The BDR who asked me for a call about the webinar campaign did not need to perform much of that computation, or at least not enough to keep her quiet. That was not an achievement I could safely infer from a policy or a leadership principle. The evidence was the interruption itself. She believed that what she saw in the market could alter work that had already passed through people with more authority and an outside expert with more status. When it did alter the work, the next interruption became a little easier to imagine.

res.p.000110That is how correctability becomes an organizational property. It is built from accumulated evidence about what happens to the person who brings the contradiction. Leaders can help by making their own revisions visible, particularly when the original position was public. “I was wrong about that” is useful less as a ritual of humility than as information about the local rules. It tells people that a judgment can move without the person who made it disintegrating, and that disagreement does not automatically become disloyalty when the disagreement survives contact with evidence.

res.p.000111The harder work is internal. Most of us have qualities we want to believe are characteristic of us: competent, generous, rational, loyal, perceptive, independent, fair. Those aspirations can orient a life. They become brittle when an individual act is required to certify the identity. A person committed to fairness has to remain capable of discovering that they acted unfairly. Someone who values rationality has to remain capable of finding an emotional motive in their own reasoning. A leader who believes they are good at reading people needs some way to survive a bad hire without converting every subsequent problem into evidence that the employee changed after the interview.

res.p.000112I prefer identities with verbs in them. I try to lead well. I make music. I write. I try to be honest. The grammar leaves room for events. A writer can write something dull without acquiring a new essence. A musician can play badly. A person who tries to be generous can discover resentment or selfishness without needing either to deny it or to rebuild the entire self around the discovery. Character still matters; repeated choices really do form patterns, and we should care about those patterns. The burden becomes unbearable when every choice has to prove the whole character at once.

res.p.000113Once that burden is reduced, decisions can become strangely ordinary again. A proposal fails and we change it. A forecast is wrong and we inspect the model. A strategy stops fitting the market and we stop funding it. Someone tells us that our interpretation of an argument is wrong and we have enough room to consider the possibility before preparing the rebuttal. None of these responses guarantees wisdom. They simply shorten the distance between encountering an error and becoming able to do something with it.

res.p.000114The rehearsal room has given me the cleanest image I know for that distance. You play the passage. You hear what happened. Perhaps somebody else hears what you missed. Then you go back to the beginning and play it again.

res.p.000115The previous attempt does not need to become anything more than information.

Chapter 4 — What Experience Is For

res.p.000116The previous chapter ended in a rehearsal room, with a mistake becoming information instead of identity. That still leaves another question. What happens to the information after the mistake is corrected?

res.p.000117Experience is often described as accumulation. Spend enough years doing something and you acquire a large inventory of situations: difficult clients, failed projects, good hires, bad hires, campaigns that worked, campaigns that did not, arguments you wish you had handled differently, technical approaches that turned out to be dead ends. The inventory matters. Familiar situations become less surprising because you have literally seen more of them. But simple accumulation has a limit. A person can experience the same kind of failure for twenty years and become extremely experienced at having it.

res.p.000118The useful transformation happens when the lesson separates, carefully, from the event that produced it.

res.p.000119I learned piano and music theory before I learned many of the other instruments I have picked up since. That did not make the instruments interchangeable. Knowing how harmony works does not give my hands the technique required by a guitar, tell me how to control breath on a wind instrument, or make the physical geometry of a new instrument disappear. The particulars remain stubbornly particular. What changed was that I did not have to relearn music every time I encountered a new way of producing it. Intervals still had relationships. Chords still had functions. Rhythm still organized events in time. Tension still wanted somewhere to go. A new instrument presented a new interface to a structure that was already partly legible.

res.p.000120That distinction between the interface and the structure is close to what I mean by experience becoming useful. If I learn only that a particular fingering failed on a particular instrument in a particular measure, the lesson may be exactly as narrow as it needs to be. Technique often is that local. But sometimes the event teaches something larger: I am repeatedly practicing too much of the passage instead of isolating the point of failure; I am speeding up before the hard section because I anticipate difficulty; I am listening to myself instead of the ensemble. Those observations can survive the piece that exposed them. Eventually a new mistake does not arrive before an empty mind.

res.p.000121Marketing gave me another portable lesson long before I thought about it in those terms. The objection someone states is not always the whole objection. A buyer says the product is too expensive, but the real problem may be that they do not believe it will work well enough to justify any price. Someone asks for another feature when their deeper concern is whether their organization can successfully implement the features already present. A prospect says they need more time when the decision is effectively already made and time is the socially easier language for no. Good marketing requires taking the stated objection seriously while remaining open to the possibility that the first available words do not completely describe what is moving the person.

res.p.000122That lesson travels, but it has to travel modestly. People often argue about the thing they have language for. An employee may ask for a title while trying to determine whether the organization understands their actual contribution. Someone may be furious about a household task while carrying a larger resentment about whether their effort is visible. An executive asking for another report may really be responding to the experience of important decisions occurring outside their field of view. None of those possibilities entitles me to tell another person what they secretly mean. Marketing has not granted me interpretive sovereignty over everyone I meet. It has given me a question I know how to ask: is the stated problem the whole problem?

res.p.000123That difference is the difference between a portable lesson and a universal theory.

res.p.000124Theories become dangerous when they are allowed to arrive in a new domain with diplomatic immunity. Business people are especially fond of this. Business becomes war, so competitors become enemies and market share becomes territory. A company is described as a family, so employment begins borrowing the moral vocabulary of kinship when sacrifice is required. Families become teams, relationships become negotiations, employees become resources, attention becomes a currency, and suddenly a clever metaphor has converted a large part of human life into things a management framework knows how to manipulate.

res.p.000125The resemblance may be real. That is what makes these metaphors useful enough to become dangerous. A family and a company both operate under constraints. Both allocate time and money. Both contain people with different needs. Both can suffer when nobody makes tradeoffs. But a company can stop employing someone when the relationship no longer makes economic sense. If that principle survives the trip into family life intact, the analogy has destroyed the thing it claimed to illuminate. The portable lesson has become larger than the reality receiving it.

res.p.000126A good principle therefore remains correctable when it crosses a boundary.

res.p.000127Consider scarcity. I can learn at work that calling six things priorities does not repeal the existence of finite capacity. That travels home perfectly well. A family calendar can also become impossible when every individual commitment has been accepted as though the others did not exist. Yet the actual ranking cannot simply be imported from work. A project can be canceled because its sponsor has lost interest. A child’s school event may carry a kind of obligation that cannot be evaluated by the same criteria. The general lesson survives: refusing to make tradeoffs does not eliminate tradeoffs. The new domain still gets to determine what the tradeoffs mean.

res.p.000128I increasingly think of principles less as rules than as compressed experience. “Bad news should travel early” may begin with one project in which a hidden problem became expensive. Then it encounters another situation and acquires a condition: bad news has to travel to someone capable of doing something with it. Another experience adds something else: people will not transmit it early if the receiver consistently punishes them for doing so. Another teaches that not every fluctuation deserves escalation. The principle survives, but it stops being a slogan. Its growing complexity eventually makes its use feel simpler because the conditions have become part of recognition.

res.p.000129The same happens with perseverance. Early in life, “do not quit when things become difficult” is often a useful correction to the fact that difficult things are unpleasant. Then perseverance encounters sunk cost. It encounters a relationship that is causing harm. It encounters a project whose original assumptions no longer hold. If the principle cannot change, it becomes stupidity with excellent branding. A mature version retains the value of persistence while learning that continued effort is not proof of courage merely because stopping would be embarrassing.

res.p.000130Honesty encounters privacy. Generosity encounters boundaries. Confidence encounters evidence. Efficiency encounters dignity. Loyalty encounters behavior that should not be protected. Each encounter adds structure to the principle without requiring us to discard what made the principle useful in the first place.

res.p.000131This is also why broad experience sometimes produces unusually fast judgment. The person is not necessarily thinking faster in the ordinary sense. They may be bringing a larger library of relationships into the situation before conscious analysis begins. A supposed execution problem resembles other situations in which nobody had established which priority was allowed to lose. A request for “visibility” resembles earlier cases in which the actual problem was missing control rather than missing information. A team repeatedly repairing the same failure begins to look less like a collection of isolated mistakes and more like a system that has learned to depend on repair.

res.p.000132The nouns differ. The structure has begun to look familiar.

res.p.000133There is a temptation, especially after recognizing a few of these patterns successfully, to become impressed with one’s own ability to see them. That is when cross-domain thinking needs another constraint. An analogy is not evidence. It is a way of deciding where to look. The fact that an organizational problem resembles a musical one does not prove they share a cause. The fact that a relationship resembles a negotiation does not mean negotiation theory contains the correct response. The pattern earns the right to generate a question, not to overrule what the new situation says back.

res.p.000134I often test an analogy by asking what it makes disappear. If a model of efficiency makes dignity invisible, the missing variable is not redeemed by the elegance of the optimization. If describing an organization as a machine makes the agency of the people inside it inconvenient, the machine is a poor model of at least one important feature of the organization. If a political metaphor requires actual human beings to become interchangeable units in order for the theory to remain clean, the messiness of the human beings is evidence against the cleanliness.

res.p.000135Reality retains veto power over the lesson we hoped to transfer.

res.p.000136Genuinely broad experience is different from merely collecting techniques. Later experiences can revise earlier ones. Something learned in a relationship can make a management failure from five years earlier suddenly legible. A professional mistake can give language to a personal habit that had previously felt like unrelated frustration. A design principle can make information architecture clearer. A musical practice can change the way criticism is received at work. The direction of transfer is not always from the supposedly serious domain into the supposedly recreational one. Sometimes a hobby has been teaching the professional self something the professional self was too busy to notice.

res.p.000137The experiences begin talking to one another.

res.p.000138When that happens, time starts behaving differently. Twenty years of experience no longer has to mean twenty separate annual deposits into twenty separate accounts. A lesson purchased in one part of life can earn something elsewhere. An old mistake can become newly useful because another domain supplies the concept that was missing when the mistake occurred. A skill can change what is perceptible before it changes what we can execute. The value of what happened is no longer confined to the reason we were doing it at the time.

res.p.000139This does not mean life should become an elaborate scheme for extracting professional value from everything we love. I have no interest in turning music into leadership training with instruments, or playing strategy games with my children into a covert curriculum whose success must someday appear on a résumé. Play is allowed to be play. Music is allowed to be music. Time with people we love is not underperforming because it fails to produce a second deliverable.

res.p.000140Yet things learned there still belong to us.

res.p.000141Playing strategy games with my children gives them occasions to encounter tradeoffs, uncertainty, resources, incentives and the consequences of decisions inside a space whose actual purpose is that we are playing together. Certain video games introduce economic relationships because economic relationships happen to make the games interesting. I do not need to convert the evening into a lesson plan for any learning to occur. The mind is quite capable of carrying useful structure home without being ordered to monetize the trip.

res.p.000142That is what integration changes. The boundaries between domains remain, because their differences carry information. Work is not family. Music is not management. A market is not a relationship. A child is not an employee. But the boundaries become permeable enough that we stop buying the same lesson from reality over and over again simply because it arrives wearing different clothes.

res.p.000143Experience compounds when the correction survives its original use.

res.p.000144The next problem will still contain facts I have never seen before. It may require expertise I do not possess and distinctions my existing models cannot make. Nothing about integration abolishes novelty. It simply means novelty does not require amnesia.

res.p.000145I get to bring what reality has already taught me.

Chapter 5 — Recognition Is Faster Than Deliberation

res.p.000146Experience changes more than what we know. Eventually it changes what we notice first.

res.p.000147A beginner confronting a complicated situation often experiences the complexity more or less at face value. Many variables are visible, many explanations remain possible, and considerable conscious work goes into deciding which deserve attention. Someone with deep experience in the same domain may encounter exactly the same information and experience a much smaller problem. They have not necessarily processed every variable faster. Much of the field has already been organized by previous encounters with similar structures. Some possibilities barely register because experience has taught them that those possibilities rarely explain this kind of situation; other details become conspicuous because they have mattered before.

res.p.000148I notice this in fields I know well when I compare them with fields I do not. Give me a set of marketing performance data and I do not experience every number as an independent fact requiring equal scrutiny. The figures begin grouping themselves almost immediately. A problem may look like demand, fit, conversion, measurement or capacity before I can fully explain why. The classification can be wrong, and the previous chapters should make clear how interested I am in preserving that possibility, but the experience of looking is different from what it was twenty years ago. The same thing happens when I read an argument in a familiar subject. Certain claims carry more weight because I know what they imply downstream. Others look routine. A sentence that might pass unnoticed by a newcomer catches my attention because it quietly depends on something contentious. Experience has changed the order in which the material becomes salient.

res.p.000149That change can look almost mystical from the outside. Someone walks into a difficult situation, listens for a few minutes, asks one basic question and apparently collapses a sprawling problem into something manageable. We call this instinct, intuition, taste, strategic thinking or occasionally genius, depending on how impressed we are. The labels suggest that the experienced person has acquired some additional faculty unavailable to ordinary deliberation. Often the opposite has happened. They are considering less because years of earlier consideration have already taught them that many available branches do not deserve equal weight.

res.p.000150Gary Klein and his colleagues encountered a striking version of this while studying experienced fireground commanders. These were people making consequential decisions under conditions in which leisurely comparison among alternatives was not especially practical. In the published account of the original research, Klein, Roberta Calderwood and Anne Clinton-Cirocco examined 156 decision points drawn from interviews with twenty-six commanders who averaged twenty-three years of experience. Evidence that the commanders had simultaneously generated several options and compared them appeared in fewer than twelve percent of those decision points. In more than eighty percent, the commander recognized the situation as resembling a familiar type and identified an action associated with it.

res.p.000151This did not mean that experienced firefighters had stopped thinking. It meant that deliberation was doing a different job. Rather than beginning with a broad set of abstract possibilities, the commander could recognize a plausible course of action and test it against the situation. Would it work here? What happens next if we do it? Is there some feature of this fire, building or crew that makes the familiar response inappropriate? If the first recognized option failed that examination, another could be considered. Klein called the resulting account recognition-primed decision making. Recognition supplied a place to begin; conscious thought tested whether the recognized pattern deserved to govern the present case.

res.p.000152The distinction is important because most formal accounts of rational decision-making begin after an expert has already done something extraordinary: decided which possibilities belong on the list. We are taught to generate alternatives, establish criteria, compare tradeoffs and select the strongest option. Those techniques are valuable, particularly for unfamiliar or irreversible decisions. Yet no decision matrix explains how its rows arrived. If ten possibilities are compared instead of a thousand, some prior act of judgment has already reduced the field. Expertise often lives in that reduction.

res.p.000153Chess gave cognitive psychologists another way to see the same thing. William Chase and Herbert Simon found that skilled players were exceptionally good at remembering meaningful chess positions after very brief exposure, while much of that advantage disappeared when the pieces were arranged randomly. The master was not simply remembering more isolated objects. Familiar relationships among pieces could be perceived as larger structures. What a novice had to hold as several independent facts could arrive for the expert as a single meaningful configuration.

res.p.000154That kind of compression is easy to underestimate because the information does not disappear. A strong player still sees the individual pieces. An experienced marketer still sees spend, conversion rate, lead quality, traffic, pipeline and revenue. A good musician still hears notes. What changes is the scale at which relationships become perceptible. Once several elements participate in a familiar structure, the structure itself becomes available as an object of thought. The mind gains capacity by no longer needing to reconstruct every relationship from first principles.

res.p.000155This is the cognitive payoff of the portable experience from the previous chapter. If lessons remain trapped inside the circumstances that produced them, every new problem arrives with all of its surface complexity intact. Once relationships begin surviving their original context, the mind has more ways to organize novelty. Responsibility without authority stops looking like a peculiar feature of one troubled project and becomes a recognizable source of delay. A metric repeatedly distorting behavior stops looking like a sequence of unrelated measurement disputes. A team that keeps solving the same emergency may eventually look less like an unusually resilient team and more like a system whose normal operation now includes emergency repair.

res.p.000156The apparent speed comes from seeing the problem at a different level.

res.p.000157This can be unnerving in organizations because requests usually arrive already categorized. We need another dashboard. We need more traffic. We need a meeting. We need another person. We need a campaign. The requested object may be exactly right, but it is also a theory about what kind of problem exists. Someone asking for a dashboard is proposing that missing information is preventing action. Someone asking for headcount is proposing that useful work exceeds available capacity. Someone asking for more leads is proposing that additional opportunities would improve the outcome. Accept the category too quickly and an organization can become impressively competent at producing the requested object while the original condition remains untouched.

res.p.000158Experienced judgment sometimes shortens that path by testing the category itself. Which decision cannot currently be made because the dashboard is missing? What work will the new employee own that nobody has capacity to perform? What is happening to the leads already entering the system? What will be different after the meeting? These questions can sound almost offensively simple after a problem has acquired enough stakeholders, terminology and work around it. Complexity creates social weight. If several teams have spent months discussing something, a small explanation can feel inadequate to the amount of effort already invested.

res.p.000159Reality has no obligation to provide an explanation proportional to the meeting hours accumulated around it.

res.p.000160But the small explanation is not always right. Fast recognition has a failure mode, and it looks uncomfortably similar to expertise.

res.p.000161Every stereotype is a recognition system. So is the manager who hears the beginning of an employee’s concern and thinks, I know this type. So is the executive who has watched three transformations fail and can explain, before the fourth begins, why people always resist change. So is the political partisan who can infer an entire moral character from one sentence. Human cognition saves enormous effort by assigning the present case to a class already known. Sometimes that assignment is expertise. Sometimes it is prejudice with a long résumé.

res.p.000162The feeling of familiarity does not tell us which one we have.

res.p.000163Daniel Kahneman and Gary Klein came from research traditions that had often emphasized opposite sides of intuitive judgment, but they eventually converged on a useful boundary. Reliable intuitive expertise requires an environment containing regularities that can actually be learned and enough opportunity for the person to encounter feedback from those regularities. Confidence alone tells us almost nothing about whether the intuition has been trained against something real.

res.p.000164Chess is unusually generous in this respect. The rules remain stable. Meaningful configurations recur. Moves produce consequences. A serious player can encounter enormous numbers of positions and receive repeated feedback from an environment that does not care how strongly the player believed a bad move was good. Firefighting is messier and vastly more dangerous, but experienced commanders still encounter recurring cues in buildings, fire behavior, smoke, crew conditions and the consequences of previous actions. There is enough structure for repeated exposure to teach something.

res.p.000165Many of the decisions human beings most want to develop intuition about offer nothing like that quality of feedback. An executive may make only a few acquisitions in an entire career, and the consequences unfold over years amid market changes, integration problems, leadership turnover, financing conditions, competitor behavior and luck. A parent may adopt some approach and watch a child thrive without ever being able to separate the effect of the approach from temperament, peers, school, chance and the thousand other variables that constitute a childhood. An investor can make a reckless decision in a rising market and receive an extremely persuasive reward for bad reasoning. A person can handle a relationship poorly, have the other person accommodate them, and experience the absence of further conflict as proof that the approach worked.

res.p.000166Some environments teach. Others merely produce outcomes.

res.p.000167Success makes the distinction particularly difficult because good outcomes feel like feedback even when they are not clean evidence about the decision that preceded them. A reckless bet can win. A careful one can lose. A strategy can succeed because the market moved in its favor rather than because its assumptions were sound. When cause and consequence are separated by enough noise, experience may increase confidence considerably faster than it increases accuracy. Thirty years in such an environment can produce less expertise than a few years in a domain where errors are frequent, visible and difficult to explain away.

res.p.000168This gives “trust your gut” a more useful follow-up question: what trained it?

res.p.000169Years alone do not create expertise. Relevant repetitions have to be dense enough for patterns to recur, and the environment has to answer clearly enough for cause and effect to connect. Bad reasoning can still produce a good outcome. Other people can absorb failures. Contrary evidence can be hidden or explained away. Experience teaches only when something in the loop is allowed to correct the model.

res.p.000170Leadership is particularly treacherous because power can damage the training environment while making the leader feel increasingly experienced. If I respond badly to disagreement and people eventually stop disagreeing, the disappearance of the signal can look like resolution. If I make weak decisions and capable employees repeatedly compensate for them, the organization may continue producing good outcomes while I accumulate evidence that my decisions work. The higher a person rises, the more likely it becomes that someone else can repair a local failure before the consequences return cleanly to the person who caused it. Success can surround a leader with an environment exquisitely adapted to preventing certain lessons from arriving.

res.p.000171That is not an argument against intuition in leadership. It is an argument for knowing which intuitions have earned their speed.

res.p.000172The pattern library remains enormously valuable. An experienced leader should be able to hear a situation and think that it resembles a priority problem, a missing-authority problem or a distorted-incentive problem before a less experienced person has finished enumerating the symptoms. The recognition gives investigation a starting point. The danger arrives when this resembles a priority problem becomes this is a priority problem because I recognize priority problems. One formulation uses experience to decide where to look. The other uses experience to eliminate the possibility that looking could produce surprise.

res.p.000173That distinction becomes especially important once experience starts crossing domains. The previous chapter argued that lessons can travel without making the domains interchangeable. The benefit is a much larger pattern library; the risk is that almost anything can be made to resemble something we already know. An organizational incentive resembles an economic system. A disagreement resembles a negotiation. A rehearsal resembles a feedback process. A campaign resembles an experiment. The analogies may expose structure that would otherwise remain hidden, but a person with enough analogies can eventually construct one for every occasion. At that point broad experience stops increasing contact with reality and becomes a way of never encountering anything new.

res.p.000174The safeguard does not require theatrical uncertainty. People with real responsibility cannot prepend a ten-minute disclaimer about epistemic humility to every decision, and hesitation is not automatically wisdom. The useful remainder is smaller: I know what this resembles, and I know that resemblance is not identity. Recognition earns the first question, not the final answer.

res.p.000175Speed becomes useful when it remains correctable. A novice may need to consider a large field because experience has not yet taught them how to compress it. An expert can often begin much nearer the structure that actually drives the case. That compression creates real speed; correctability keeps it from becoming a sealed system. The person acts on what experience has taught while leaving enough of the situation unclaimed for the situation to answer back.

res.p.000176The thinking did not vanish.

res.p.000177Much of it happened earlier, slowly, over years of mistakes, repetitions, corrections and comparisons that taught the mind which differences deserve attention. What looks like intuition in the present may be an enormous archive of deliberation that has finally become perceptual.

res.p.000178Good judgment can be fast without being careless.

res.p.000179It is also why speed alone proves nothing.

Chapter 6 — Decide in Better Weather

res.p.000180There are decisions I would rather not leave entirely to the version of myself who most wants a particular answer.

res.p.000181This is not because that version is uniquely dishonest. He may be tired, flattered, frightened, ambitious, impatient, angry, embarrassed or simply standing close enough to an immediate reward that everything farther away has become visually smaller. The problem is less moral than perceptual. A decision made under pressure contains the same facts that may have been available a month earlier, but the facts do not arrive with equal force. The money is here. The disappointed person is here. The deadline is here. The awkwardness of saying no will occur in the next thirty seconds. The rework, exhaustion, damaged trust or lost evening belong to a future that has not yet acquired a face.

res.p.000182The present is very good at arguing its case.

res.p.000183Organizations experience this constantly around customer commitments. A large prospect wants a capability that does not yet exist. Product has discussed it. Engineering believes something adjacent may be feasible. The customer wants to know whether it will be available by October, and everyone on the call understands that the answer may decide the deal. Nothing requires anyone to tell a direct lie. “It is planned” may be true. “We expect to support that” may be defensible. A salesperson can choose language that allows the customer to hear more certainty than the organization actually possesses while retaining enough literal accuracy to explain the sentence later.

res.p.000184If the organization has never decided what its promises are allowed to mean, the entire philosophy of customer trust enters the call at exactly the moment when revenue has its strongest advocate.

res.p.000185That is a bad time to invent the philosophy.

res.p.000186An organization that has already learned something about these situations can arrive differently. Perhaps it has suffered through enough implementation failures to know that selling a future possibility as a present commitment merely moves the uncertainty downstream. Perhaps product teams have inherited promises made on their behalf and spent months converting someone else’s optimism into a roadmap. Perhaps sales has discovered that winning the wrong expectation produces a customer relationship organized around recovering from the first conversation. Over time, those experiences can become a principle: describe current capability as current, planned capability as planned, and do not convert another team’s possibility into a promise without that team participating in the commitment.

res.p.000187The principle does not answer the customer’s question. October may genuinely be achievable. The account may be important enough to change the roadmap. Engineering may decide the risk is worth taking. What the principle removes is a different debate: whether the attractiveness of this particular deal is itself sufficient reason to blur the distinction between possibility and commitment. That decision was made when the commission was not sitting in the room.

res.p.000188This is what I mean by deciding in better weather.

res.p.000189We tend to speak about principles as though their value lies mainly in moral seriousness. A person “has principles.” A company “has values.” The language makes them sound like possessions kept somewhere inside character. Their practical usefulness is often more mundane. A principle allows reasoning done under one set of conditions to remain available under another. It gives a future self access to a judgment made when a particular fear, desire or incentive did not occupy most of the visual field.

res.p.000190Human beings already rely on this mechanism in settings where improvisation would obviously be foolish. Fire alarms do not invite a fresh debate about whether remaining inside might be more convenient. Pilots use checklists for known failure conditions because emergency is an atrocious moment to discover which steps memory happens to preserve under stress. Technical systems route traffic away from a failed server according to conditions established before the failure. We understand that the very situations in which a response matters most may also be the situations least suited to constructing the response from scratch.

res.p.000191Outside those settings, we can become strangely sentimental about keeping every option open.

res.p.000192Openness feels like freedom. A rule feels constraining. Yet leaving a recurring decision permanently unsettled does not remove constraint; it hands more power to whatever pressure happens to be strongest when the decision returns. Hunger gets another vote on what you intended to eat. Flattery gets another vote on the boundary you thought you had established. Fear gets another vote on the risk you had previously decided was worth taking. Work gets another vote on whether the evening belonged to your family because work has a meeting request, a deadline and a person waiting for an answer, while the life you meant to protect is represented by a general conviction that family is important.

res.p.000193Repeated deliberation can become a way of ensuring that the decision is always made in worse weather.

res.p.000194Psychologist Peter Gollwitzer’s work on implementation intentions offers a narrower but useful version of the same mechanism. Instead of holding only a goal—exercise more, finish the task, avoid some habitual response—a person links a recognizable condition to an intended action: if this situation occurs, I will do this. The research does not tell us which principles deserve to govern a life, but it helps explain why prior decisions can change behavior later. Once the cue appears, some portion of the work of deciding has already been done.

res.p.000195The larger version is familiar in anyone whose values have ever lost to logistics. “My family is more important than work” can be completely sincere and causally weak. Work rarely arrives asking whether you would like to betray your family for the evening. It arrives as a meeting that can only happen at six, a trip that would be useful, a deadline that became urgent for reasons nobody planned, or one email that will take twenty minutes. Every request can be reasonable on its own. Meanwhile, the family is expected to defend its place through memory.

res.p.000196A value becomes more real when it acquires an interface with decisions.

res.p.000197That interface does not have to be grand. Some commitments enter the calendar before discretionary work does. Some evenings are unavailable except for circumstances that have earned the word emergency. Some forms of travel require a higher threshold than others. The details vary because lives vary, and a workable boundary in one season can become absurd in another. What matters is that “this is important to me” eventually changes something that would otherwise have happened.

res.p.000198A value that never survives a collision with another incentive is difficult to distinguish from a preference for describing ourselves a certain way.

res.p.000199Organizations face the same problem at larger scale. Safety, customer trust, employee welfare, sustainability, and integrity all sound admirable in a values statement. Their meaning becomes visible when a live decision puts revenue, speed, or executive preference on the other side. A value with no authority, budget consequence, decision right, veto, or class of behavior that remains unacceptable when violating it would be useful has been delegated to whichever employee happens to encounter the conflict.

res.p.000200The employee is then expected to reconstruct corporate morality while a customer is waiting.

res.p.000201Some organizations deliberately move certain principles into architecture instead. Patagonia offers a visible example: the company’s governance gives its purpose commitments formal structural protection, and its sourcing process gives social and environmental review the ability to stop a supplier decision rather than merely advise the people making it. Whatever one thinks of Patagonia or any individual decision it makes, the structural move is interesting. “This concern matters” has been translated into “this concern has authority here.”

res.p.000202Once that translation occurs, every supplier does not reopen the organization’s entire moral vocabulary for negotiation. Some possible trades have already been removed from ordinary circulation.

res.p.000203The danger is obvious.

res.p.000204Bad principles work by exactly the same mechanism.

res.p.000205“Never admit weakness” can make leadership remarkably efficient. “Protect the family no matter what” can eliminate the uncomfortable work of deciding whether a family member should face consequences. Racism is capable of extremely fast classification. Authoritarian systems save enormous amounts of deliberation by deciding that certain voices never deserve a vote. Fanaticism is cognitively cheap because almost every new situation arrives with its meaning already assigned.

res.p.000206Coherence is not the same thing as truth.

res.p.000207A principle reduces branching by refusing to reconsider something. That property has no moral content on its own. The question is whether the compressed judgment remains reachable by the reality it is supposed to help us navigate.

res.p.000208A principle is prior learning that remains capable of further learning. Dogma keeps the speed of compression by giving up the right to be corrected. Response depends on the first without quietly turning it into the second.

res.p.000209Good principles therefore acquire scars.

res.p.000210Persistence works until the objective has changed. Loyalty works until loyalty means protecting wrongdoing. Generosity works until generosity is financing exploitation. Trust works until contrary evidence makes continued trust an act of denial. Speed works until the decision becomes difficult to reverse. Honesty works alongside privacy because refusing to deceive someone does not entitle us to disclose another person’s confidence. Each encounter reveals something about scope. The principle becomes more mature because reality has found its edges.

res.p.000211This is different from manufacturing an exception whenever the principle becomes inconvenient. Human beings are talented enough to make those two operations sound nearly identical.

res.p.000212The rule applies, of course, but this customer is strategically important.

res.p.000213This deadline is unusual.

res.p.000214This political moment is existential.

res.p.000215This opportunity will never come again.

res.p.000216This relationship is complicated.

res.p.000217Any of those sentences can be true. They are also approximately the sentences we would generate if we simply wanted what the principle was preventing us from doing. Urgency produces very similar prose whether the exception is wise or desired.

res.p.000218A useful principle gives that prose something to push against.

res.p.000219If I believe I should not volunteer another person’s time without their involvement, then the executive request does not become impossible; it becomes legible as an exception if I make it anyway. Perhaps I decide that the emergency justifies the call. Perhaps the person explicitly authorizes me to make it. Perhaps the rule turns out to have been too rigid. What I lose is the ability to violate the principle repeatedly while preserving a story in which nothing has changed.

res.p.000220Three exceptions in a month are data.

res.p.000221Maybe the principle is unrealistic. Maybe my behavior is. Either conclusion gives reality somewhere to enter.

res.p.000222I do not find enormous personal rulebooks especially attractive. There is a point at which a system designed to reduce deliberation becomes another object that requires administration. A rule for mornings, email, meetings, exercise, money, travel, reading, relationships, food, planning, deep work and what to do when all those rules conflict can produce a person who no longer makes decisions so much as administers an internal compliance department.

res.p.000223The purpose of a principle is compression.

res.p.000224The few I actually trust tend to operate lower in the stack. Reality gets to answer. Other people remain real. A promise should resemble what I believe I can deliver. If I claim that something matters, some allocation of time, attention, risk or opportunity should eventually make the claim visible. Being wrong should remain cheaper than defending something I know is wrong. Money is valuable and useful without being competent to determine the value of everything else.

res.p.000225Those principles do not tell me whether to take a meeting at three or four. They do not decide which campaign to run or how much money belongs in a particular account. They constrain the arguments available when those ordinary decisions begin recruiting larger values on behalf of an immediate preference.

res.p.000226They also remain vulnerable to revision.

res.p.000227That vulnerability is part of why they are useful. If a principle repeatedly produces consequences that contradict the thing it was meant to protect, the principle has returned to deliberation. If a boundary designed to protect family life becomes a way of avoiding a responsibility only I can meet, reality has supplied new information. If an insistence on transparency becomes cruelty disguised as honesty, something was compressed badly. The prior decision has earned influence over the present. It has not acquired sovereignty over it.

res.p.000228This is the balance I want from a principle: strong enough that every pressure does not receive a fresh hearing, open enough that reality never loses its vote.

res.p.000229That balance changes the branching structure of a day. Some choices really are no longer live. Some familiar arguments do not need another performance. Some values have acquired enough practical consequence to survive a moment when violating them would be locally attractive.

res.p.000230The speed comes later.

res.p.000231The thinking happened when the weather was better.

Chapter 7 — What Work Leaves Behind

res.p.000232Some work is genuinely finished when it is finished. The dishes are clean. The package has been mailed. The form has been submitted. Nothing has gone wrong merely because the hour leaves behind no enduring capability, relationship or artifact. A life in which every ordinary task had to justify itself through some additional future return would become intolerable very quickly.

res.p.000233Other work has a longer afterlife, and the difference is large enough to affect how I choose what to do and how I arrange for it to be done.

res.p.000234Two projects can consume the same week, solve equally legitimate problems and receive equally competent execution. At the end of one, the result exists and almost everything required to produce it disappears with completion. At the end of the other, someone knows how to solve a class of problem they could not solve before. A useful template remains. A relationship has acquired trust because two people have now done difficult work together. The reasoning behind a decision has become legible to people who were not in the room. A tool built for one use can be adapted to another. The next version of the problem will still require effort, but it will not arrive at the same starting line.

res.p.000235The hours were spent in both cases. Their consequences were different.

res.p.000236This property is easy to miss because most systems of work are built to observe completion. The campaign launched. The analysis was delivered. The customer issue was resolved. The employee received an answer. Completion is visible, countable and often urgent. The residue is distributed across the future, where it may never be traced back to the project that produced it. Six months later, a person makes a better decision without needing help because they understood the reasoning the first time. Nobody records that as an output of the earlier conversation. A team solves something in half the time because somebody documented the strange part of the first implementation. The saved hours appear nowhere beside the original project’s cost.

res.p.000237I learned a surprising amount about design by sitting with designers while they were designing.

res.p.000238That is different from receiving a finished design and asking why it looks good. By the time the finished thing exists, many of the decisions that produced it have disappeared into the object. The designer has already tried the wrong spacing, moved the element that felt stranded, rejected the treatment that was technically consistent but somehow made the page feel cheap, and noticed that the hierarchy stopped working when a line wrapped. Ask afterward and you get the decisions large enough to remember. Sit beside the work while it is happening and you begin to acquire the smaller distinctions.

res.p.000239I have spent enough time in live working sessions with designers that some of their shorthand became available to me without my ever becoming a designer in the way they were designers. I learned to see when a page lacked hierarchy before I could always explain what should change. I became more sensitive to spacing, visual weight and the relationship between the amount of information present and the amount the eye could usefully organize. A designer could say something as compressed as “that’s fighting the headline” and I knew increasingly well what they meant. The work still had to get done. Nobody had added a class called Visual Design for the Marketing Guy. Learning was occurring inside the production of something we already needed.

res.p.000240The arrangement cost time. A designer working alone could make some decisions faster than a room containing me. I asked questions whose answers were obvious to them. They occasionally had to expose reasoning that would otherwise have remained almost automatic. I sometimes argued for an idea long enough to require them to demonstrate why it failed rather than simply discarding it. The next conversation started from farther along, though. I could bring a better brief, distinguish a structural problem from a decorative preference and recognize earlier when my own request contained a contradiction. Some of the designer’s judgment had become legible to another person, and some of my context about the underlying problem had become available to the designer. The work between us improved without either of us acquiring the other’s profession.

res.p.000241This creates a persistent bias toward local efficiency. If I can perform a task in thirty minutes and someone else will need two hours because they are learning, doing it myself is plainly faster. Once may be entirely sensible. Repeated often enough, though, the same calculation produces an organization in which I remain the only person who can do the task in thirty minutes. Each locally efficient choice protects the condition that makes the next locally efficient choice obvious.

res.p.000242Delegation exposes this particularly clearly. If I delegate because I want an item removed from my list, the relevant output is completion. The work returns and I recover the time. If I delegate a problem that is slightly beyond someone’s current range and stay available enough that the stretch remains survivable, the same work has another possible output: the range of problems that person can own has changed. The second approach may consume more of my time today. I may have to explain context, answer questions, tolerate an approach different from mine, allow a harmless mistake to occur instead of intercepting it, and resist the particularly seductive efficiency of taking the keyboard back.

res.p.000243The return appears later, if the learning was real. A problem that once required me now has another route through the system.

res.p.000244That is capacity in a more durable sense than getting more done this week. An organization does not become less dependent on a leader because the leader finishes tasks quickly. It becomes less dependent when knowledge, judgment and authority can move far enough outward that work no longer has to return to the same person for integration. A manager who personally synthesizes five departments can look indispensable and be performing terribly important work while simultaneously preserving the exact dependency that makes the work necessary. The organization experiences the manager’s competence as a solution. From another distance, the manager may be functioning as an expensive adapter between parts of the system that never learned how to speak directly.

res.p.000245Communication can leave the same kind of residue. A leader can announce a decision in three minutes. Sometimes that is exactly what should happen; nobody needs the complete intellectual history of the lunch order. Other decisions contain a model worth distributing. Spending another fifteen minutes on which variables controlled the choice, what tradeoffs were accepted and what evidence would have produced a different answer may allow several people to make related decisions later without returning for permission. The explanation has then done more than communicate a result. It has moved part of the decision-making system into more minds.

res.p.000246Documentation works similarly when it is created at the point where the knowledge is alive. Writing down what happened after a project can feel like administrative overhead, and much documentation deserves that reputation because it is produced for an archive nobody consults. Useful documentation captures the part someone would otherwise have to rediscover: the non-obvious dependency, the reason an apparently easier approach failed, the decision rule, the point where the official process no longer describes the real one. The artifact becomes valuable when it reduces the amount of institutional archaeology required the next time reality produces a similar shape.

res.p.000247Writing has this quality for me even when nobody else reads it. An idea I have only thought about can retain convenient ambiguity for a long time. Putting it into sentences forces one possibility to exclude another. Gaps become visible because the paragraph has to cross them. The finished piece preserves the result after I have forgotten most of the internal path that produced it, and another reader may find the weakness I had unconsciously written around. The writing therefore participates in the thinking, stores some of the thinking, exposes it to correction and occasionally creates a relationship around the thing being examined. One activity can have several effects because those effects grow naturally from the same act.

res.p.000248Teaching often produces a similar surplus. I can know how to perform something for years without knowing how to explain it. The attempt to make it legible to someone else discovers which parts of my competence are procedural, which parts depend on intuition I have never unpacked and where my own explanation contains a gap. The student receives something, but the teacher is also forced into another encounter with the material. A good question can expose an assumption that years of successful performance never made visible.

res.p.000249Benjamin Franklin’s Junto is an unusually clear historical example of this kind of architecture. Franklin organized the Philadelphia club in 1727 as a group for what he later described as mutual improvement. Members met, brought questions, discussed politics, morals and natural philosophy, and periodically wrote essays for the group. The arrangement educated them, but Franklin’s own account suggests that education was only one of the things moving through the network. The relationships also carried information about the city, professional reputation, business referrals, potential collaborators and opportunities for civic action. Books needed for discussion were pooled among members, and that local solution eventually contributed to the creation of the Library Company of Philadelphia.

res.p.000250What interests me about the Junto is not Franklin’s legendary industriousness. Industriousness is easy to admire and hard to operationalize. The structure is more useful. Reading could produce a question for discussion. Discussion could sharpen an argument. The argument could become writing. Writing could travel through Franklin’s publishing work. A problem noticed in the city could enter conversation among people who possessed different knowledge and relationships, and the conversation could become a civic project. The Friday meeting did not somehow contain more than sixty minutes in every hour. It was arranged so that the consequences of one activity had somewhere else to go.

res.p.000251Modern professional life often fragments the same functions and then treats their reconnection as additional work. We work and separately network because the work does not create many durable relationships. We execute and separately train because the work has ceased to teach. We produce things that cannot be shown and then invent portfolio projects to demonstrate capabilities the real work already required. We hold meetings to make decisions and later write documents to explain why the decisions were made. We schedule retreats in order to think because the normal flow of execution contains no place where reflection can survive.

res.p.000252Every one of those separate activities can be useful. Sometimes separation is the safer or better design. Formal education exists for reasons deeper than inefficiency, protected practice matters when beginner mistakes can hurt people, and many relationships become healthier when they are not asked to carry professional usefulness at all. The pattern becomes interesting when a system repeatedly requires a second activity to recover value that the first activity could have produced naturally if it had been arranged differently.

res.p.000253A project that teaches nothing reusable may be perfectly fine. A sequence of nearly identical projects in which every team rediscovers the same failure is harder to defend. A manager solving an employee’s problem may be exactly the help the employee needs. A manager solving every instance without improving the employee’s ability to recognize the next one has preserved a dependency while repeatedly experiencing themselves as supportive. A meeting that produces a decision has done its minimum job. If everyone leaves knowing only what was decided and nobody understands enough of the reasoning to act when the next case differs slightly, the meeting may also have guaranteed another meeting.

res.p.000254The relevant accounting therefore depends heavily on where we decide the work ends.

res.p.000255Training someone takes longer than doing the task yourself if the calculation stops at today’s task. Building the reusable version takes longer than building the one-off version if the calculation stops at today’s requirement. Writing down what you know takes longer than keeping it in your head if the calculation ends before anyone else needs the knowledge. Inviting a difficult question lengthens the meeting if the only output being measured is the meeting ending.

res.p.000256Widen the horizon and the numbers can change.

res.p.000257They do not always change. This idea becomes obnoxious the moment it turns into a requirement that every action produce five forms of capital. Sometimes the one-off version is correct because there will never be a second use. Sometimes the senior person should simply do the work because the deadline is real and training can happen somewhere else. Sometimes people are exhausted and the most intelligent use of an hour is to let nothing productive happen during it.

res.p.000258Rest does not owe the future a business case.

res.p.000259Neither does play. Neither does friendship. Neither does time with children. Music does not need to improve my executive function in order to deserve the hours I spend making it. A novel does not need to teach empathy. A walk is not wasting its potential because I left the audiobook at home. The language of compounding becomes grotesque if it converts everything valuable into a secondary form of work. And then sometimes I play music. Nothing else has to come out of the hour.

res.p.000260There are parts of life whose value depends partly on their refusal to be instrumentalized.

res.p.000261That boundary also protects other people. One person’s elegant arrangement can be another person’s invisible labor. A manager can describe unwanted work as a development opportunity because the phrase makes the transfer sound mutually beneficial. A company can ask employees to build an unpaid community and call the exposure career capital. A person can treat every friendship as partly a professional network and become puzzled when other people experience the relationship differently. An arrangement does not become well integrated merely because several benefits flow toward the person designing it.

res.p.000262Whenever one activity appears to solve several problems at once, the distribution belongs in the analysis. Who is doing the additional work? Who receives the residue? Who bears the uncertainty? Does another person gain capability, trust or opportunity from the arrangement, or have I simply found a sophisticated description for exporting cost?

res.p.000263The clean cases are reciprocal without requiring mathematical equality. A client receives useful work while the person doing it acquires legitimate repetitions. A junior employee grows into ownership while the organization gains another capable decision-maker. A student receives an explanation while the teacher discovers whether the explanation actually works. Collaborators solve a problem that neither could have solved as well alone and leave the exchange knowing more about one another’s judgment. The value moves through the relationship instead of accumulating exclusively at one end.

res.p.000264This changes how opportunity looks. Salary, title, revenue and prestige remain real. They are simply not the whole transaction. An opportunity also determines which problems will become familiar, which people will become close, which skills will receive hundreds of repetitions and which will slowly decay. It can increase autonomy or teach dependence. It can create visible evidence of capability or years of excellent work that remains inaccessible outside one institution. It can place someone near people who make them larger or inside a system that rewards a narrower version of them.

res.p.000265Two jobs offering similar compensation can therefore be very different pieces of a life without either one being secretly superior in the abstract. One may produce the income a family needs while leaving time and attention available elsewhere. Another may produce intense learning and relationships at the cost of energy that the person is willing, for a season, to spend. A promotion that creates remarkable professional leverage can still be the wrong choice if it consumes something the person has already decided is not for sale. Looking at the whole output is useful precisely because money and career growth do not automatically win once the accounting becomes more complete.

res.p.000266The most interesting opportunities are often hiding inside work we already have to do. Before adding another course, meeting, networking event, development project or personal system to an already crowded life, I increasingly want to know whether some necessary activity can leave a little more useful residue. Can the real project contain the practice? Can the explanation become the documentation? Can someone participate in a decision early enough to learn the model instead of receiving the answer afterward? Can a customer conversation improve the product team’s understanding of the market? Can a piece of work solve the immediate problem and make the next version of the problem less dependent on the same people?

res.p.000267Those questions are modest enough to survive contact with an ordinary week.

res.p.000268They also change the meaning of efficiency. The shortest route to finishing today’s task is not always the route that leaves the system in the best condition when the task is gone. Sometimes it is. The difference cannot be settled by admiring either speed or development in the abstract. We have to look at what the work consumed, what it produced, who carried the cost and what remains available afterward.

res.p.000269Completion is only one boundary we can draw around an activity.

res.p.000270Some of the most valuable work keeps moving after we stop.

Chapter 8 — Work at the Edge of What You Can Do

res.p.000271At some point in a career, somebody will tell you to become more strategic.

res.p.000272The advice is usually sincere and often correct. It is also difficult to act on because strategic is one of those professional words whose meaning expands as it rises through an organization. You can generally infer the criticism. You are too deep in execution. You solve the problem handed to you without asking whether it is the right problem. You know your function but not enough of the business surrounding it. You can optimize inside a frame someone else supplied, and the next level of work requires some ability to examine the frame itself.

res.p.000273Fine. How?

res.p.000274There are books for this, and courses, frameworks, case studies and off-sites where markets become quadrants and arrows move confidently from insight to action. Some are excellent. Formal instruction can give names to patterns we have felt without understanding, expose us to decisions we would otherwise have no way to observe, and save us from paying personally for every useful idea. What it cannot entirely reproduce is the moment when several legitimate claims are sitting in front of you at once, the information is incomplete, everyone owns a different part of the truth, somebody will have to be disappointed, and the recommendation will still belong to you when the meeting ends.

res.p.000275Strategy stops being vocabulary.

res.p.000276Suppose a part of a business is growing unusually fast. Revenue is attractive, but the customers producing it behave differently from the customers the company normally serves. Sales sees an opening and reasonably wants to move before a competitor does. Operations has started noticing that these customers require more exception handling than the revenue number reveals. Product sees requests that might indicate an adjacent market or might become a permanent diversion from the product it already knows how to build. Finance can make the opportunity look excellent or terrible depending on assumptions nobody can yet verify confidently.

res.p.000277No participant is obviously wrong. That is what makes the problem developmental.

res.p.000278A person remaining entirely inside one function can contribute useful information and wait for someone else to integrate it. A marketer can describe demand. Sales can describe the conversations. Operations can describe delivery. Product can describe roadmap consequences. Finance can describe the economics. Organizations need specialists who do exactly this. But the person trying to form a view of the whole problem has to cross those boundaries far enough to understand what the other claims actually mean.

res.p.000279The gaps in competence become specific very quickly. You need to know whether the revenue remains attractive after the sale rather than merely whether the contract is large. You need to understand why operations calls something difficult rather than carrying the word difficult back into the discussion as information from another country. You need to know which parts of sales performance are repeatable and which depend on a few unusually capable people. You need to understand what product would stop doing if the new opportunity became a priority, because a roadmap is partly a description of who will not get what they want. Eventually you have to distinguish the uncertainties that can remain unresolved from the ones that would change the decision.

res.p.000280Nobody had to invent a strategy exercise. The problem supplied one.

res.p.000281This is one of the most useful properties a real problem can have. It can reveal ignorance at a useful scale. “I need to understand finance better” is nearly infinite. “I cannot tell whether this apparently attractive revenue is actually good business” has a center. The question may send you into contribution margin, cost to serve, retention, implementation expense or pricing, but the new knowledge has something to attach itself to. You know what you are trying to become able to see.

res.p.000282The same difference appears in almost every serious capability. “Become a better leader” is too large to practice directly. Managing a situation in which a capable employee is underperforming for reasons you do not yet understand is smaller. “Learn data analysis” describes a continent. “Figure out why qualification fell while lead volume remained stable” gives the analysis somewhere to begin. “Become a better writer” is mostly an aspiration until another mind has to understand a difficult idea you have failed to make clear.

res.p.000283Ignorance becomes more tractable when it has an address.

res.p.000284There is an important boundary here. Work only teaches well when enough of the problem is already legible that the missing capability can be distinguished from everything else going wrong. Give a beginner a musical passage modestly beyond their technique and the difficult measure becomes visible. They can slow it down, isolate the movement, understand what the hand cannot yet do and practice toward it. Give the same player something wildly beyond their current level and everything fails at once. Reading collapses while rhythm collapses while the hands tense while tempo disappears. The performance offers one broad diagnostic result: no.

res.p.000285Useful difficulty has shape.

res.p.000286“Get outside your comfort zone” has always struck me as imprecise advice. Distance from comfort is not the variable that produces learning. A problem can be extremely uncomfortable because it is politically toxic, chronically under-resourced, structurally impossible or simply too large for the person being asked to carry it. None of those conditions guarantees development. A useful stretch lets someone encounter the edge of a capability while retaining enough footing to understand what the edge is.

res.p.000287Good stretch assignments are designed around that distinction. Imagine someone who needs broader business judgment. Giving them ownership of a cross-functional recommendation can be developmental if they can reach the people who hold necessary information, have enough authority to investigate the problem honestly, and remain close to a more experienced person who can interrogate their reasoning without taking the decision away from them. The assignment contains real stakes, and the employee has enough ownership for those stakes to answer their judgment.

res.p.000288There is another version of the same assignment. A failing initiative has become politically inconvenient. No senior leader wants it. Someone identifies a promising employee who “needs visibility,” hands them the problem without the authority to change its basic conditions, adds no capacity, and explains that the exposure will be valuable for their career.

res.p.000289That is development in approximately the sense that fabric develops character while being pulled apart.

res.p.000290Organizations confuse these situations because the vocabulary of growth can make neglect sound generous. Workload becomes opportunity. Lack of support becomes autonomy. A person surviving chaos becomes evidence that the chaos built resilience. Managers can persuade themselves that an employee should be grateful for access to a problem whose senior sponsors have carefully arranged not to own it.

res.p.000291Suffering has no obligation to become wisdom.

res.p.000292A problem trains its own solution only if contact with the problem can actually change the person’s capability. That requires feedback. It requires some connection between action and consequence tight enough for the learner to recover what happened. It often requires another person who can see distinctions the learner cannot yet see. It requires stakes real enough that the answer matters and bounded enough that the cost of a mistake does not make experimentation irresponsible.

res.p.000293Other people are often what make the learning real.

res.p.000294Management books become different after another human being refuses to behave like the example. An instruction that seemed perfectly clear is misunderstood. Feedback meant as supportive lands as threat. Someone you manage is better than you at something around which you have built part of your own professional identity. An employee conceals a mistake, and examination reveals that your own reaction to earlier mistakes helped teach them to conceal it. A person you like is failing in a way that imposes costs on everyone around them, and compassion for that person now has to coexist with responsibility to the rest of the team.

res.p.000295A sentence about accountability that seemed obvious on the page has acquired a pulse.

res.p.000296Leadership becomes learnable partly because people answer back. You cannot understand another person exclusively as a problem to be solved, because the person possesses information about the relationship you do not control. They interpret you. They adapt to you. They occasionally contradict your understanding of what you are doing. The capability develops through repeated encounters in which your model of people is forced to meet people.

res.p.000297Some professional skills have this character more than others. Judgment cannot be separated completely from decisions whose consequences might prove the judgment wrong. Negotiation eventually requires another party with interests that do not reorganize themselves around the framework you learned. Writing for readers requires readers. Difficult conversations become a skill only after someone responds in a way the script did not anticipate.

res.p.000298The model becomes yours when reality stops cooperating with your first attempt.

res.p.000299This changes how I think about preparation. Preparation is valuable, and civilization has done something very useful by preserving other people’s discoveries so that we do not each have to rediscover structural engineering through bridge collapse. There are legitimate gates around medicine, aviation, engineering and many other forms of work because some beginner mistakes are too expensive to make on live subjects. Formal instruction, simulation, supervised practice and credentials exist partly because reality does not need another opportunity to teach certain lessons through catastrophe.

res.p.000300But many professional gates are less absolute than the person standing before them imagines.

res.p.000301“I am not ready” often treats capability as binary when the real question is the size of the claim. You may not be ready to run a function. You may be ready to own the analysis that informs one of its decisions. You may not be ready to build an application. You may be able to automate one painful part of a workflow and learn from someone who understands the rest. You may not be ready to advise an executive team, but you may be ready to form a recommendation strong enough that your manager has to argue with it instead of rewriting it.

res.p.000302The legitimate first claim is often smaller than the identity we hope eventually to occupy.

res.p.000303This is useful because actual participation creates information that preparation cannot. The person considering a writing project can read another book about writing, and perhaps should. Eventually the unanswered question is no longer what writers say about writing. It is what happens when a reader encounters these sentences. Someone preparing for leadership can study conflict for years. Eventually they need to discover how their own nervous system behaves when another person is angry with them and a decision still has to be made.

res.p.000304Participation produces the next curriculum.

res.p.000305That is very different from glorifying risk. There is no prize for choosing the most difficult available version of a problem. A person raising children, caring for someone, managing illness, living with financial precarity or simply exhausted after a difficult season may quite rationally choose work they already know how to do. Stability is not a failure to develop. Some periods of life create room for expansion. Others require enough familiar ground that the person can remain intact.

res.p.000306A theory of growth that treats every stable season as cowardice has confused a human life with a venture portfolio.

res.p.000307The better question is what kind of problem the opportunity will repeatedly place in front of you.

res.p.000308Titles tell us something. Compensation tells us something. Organizational prestige, authority and scope matter. But a role also determines the problems that will become familiar through repetition. A larger title can make a person smaller if the new job consists mainly of repeating a narrow judgment they already know how to make. Organizations often promote people because they are exceptionally good at solving one class of failure and then arrange for that class of failure to remain near them permanently. The employee becomes more senior while the range of problems capable of teaching them contracts.

res.p.000309A less impressive move can do the opposite. A role near unfamiliar functions may require someone to understand customers, money, product and operations at the same time. A sideways move may place them close to a kind of decision they have never been allowed to observe. An ordinary project can expose a system whose behavior no framework had previously made concrete.

res.p.000310The question is not whether the opportunity looks developmental from the outside. It is what you will have to become able to do for the work to succeed.

res.p.000311That question also helps distinguish skills by the worlds they open. Some capabilities remain relatively local, and there is nothing wrong with that. Others alter the number of situations in which a person can meaningfully participate. Learning to write well changes more than the production of documents; it changes which ideas can be made precise, preserved and sent beyond the room in which they originated. Financial literacy changes which business arguments can be evaluated without borrowing somebody else’s conclusion. Enough technical fluency to automate recurring work changes the relationship between certain problems and time. Learning to conduct conflict without immediately defending against it changes how many disagreements can be addressed before avoidance has built an infrastructure around them.

res.p.000312Some skills are tools. Others alter the size of the reachable world.

res.p.000313Those are particularly valuable places to look for developmental problems because their residue spreads. The capability learned in one project changes the range of projects that can be understood afterward. It can also expose the limits of expertise more clearly. The more of the system becomes legible, the easier it is to see which parts still depend on borrowed judgment.

res.p.000314This brings the argument back to Response. Growth is not merely acquisition. A capability becomes real when it changes what can happen after reality reaches you. More information becomes usable. More kinds of feedback can be understood. A contradiction that once looked like noise can be recognized as a problem worth investigating. A decision that once had to be escalated can be made closer to where the relevant information lives.

res.p.000315The developmental problem earns its value by changing the next response.

res.p.000316When choosing among opportunities, then, I want the ordinary questions about money, people, stability and meaning. I also want another one: What will this work teach me because the work cannot be completed without learning it?

res.p.000317The answer should not be “endure more.”

res.p.000318It should name a capability.

res.p.000319And the problem should be close enough to reach, real enough to answer, and bounded enough that learning remains possible while the answer is still useful.

Chapter 9 — Let the Result Travel Backward

res.p.000320A great deal of work is organized as though completion were the end of the causal chain.

res.p.000321Marketing produces leads and hands them to sales. Sales produces customers and hands them to implementation. Implementation gets the customer live and hands the relationship into whatever comes next. Each function can become highly competent at moving its object across the boundary, and most measurement systems make that competence easy to see. Leads were generated. Opportunities were created. Contracts were signed. Projects launched. Tickets closed. The sequence advances.

res.p.000322Then a customer six months downstream reveals that the beginning was wrong.

res.p.000323Perhaps the customers who stay, expand and refer other buyers share a more specific problem than the audience marketing originally targeted. Perhaps the customers who require heroic implementation effort arrived through promises that sounded attractive during the sale and became expensive only after signature. Perhaps the prospects sales dismissed quickly were responding to language that attracted the wrong need in the first place. Once enough of those outcomes are visible, the later stages contain information about decisions made much earlier.

res.p.000324Whether the organization learns depends on whether that information is allowed to travel backward.

res.p.000325A funnel is a useful picture of attrition and a poor picture of learning. It shows how much survives each stage, which is often exactly what someone needs to know. Its geometry also encourages a particular imagination of responsibility. Work descends. Marketing owns the top, sales the middle, another group the bottom. A clean handoff allows each function to define success at the boundary where its direct control ends. Marketing hit the lead target. Sales closed the deal. Implementation delivered the contracted scope. Customer success managed the relationship. Everything can be true while the customer still wishes they had never bought.

res.p.000326The customer’s experience then becomes inconveniently cross-functional. It may contain evidence that the targeting was wrong, the promise was wrong, the product was wrong for that use, the implementation model was wrong for that customer, or several of those at once. Learning requires the outcome to cross boundaries in the opposite direction from the work that produced it. The end of the process has to earn some authority over the beginning.

res.p.000327The same principle operates far beyond customers.

res.p.000328An employee brings a problem to a leader earlier than strictly necessary. The project has not failed. Something that was expected to work has stopped working, the margin for recovery is shrinking, and the employee decides the uncertainty itself is worth surfacing. The leader is having a bad day. The project is visible to senior management. They respond with irritation, ask why the problem was not anticipated, enumerate several things the employee should already have considered and end by saying they never want to be surprised like this again.

res.p.000329The facts in the response may all be correct. The response still changes the next report.

res.p.000330The employee has learned something about when information becomes safe to transmit. Next time they wait. They investigate longer, try another workaround and bring the issue forward once the evidence is mature enough to survive questioning. The leader receives the problem later and, reasonably, concludes that people are not escalating soon enough. A weekly status report appears. Ambiguous conditions must now be compressed into green, yellow and red, and nobody wants to turn a visible project red while there remains a plausible route back to green. The organization is collecting more information and discovering problems later.

res.p.000331No one involved needs to be irrational. Each response is locally intelligible inside the conditions produced by the previous one. Early reporting was painful, so employees delay until they have stronger evidence. Delayed reporting frustrates leadership, so leadership creates tighter controls. Tighter controls increase the stakes attached to each report. Higher stakes encourage more careful management of what gets reported. The information becomes cleaner as it travels and less capable of surprising the people receiving it.

res.p.000332Fear can become a feedback loop.

res.p.000333The opposite loop is usually quieter. Someone brings the problem early. The leader does not have to enjoy hearing it, and the employee does not need ceremonial praise for having spoken. They work the problem. Perhaps the employee missed something obvious and hears that plainly. Perhaps the risk was genuinely difficult to foresee. Perhaps the plan changes while there are still several reasonable ways to change it. What survives the encounter is a piece of evidence about the system: bringing uncertain bad news forward while it is still actionable can improve the outcome.

res.p.000334The next warning arrives a little sooner. Earlier warnings permit smaller corrections. Smaller corrections make early escalation look increasingly sensible instead of alarmist. A leader who repeatedly sees someone exercise good judgment before a problem becomes undeniable can grant that person more autonomy. Autonomy creates opportunities for judgment to become visible under more consequential conditions. Trust grows from observed behavior rather than from a declaration that the culture values trust.

res.p.000335Eventually the person who once required approval may become someone whose judgment can substitute for approval.

res.p.000336There is no moment when the relationship steps onto the rung labeled TRUSTED. Repetition changes the structure between the people involved.

res.p.000337Careers look more ladder-like on a résumé than they feel while being lived. Titles can be written in chronological order. The forces producing the next opportunity are more recursive. Someone handles an ambiguous problem well, so another person remembers them when a harder one appears. The harder problem exposes their judgment to people who have never worked with them before. Those people now possess information about what the person is like under conditions no interview could reproduce. One of them later has a problem of their own. Work generates evidence, evidence changes trust, trust changes which work becomes available, and the new work produces another kind of evidence.

res.p.000338The later opportunity did not merely sit above the earlier one waiting to be climbed into. What happened earlier changed the probability that the later opportunity would exist at all.

res.p.000339Organizations sometimes construct this kind of circulation deliberately. The development history of Linux makes the structure unusually visible. The early kernel was incomplete, ran in a narrow set of environments and became available to other people while its limitations were still abundant. Different machines exposed assumptions its original environment did not. Other developers could inspect the source, attempt changes, report failures and contribute work. The contemporary development process is vastly more formal, but the same broad circulation remains: code moves outward, review and failure produce information, accepted work receives broader exposure, and broader exposure produces another class of information about the work.

res.p.000340At sufficient scale, no single person can inspect everything. Judgment itself becomes distributed. Maintainers develop reputations for the quality of what they accept. Prior judgment affects how future contributions move through the system. Trust allows authority to be distributed, distributed authority allows more work to receive meaningful review, and the expanded system creates more opportunities for defects and weak assumptions to become visible.

res.p.000341The interesting feature is not that Linux contains feedback. Almost every system does in some loose sense. The useful feature is that output remains capable of returning as information.

res.p.000342That distinction protects us from romanticizing loops.

res.p.000343A system can learn beautifully in the wrong direction. A marketing platform optimized for clicks can become extraordinarily good at finding people who click. The system observes who responds, searches for more people who resemble them, identifies creative associated with the response and allocates more exposure toward what appears to work. Every cycle sharpens the pattern. Revenue can remain entirely unimpressed.

res.p.000344The loop is functioning.

res.p.000345So is the organization in which a leader promotes people who understand what the leader wants to hear. Those people learn which arguments receive approval and bring more of them. Meetings become smoother. Objections diminish. Agreement increases. The organization experiences growing coherence while the people carrying incompatible information either translate it into acceptable language or stop offering it. Eventually the ease of reaching consensus becomes evidence that the strategy is obvious.

res.p.000346A loop does not require truth. Recurrence is enough.

res.p.000347The same fact should make us careful about the loops inside our own minds. Expect rejection, approach someone defensively, receive a guarded response and the response confirms the expectation. Believe that nobody else will do something properly, inspect every detail of their work, make independent action slow and unpleasant, and eventually people stop acting without you. Their dependence becomes evidence that your supervision was necessary. Enter a familiar disagreement expecting hostility and the preparations for hostility can help produce exactly the tone you expected to encounter.

res.p.000348Feedback is only as good as the independence of whatever is feeding back.

res.p.000349A customer matters partly because the customer can leave. An employee’s disagreement can teach a leader something partly because the leader did not author the disagreement. A reader can reveal that a sentence is unclear because the reader has no obligation to understand the sentence in the way the writer intended. Code encounters hardware the developer does not control. A financial result eventually places limits on how creatively a business can describe the success of an investment. The system learns when something outside its preferred explanation retains the ability to answer no.

res.p.000350That answer then has to reach someone capable of changing the next attempt.

res.p.000351Organizations are frequently rich in feedback and poor in loops. Customer surveys live in one platform. Support complaints live in another. Sales objections become CRM fields. Product usage enters analytics. Employee sentiment arrives through surveys. Churn reasons appear in reports after the customer is gone. Financial outcomes arrive on a cadence determined partly by accounting rather than by the decisions that created them. An organization can become extraordinarily sophisticated at collecting all of this information while leaving the activities that produced it largely untouched.

res.p.000352The information exists. Nothing has closed.

res.p.000353Closing the loop tends to be harder because the consequence has to cross an ownership boundary. If poor retention suggests that marketing is attracting the wrong customers, acquisition can no longer define its success entirely at acquisition. If implementation repeatedly discovers commitments that cannot be delivered efficiently, the quality of the sale extends beyond the signature. If employee turnover contains information about a leader’s behavior, the people data has crossed into a domain hierarchy may prefer to describe as management judgment.

res.p.000354Downstream reality acquires the right to revise an upstream decision.

res.p.000355That right cannot become automatic blame. A customer failing does not prove the sale should never have happened. An employee leaving does not prove the manager mistreated them. A product defect does not establish that an engineer was careless. Consequences contain randomness, competing causes and other people’s agency. Treat every bad outcome as a verdict and people will spend their energy contesting jurisdiction instead of examining the signal.

res.p.000356This is one of the easiest ways to destroy a feedback loop after paying to build it. Sales explains why implementation caused the customer failure. Implementation documents why sales made an unreasonable promise. Marketing refines attribution until the customer entering the funnel falls outside its responsibility. Customer success learns to code churn reasons in language that cannot become an accusation. The data becomes increasingly defensible and progressively less useful.

res.p.000357Blame asks who should carry the outcome. Learning asks what the outcome should be allowed to change.

res.p.000358Those are sometimes the same question. Often they are not.

res.p.000359A mature loop needs enough accountability that consequences cannot simply be exported and enough psychological room that receiving information does not require accepting a total theory of fault. If a sales team repeatedly wins customers who fail for the same reason, that pattern should alter how the organization sells even when nobody can prove that any individual deal was obviously wrong at the moment of signature. If a leader loses several strong employees under similar conditions, the recurrence deserves examination even if each departure has its own legitimate explanation. Patterns can modify future behavior before a courtroom standard of causality has been reached.

res.p.000360This returns us to the practical horizon of Response. The value of feedback depends partly on how much of the next decision still remains changeable when the feedback arrives. A customer problem that reaches product after ten more customers have encountered it has taught the organization something, but late. A campaign reviewed only after the quarter has spent most of its budget leaves fewer possible responses than the same information in week two. An employee warning that reaches a leader while the plan is still provisional has a different economic and social value from the same warning after everyone has publicly committed to the date.

res.p.000361Delay increases the amount of reality that must be undone before reality can be answered.

res.p.000362Distance does something similar. A support agent may see a recurring customer misunderstanding months before the pattern becomes large enough to appear in an executive report. A BDR may hear the market reject a phrase long before campaign performance reaches statistical significance. An implementation team may know which promises generate downstream work while leadership still sees the deals primarily as bookings. The people closest to the consequence often possess lower-status information: anecdotal, incomplete, difficult to aggregate, occasionally wrong.

res.p.000363Waiting for that information to become clean can destroy much of its value.

res.p.000364The solution is not to elevate every anecdote into strategy. Noise remains noise. The organization needs ways to distinguish a local irritation from a recurring structure. It also needs routes by which provisional information can travel before it has become undeniable. That is a different design problem from reporting. Reporting asks how to describe what has happened. A learning system asks how early consequences can begin revising the choices that make the next consequence more likely.

res.p.000365I find one question useful after almost any recurring piece of work: what did this make more likely next?

res.p.000366A meeting can solve today’s coordination problem while teaching everyone that cross-functional decisions require that meeting. Or it can expose the information gap the meeting was compensating for and make some future meetings unnecessary. Solving an employee’s problem can restore progress while making it more likely that the employee brings the next problem upward in the same form. Or it can transfer enough judgment that the next version never reaches you. Winning a customer can improve the organization’s understanding of who succeeds and why. It can also create pressure to find another superficially similar customer before anyone learns what happened to the first.

res.p.000367The result is never only the thing that happened.

res.p.000368Every response changes, however slightly, the environment in which the next signal will arrive.

res.p.000369A good loop makes that change worth inheriting.

Chapter 10 — Who Has Editing Rights?

res.p.000370Calendars are strange documents. They look factual in a way that most accounts of a life do not. At three o’clock there is a meeting. At five-thirty there is a dentist appointment. Thursday morning is blocked. A child needs to be somewhere at seven. The rectangles make no argument about whether any of this is wise. They simply sit there, occupying the same finite surface.

res.p.000371What the rectangles conceal is how differently each one got there.

res.p.000372Somebody asked for an hour and I gave it to them. I promised another hour months ago and the promise has finally arrived to collect. Work took an afternoon because I accepted a responsibility that includes afternoons like this one. My child needs something from me because children have a notoriously weak respect for quarterly planning. A friend has an emergency. A recurring meeting persists because nobody has quite developed enough irritation to kill it. Something I badly want to accomplish has been squeezed into the white space between commitments that arrived with other people attached to them.

res.p.000373Seen that way, a calendar is less a schedule than a record of negotiated access.

res.p.000374The negotiation is not always explicit. Often it happened long before the invitation arrived. A boss asks whether I can move something, and the request contains more force than the same words from someone who reports to me. A customer has access because somewhere upstream we agreed that serving customers would occasionally disrupt plans. A child has extraordinary access because I chose a relationship in which somebody with almost no institutional power can legitimately reorganize an entire day by developing a fever.

res.p.000375The software treats these claims as though they were comparable. Accept. Tentative. Decline. Three little buttons, available to everyone.

res.p.000376Only the software thinks the buttons mean the same thing.

res.p.000377A junior employee invited to a “quick sync” by three vice presidents may technically possess the ability to decline. Their actual decision contains several other variables: how unusual the request is, whether the topic touches work they own, what declining will communicate, whether the organization means what it says about protected time, how the particular leaders involved interpret resistance, and whether the hour they are protecting will look sufficiently important to survive explanation. Someone senior enough may face none of those calculations. They decline because they are busy. Their busyness is presumed rather than litigated.

res.p.000378Power often appears in a calendar as the right to create inconvenience without experiencing the inconvenience as power.

res.p.000379A senior leader asks if a meeting can move to three. Six people say yes. It was a reasonable request and nobody behaved badly. One person moves a doctor’s appointment, another loses the only uninterrupted hour in the afternoon, somebody else does the work after dinner. None of this is visible from the leader’s side of the exchange. They asked. Everyone could make it work.

res.p.000380After enough repetitions, the asymmetry becomes part of the environment. Some people’s time is treated as terrain; other people’s time is treated as weather.

res.p.000381I have become more aware of this as I have gained authority. Early in a career, the asymmetry is hard to miss because you are usually standing on the receiving end of it. Later, people start accommodating you before you have experienced yourself as asking for accommodation. A calendar opens because someone moved something. A meeting becomes possible because another person handles the collision. You begin to feel unusually disciplined about your schedule without noticing that some of the discipline is being performed elsewhere.

res.p.000382This does not make protecting time illegitimate. There is work that disappears if nobody protects a place for it. Strategy rarely arrives with an Outlook invitation demanding immediate attention. Thinking loses routinely to things with participants. The work that matters most over a year can be astonishingly weak in a contest with the work that matters most over the next twenty minutes.

res.p.000383Senior people often have to refuse access precisely because their authority attracts claims. The trouble begins when the ability to refuse becomes a morality tale. Someone blocks every morning for deep work and concludes that people who lack deep work simply need stronger boundaries. Perhaps. Or perhaps their own boundary survives because an assistant reorganizes requests, a deputy takes early calls, their title makes a decline cheap, and nobody invites them to meetings whose value is doubtful simply because declining them would look politically strange.

res.p.000384The calendar does reveal priorities, but only after power, money, responsibility and obligation have had their say.

res.p.000385A nurse on a twelve-hour shift does not need a lecture from an executive about protecting mornings. Someone working two jobs because one does not cover rent is not revealing a hidden value system through a lack of white space. A person caring for a sick parent may have exactly the schedule their values demand and still desperately wish the schedule were different. In all of these cases, the allocation is meaningful, but the meaning cannot be recovered by pretending allocation and preference are identical.

res.p.000386There are also constraints we chose and later experience as constraints anyway.

res.p.000387That category interests me more because it is easy to tell the story badly. Parenting narrows the set of things I can do at particular times. So do friendships, commitments, jobs, partnerships, debts, teams and promises. The language of personal freedom encourages us to notice the narrowing and forget that some of it is the physical shape taken by things we value.

res.p.000388A promise is a peculiar use of time because it deliberately gives part of the future away before the future arrives.

res.p.000389“I’ll be there Thursday” means that Thursday-me does not receive a fresh vote equal to the one I am casting now. If showing up remains optional according to whatever I happen to want on Thursday, I did not make much of a promise. I predicted.

res.p.000390I am suspicious of accounts of boundaries that treat every external claim as an intrusion. Other people are inconvenient by nature. They become sick at bad times. They need conversations whose duration cannot be known in advance. They occasionally require help on the night when I wanted to be left alone. If the goal of boundaries is a life in which nobody can successfully interrupt my preferred allocation of attention, then the project has overshot autonomy and arrived somewhere much lonelier.

res.p.000391There are claims on my time I want to be strong enough to defeat my convenience.

res.p.000392There are others I have allowed to acquire that strength without ever deciding that they should.

res.p.000393The calendar is useful partly because both leave the same kind of trace.

res.p.000394A recurring commitment can begin as a choice and gradually lose any living relationship to the reason it was created. A meeting once existed because information could not move another way. The system changes, the constituency changes, the original problem disappears, and the meeting remains. Nobody particularly wants it. Nobody feels individually entitled to remove it. Twenty people keep attending because the cost of one person questioning its existence feels larger than another hour.

res.p.000395I have had a version of this happen with a cross-functional call. It began under one leadership regime partly as a way to route around micromanagement. People needed a place where work could become coordinated without every interaction being pulled upward. By the time someone finally asked, about six months later, whether we still needed the meeting, the original reason had largely disappeared.

res.p.000396The answer was still yes.

res.p.000397But the constituency had changed, and so had the function. What had started partly as a shadow process had become useful for reducing communication gaps among people who otherwise would not have had a reliable place to see one another’s work. Keeping the meeting was not inertia once we could name the new reason. We had examined the claim and chosen it again.

res.p.000398That difference is enormous.

res.p.000399A recurring obligation does not become illegitimate merely because its origin has expired. Nor does longevity earn it permanent rights. The important change happened when the meeting stopped surviving automatically and had to present itself again for judgment.

res.p.000400Calendars accumulate these little inheritances. Wednesday at eleven belongs to a decision made eighteen months ago. Tuesday night belongs to a season of life that was supposed to last one year and somehow entered its fourth. An annual commitment becomes monthly work through increments nobody would have accepted as a package. Travel described during hiring as twenty percent becomes a collection of four-thirty alarms, missed dinners and airport meals that feels quite different from the percentage.

res.p.000401The future receives simultaneously what the past accepted separately.

res.p.000402That is how people end up staring at weeks they mostly constructed and feeling that somebody else constructed them.

res.p.000403Sometimes that feeling is dishonest. I can blame work for evenings I volunteered because I wanted the recognition attached to the work. Ambition is good at wearing the clothes of obligation. “I have to” allows me to enjoy the status or opportunity produced by a choice without acknowledging that I chose the trade.

res.p.000404At other times, “I chose this” is the dishonest sentence. Economic pressure can transform formal choice into something much narrower. An employer can describe overtime as optional while consistently promoting the people who provide it. A networking event can be technically voluntary and practically important enough that absence has consequences. Vacation can be encouraged in policy while everybody who matters remains reachable from the beach.

res.p.000405The language around time is full of these small evasions because the truth is frequently mixed. I chose the job, but I did not choose every condition of the labor market in which I needed a job. I chose to become a parent, but I did not choose tonight’s emergency. I chose the ambitious project, and I also underestimated what it would require. I did not choose the meeting, though perhaps I have enough authority to stop attending and prefer not to absorb the social cost.

res.p.000406Clean categories are attractive precisely where reality is mixed.

res.p.000407Nineteenth-century factory labor made some questions of control less ambiguous. Women working in the textile mills of Lowell, Massachusetts, had opportunities for wages and independence that were significant in the world available to them, while also working extraordinarily long days structured by bells and company rules that extended into boardinghouse life. By the 1840s, women in Lowell were organizing for a ten-hour day and petitioning the Massachusetts legislature. Their argument was not merely that shorter hours would make them more productive. They were contesting how much of a human life an employer should be permitted to claim.

res.p.000408The distance between that world and modern salaried office work should remain distance. A professional annoyed by a crowded Outlook calendar has not rediscovered the industrial labor conditions of the nineteenth century. The historical case is useful because the factory bell made control explicit. The bell rang and somebody else’s authority over the hour became audible.

res.p.000409Professional work can distribute similar questions across softer mechanisms. There may be no bell, no formal rule, sometimes not even an explicit request. People notice who answers at night. They notice whose vacations are genuinely vacations. They notice whether declining a meeting requires explanation, whether “flexible” means the company flexes around the employee or the employee flexes around the company, and whether someone who becomes unreachable for a week returns to admiration or suspicion.

res.p.000410Culture can claim time without putting the claim in writing.

res.p.000411Money can buy resistance to some of these claims. Enough savings changes the meaning of an intolerable boss because leaving can happen before another employer grants permission. Childcare can return hours that would otherwise have to be supplied personally. Paying someone to repair the furnace can turn a Saturday into something else. A shorter commute may cost more money and return time.

res.p.000412Money cannot buy time in the literal sense. Everyone gets the same twenty-four hours. It can alter the number of hours over which someone else has practical authority.

res.p.000413Advice about time often smuggles class position into character. A person who can outsource an obligation has solved a different problem from the person who still has to perform it. The first person’s calendar may become cleaner because the second person’s calendar became more crowded.

res.p.000414The transfer is not automatically unjust. People trade labor for money because both can benefit. I care about noticing the transfer because apparent personal efficiency often has a system around it. Somebody cleans the room, watches the child, takes the first call, covers the shift, prepares the material, drives the package, answers the customer or absorbs the scheduling conflict. What appears as a beautifully defended hour from one position in the system may appear as paid work from another.

res.p.000415Once you start seeing calendars this way, they become surprisingly social documents.

res.p.000416Mine contains the people to whom I have made commitments and the institutions with enough authority to make claims. It contains traces of how much financial flexibility I have, what I am trying to become, what I am afraid of losing, and which relationships I permit to interrupt me. It also contains accidents, seasonal obligations, badly estimated projects and things that simply have to get done.

res.p.000417No honest reading can reduce all of that to “these are your priorities.”

res.p.000418Still, patterns become difficult to explain forever.

res.p.000419If I say that something is central to my life and it loses every conflict for years, eventually the word central needs work. Perhaps the circumstances really have been extraordinary for years. Perhaps the sacrifice is conscious and temporary on a longer horizon than anyone else can see. Perhaps I have almost no control over the relevant time. Those are all possible.

res.p.000420Perhaps I also like believing that I value something more than I like paying the costs required to value it in practice.

res.p.000421The calendar cannot decide among those explanations. It can prevent me from pretending there is nothing to explain.

res.p.000422That is enough.

res.p.000423I do not need the calendar to become a moral instrument. I already have plenty of ways to feel inadequate. I need it to preserve a record that memory cannot rewrite quite as easily. There was an hour. Something got it. Then another hour. Then another.

res.p.000424If the same claimant keeps winning, I can become curious about the arrangement.

res.p.000425Sometimes the arrangement should change. The standing meeting should end. The boss has acquired too much informal access. Work has been living inside a supposedly temporary exception. I keep volunteering for obligations whose resentment I later blame on the people receiving the help.

res.p.000426Sometimes the calendar is telling me almost the opposite. I am tired because I am caring for people I love, keeping promises I still endorse and doing work I deliberately chose in a difficult season. There may be changes available at the margins, but no scheduling system is going to reveal that an integrated life contains zero competing obligations. I do not need to resent those claims merely because they make the week harder.

res.p.000427The point is to know what kind of difficulty I am looking at.

res.p.000428There is difficulty produced by a claim that should no longer have authority. There is difficulty produced by a claim I never meaningfully authorized. There is difficulty produced by structural conditions I cannot currently change. And there is difficulty produced by belonging to other people in ways I still consider worth the cost.

res.p.000429On a screen, they are all rectangles.

res.p.000430In a life, they are not.

Chapter 11 — What the Receiver Has to Invent

res.p.000431One of the easiest ways to save time at work is to make somebody else finish your thought.

res.p.000432The transfer can be almost invisible. I have a document open, know why I am looking at it, know which part bothers me, know whose judgment I want and roughly what I will do depending on the answer. I send the document with “Thoughts?” The message takes seconds. From my side, almost all of the relevant context remains available because I was just thinking it.

res.p.000433The person receiving it has none of that advantage.

res.p.000434They have the document and one word. They have to decide whether I want approval, criticism, copy editing, strategic disagreement, reassurance, a quick reaction or a recommendation I can act on. They have to decide how much time the request deserves, whether I have already formed an opinion, whether I expect a response now, and whether “thoughts” means that the underlying decision is still genuinely open. If I have more authority than they do, they also have to estimate the cost of guessing wrong.

res.p.000435The missing work did not disappear when I omitted it. It changed heads.

res.p.000436Language always depends on omission, of course. Human communication would become impossible if every sentence had to contain everything necessary to derive its meaning from first principles. “The meeting moved to ten-thirty” works because sender and receiver share enough of a world that nobody needs a definition of meeting, an explanation of clocks or a clarification that the furniture has not moved. Communication is economical because context can be shared.

res.p.000437Professional ambiguity becomes expensive when we mistake context available to me for context shared between us.

res.p.000438I know why the customer complaint worries me, so forwarding the email to five people feels informative. One recipient thinks I am escalating it. Another assumes I want diagnosis. Someone else believes I am assigning action because their team appears in the thread. A fourth person sees the forward as courtesy and does nothing. If nothing happens, I may conclude that nobody took ownership. They may reasonably conclude that no ownership was ever transferred.

res.p.000439The grammar was fine.

res.p.000440What was missing was the transaction.

res.p.000441A surprising amount of workplace communication leaves the receiver trying to determine what kind of event has just occurred between two minds. Is this information or a request? Is the request for work or judgment? Has a decision been made? If not, who gets to make it? Is the sender asking me to improve their proposal or tell them whether there should be a proposal at all? Is this urgent, or did the sender happen to send it urgently?

res.p.000442“Can you take a look?” may contain almost none of that structure and still work perfectly between people who know one another well. Shared history fills the sentence. I know that when a particular colleague sends me a page and asks me to look, they want me to test the argument rather than touch the prose. They know that if I think the premise is wrong, I will say so even if they asked about something narrower. Years of successful exchanges have compressed a complicated protocol into ordinary speech.

res.p.000443That compression becomes less admirable when an organization depends on private decoding.

res.p.000444People who have worked with a leader for years often know things that are nowhere in the leader’s vocabulary. “Interesting” means the proposal is probably dead. “Can you sharpen this?” means the argument needs to be rebuilt, not that the slides need fewer words. “We should think about that” means the idea is not going anywhere unless somebody returns with evidence. “I’m not sure this is landing” means tomorrow’s presentation has just become tonight’s work.

res.p.000445A newcomer hears the literal sentence.

res.p.000446The experienced employee hears the organization.

res.p.000447Some of that is genuine context. Work among human beings will always contain tone, history and tacit knowledge that no onboarding guide can capture. The experienced person knows which objections are old, where previous attempts failed, which stakeholder owns a dependency that never made it into the process diagram and why a seemingly small request has larger consequences than its wording suggests. That knowledge is real.

res.p.000448Some of the apparent expertise, though, is compensation for somebody else’s failure to become intelligible.

res.p.000449This becomes visible when managers compare employees who have unequal access to their unstated model. Ask several people to “put together some thoughts on our webinar strategy” and you may receive entirely different artifacts. One person writes three bullets. Another builds a fifteen-slide presentation. Another investigates performance by audience and format, interviews sales and returns with a recommendation. A fourth asks what problem the manager is trying to solve before beginning.

res.p.000450The manager can experience those responses as differences in initiative or strategic ability. Sometimes they are. The employees may also have been solving four different assignments because the actual assignment never fully left the manager’s head.

res.p.000451A particularly capable employee often protects the manager from discovering this. They infer the audience, reconstruct the problem, choose the appropriate level of effort, locate data that was never mentioned and return something useful. The work succeeds, so the original request appears adequate. Over time the manager becomes accustomed to the quality of output produced by people who know how to complete the instruction.

res.p.000452Then somebody new arrives and appears strangely literal.

res.p.000453The new person asks when the work is needed. They ask who will see it. They want to know whether the manager is looking for options or a recommendation. The manager, who has become accustomed to colleagues supplying this structure invisibly, may experience the questions as evidence that the employee lacks autonomy.

res.p.000454An organization can accidentally define autonomy as knowing what your boss forgot to say.

res.p.000455The same problem appears between functions even when no individual is especially vague. A word acquires meaning locally and crosses a boundary intact while its meaning does not. Marketing says a campaign is “ready” because creative and configuration are complete. Legal hears “ready” and assumes review is complete. Sales hears “ready” and expects customers to see it tomorrow. Product says a feature is “committed,” meaning that the roadmap currently includes it, while somebody closer to a customer hears a promise.

res.p.000456“Forecast” can describe several different numbers depending on who is producing it and what uncertainty they believe belongs inside the number. “Qualified” can become a political argument disguised as a field definition. “Launch” may mean released to production, available to customers, publicly announced, actively promoted or merely technically capable of being turned on.

res.p.000457Organizations respond by building glossaries, templates, stage definitions and process documents. Sometimes that is exactly what is needed. Shared language is infrastructure. If five teams repeatedly collide because “ready” means five things, defining the word can save months of correspondence over the life of the system.

res.p.000458No glossary can decide what the sender has not decided.

res.p.000459A standardized status marked “yellow” still leaves work behind if the message accompanying it says only “let’s connect.” The receiver has to determine whether they are being informed, asked to help, asked to approve a recovery plan or quietly summoned because the project is closer to “red” than the field allows anyone to say.

res.p.000460Digital communication makes these transfers unusually easy because the cost of transmission has become so small. A thought can leave my head before it has become a request. Forwarding an email takes less effort than deciding why another person needs to see it. Mentioning somebody in a Slack thread is nearly free. “Adding Maya for visibility” can bring Maya into a conversation while leaving Maya to decide whether visibility means awareness, accountability, consultation or future blame if the issue goes badly.

res.p.000461The thread grows from there.

res.p.000462A customer asks whether a feature will be available in June. Someone says engineering is targeting June. Another person asks whether that means June can go into the contract. Engineering says they would rather not commit until final scope is known. Someone asks whether the customer needs the entire feature or one component. Sales joins late, sees several references to June and asks whether the date has been confirmed.

res.p.000463Each contribution can be locally reasonable. Nobody sat down to write an incoherent twenty-message exchange. The incoherence was generated collaboratively, one unresolved distinction at a time.

res.p.000464By the end, the organization may contain several versions of the same fact. One person remembers June. Another remembers targeting June. Someone else correctly remembers that engineering explicitly refused to commit to June. The date enters a slide without its qualifier. The slide travels farther than the thread. A month later, somebody asks why the team moved a date that nobody technically promised.

res.p.000465Ambiguity has become state.

res.p.000466Slack makes this especially strange because it occupies a space somewhere between speech and record. People write with the casualness of conversation into a medium capable of preserving the sentence indefinitely. “We should be able to do that” feels provisional when typed and becomes remarkably solid when screenshotted six months later. “June seems realistic” loses the uncertainty supplied by the surrounding conversation. A thumbs-up reaction can become evidence of approval because nobody wrote anything more explicit.

res.p.000467Turning every workplace conversation into contractual language would be miserable. Teams need rooms in which ideas can remain incomplete. People have to be able to think aloud, suggest something they may abandon ten minutes later and discuss uncertain dates without creating commitments merely by mentioning them.

res.p.000468The language should be allowed to be provisional when the underlying state is provisional.

res.p.000469That is different from allowing certainty to leak into the sentence because uncertainty feels embarrassing.

res.p.000470A customer asks when something will be available. “I don’t know yet” feels weak, so we choose a date that sounds plausible and surround it with enough soft language to preserve an exit. A senior leader asks whether the plan will work. Nobody knows yet, but saying so feels like failure to prepare, so the room converges on confidence slightly faster than the evidence does. Someone proposes a direction we dislike, and instead of disagreeing we say we are “not fully aligned,” preserving harmony while leaving everyone to determine whether the objection is substantive enough to stop anything.

res.p.000471The discomfort avoided by the sender does not vanish either.

res.p.000472Vagueness can distribute emotion as efficiently as it distributes cognitive work.

res.p.000473A clean no creates an unpleasant moment now. A vague maybe creates several smaller moments later, usually for somebody else. “Let’s revisit this next quarter” may be a genuine timing decision. It may also be a refusal whose sender would rather not experience themselves refusing. The recipient keeps the possibility alive, schedules a reminder, preserves capacity or delays another route because the answer was never sufficiently closed to act against.

res.p.000474“I’ll try” may be an accurate description of uncertainty. It may mean no spoken by a person who wants the receiver to perform the disappointment gradually.

res.p.000475Unclosed possibilities occupy space.

res.p.000476They remain in project trackers and inboxes. Somebody follows up. Somebody waits. Someone keeps a contingency plan alive. The sender bought a cleaner interaction by leaving another person responsible for discovering when the possibility had actually died.

res.p.000477Organizations often regard this as a communication problem only when the administrative residue becomes visible. There are too many meetings. Email chains are too long. Decisions keep getting revisited. People complain that they lack ownership. Another tool is purchased to establish a single source of truth.

res.p.000478Sometimes the source was never the problem.

res.p.000479If the human transaction remained unresolved, the project system faithfully preserves the unresolved transaction.

res.p.000480The aviation world offers a deliberately extreme contrast because certain kinds of ambiguity become dangerous too quickly to romanticize. Air traffic control cannot function if pilot and controller leave an exchange with neighboring interpretations of an altitude or heading. The FAA maintains standardized terminology, and the control system makes acknowledgment part of communication rather than treating transmission as sufficient. When pilots read back an instruction, controllers listen for whether the instruction that returned is the one they meant to send.

res.p.000481Office work is not aviation. Nobody needs a formal readback after I ask for edits to a presentation, and importing safety-critical protocol wholesale into ordinary collaboration would be unbearable. The useful thing exposed by the comparison is much smaller: saying what I meant and successfully communicating it are different events.

res.p.000482The sender controls only the first.

res.p.000483This should not become a theory in which every misunderstanding proves that the sender failed. Receivers bring their own assumptions, distractions, fears and incentives. People can misread excellent writing. They can hear accusations nobody made, miss information plainly stated and confidently remember sentences that were never spoken. Language cannot eliminate the fact that another mind remains another mind.

res.p.000484What the sender can do is expose the distinctions whose loss would create consequential divergence.

res.p.000485If I need a decision, the decision should be locatable. If I am supplying information only, the receiver should be able to remain merely informed without worrying that silence will later be interpreted as neglect. If the date is uncertain, the uncertainty should survive transmission. If I want someone to review the argument rather than polish the wording, a few words can keep three hours of copy editing from becoming the wrong kind of contribution. If responsibility is changing hands, both people should be able to identify roughly when that happened.

res.p.000486Ordinary language is entirely capable of this.

res.p.000487“For visibility only; nothing needed from you.”

res.p.000488“I recommend B. I need your decision by Thursday because procurement starts Friday.”

res.p.000489“The customer wants June. Engineering thinks June is plausible but has not committed to it. Please don’t promise the date yet.”

res.p.000490“I don’t know whether this changes launch. I should know tomorrow and wanted you to have the risk now rather than the answer later.”

res.p.000491None of those sentences is impressive. They simply contain enough of the transaction that another person can enter at the point where the sender actually is.

res.p.000492Sometimes writing the clean sentence is hard because the sentence exposes a social fact we were using language to blur. “I need your approval” admits that authority resides somewhere else. “I’ve decided” closes a discussion that we may prefer to keep cosmetically collaborative. “I disagree” creates a sharper moment than a paragraph about alignment. “I missed the deadline” attaches the problem to a speaker more clearly than “we’ve had some timing challenges.”

res.p.000493A large amount of corporate language has evolved to manage precisely these little abrasions.

res.p.000494Some of that softening is humane. Human beings are not packet-switched networks, and maximum explicitness is not the same thing as good communication. A person receiving terrible news may deserve time and gentleness rather than the shortest sentence capable of conveying the fact. Criticism may belong in private even when stating it publicly would be more efficient. Confidentiality can make omission an obligation. Negotiations sometimes need room before positions harden. Early thinking may genuinely be too uncertain to compress without distortion.

res.p.000495Clarity is compatible with tact because clarity concerns the shape of the reality being shared, not the violence with which we throw it at another person.

res.p.000496Brevity has no special claim on that virtue. “Fix?” beneath a screenshot is admirably short and may produce an hour of work determining whether the sender wants the visible error corrected, the process investigated, the owner identified, or an explanation of how the mistake escaped review. The message saved characters, not labor.

res.p.000497The opposite extreme merely hides the center differently. An email can contain every relevant fact, every caveat and four years of historical context while leaving the requested decision somewhere in paragraph eleven. Nothing is missing except the hierarchy of the information. The receiver opens the wall of text, understands that it deserves concentration, postpones reading until enough concentration becomes available and eventually receives a follow-up asking whether they saw it.

res.p.000498A clear message has proportions.

res.p.000499The proportions depend on the work. A difficult board recommendation may need twenty pages. A friend may need six words. Complexity is allowed to remain complex. The sender’s job is not to compress every subject into a slogan but to give the receiver some way to understand what governs the rest.

res.p.000500That is partly respect for another person’s attention. Every communication asks another mind to load context that currently lives somewhere else. Sometimes loading that context is exactly the collaborative work. I cannot send somebody three bullets and expect them to understand a market, a marriage, a strategy or a moral problem that actually requires immersion.

res.p.000501But when I already know the distinction that would organize the receiver’s effort—decision or discussion, committed or possible, action or information, urgent or merely important—and omit it because finishing my own thought would take another minute, I have made a small allocation decision.

res.p.000502I have decided whose mind will contain the unfinished work.

res.p.000503Power affects that allocation just as it affected the calendar. A vague message from a peer can often be returned with a casual question. A vague message from somebody several layers above you may reorganize the afternoon before anyone dares discover that no action was required. The sender may experience themselves as approachable and concise while the organization has learned to treat every fragment of their attention as potentially consequential.

res.p.000504Leaders therefore need information about the wake their language creates.

res.p.000505If “Could you look at this?” repeatedly produces polished work you did not request, the conclusion need not be that everyone around you overthinks. If people consistently hear tentative observations as decisions, the hierarchy may be supplying certainty your words did not intend. If a casual comment becomes a project, your authority has become part of the sentence whether you put it there or not.

res.p.000506People close to power listen for implications because implications have consequences.

res.p.000507The responsibility is not to sterilize speech until nobody can mistake a thought for an instruction. Leaders should be allowed to wonder aloud. They should also become interested in what the room has learned to do with their wondering.

res.p.000508The best communication cultures I know are not necessarily quieter because people write less. They are quieter because fewer messages generate interpretive work that has to be resolved somewhere else. Fewer side conversations exist solely to decode the main conversation. Fewer meetings begin by discovering why the meeting exists. Fewer people work under incompatible assumptions that only become visible after the work has diverged.

res.p.000509Disagreement becomes cleaner too. It is much easier to argue with the thing another person actually proposed than with a reconstruction of what their message probably meant.

res.p.000510None of this will rescue us from email or Slack. “Quick question” will continue introducing questions that are neither quick nor singular. People will forward threads without explanation. Somebody will write “gentle reminder” beneath a reminder whose gentleness is open to interpretation. Language remains a human technology and will continue carrying everything human beings bring to it.

res.p.000511The standard can remain ordinary.

res.p.000512Before I send the message, has enough of the thought actually left my head that the next person can begin with the work rather than with me?

res.p.000513When it has, a conversation may still be necessary. The problem may still be difficult. We may disagree completely.

res.p.000514At least we can disagree about the same thing.

Chapter 12 — What the Meeting Is Carrying

res.p.000515There is a kind of meeting that everybody complains about and nobody can quite cancel.

res.p.000516It has capable people in it. The agenda is real. Decisions get made. Without the meeting, something would probably go wrong. Sales needs an answer from operations. A regional team needs an exception from a central process. Finance needs to understand a commitment before releasing money. Product and engineering have arrived at different conclusions about what can responsibly be promised. None of the participants is wasting time in the ordinary sense. They are resolving problems the organization genuinely has.

res.p.000517Then the same people come back next week and resolve approximately the same problem again.

res.p.000518The customer changes. The campaign changes. The dollar amount changes. The nouns move around enough to keep the work feeling fresh, but the argument is strangely durable. One function wants speed and another needs control. One owns a revenue number and another inherits the work required to produce it. Someone has nominal authority but knows another group can reverse the decision later. A process says one thing while commercial reality keeps rewarding people for violating it. The meeting is where these local logics meet long enough for somebody to construct a temporary peace.

res.p.000519People can become exceptionally good at this. They learn what concessions are possible, which objections are ceremonial and which are real, who needs to hear about an exception before the formal request arrives, what finance actually means by unlikely, what engineering means by possible, and how far a commercial commitment can travel before it becomes dangerous. The recurring meeting begins accumulating its own expertise. New people watch the veterans navigate it and conclude that this must simply be what cross-functional work looks like.

res.p.000520Sometimes it is.

res.p.000521Organizations contain legitimate competing interests. Sales should care about the customer in ways legal does not. Security should notice risks that product would rationally tolerate in pursuit of ease. A central brand function and a regional marketer can both be correct about different consequences of the same decision. Finance is supposed to place constraints on desires other people sincerely believe are worth funding. A company in which all of these functions agree instantly may have achieved a remarkable degree of integration, or it may have arranged matters so that one interest no longer has meaningful representation.

res.p.000522The existence of tension tells us very little by itself.

res.p.000523What becomes interesting is whether repeated exposure to the tension changes anything about the system that produces it.

res.p.000524A meeting can resolve a real exception. A customer is genuinely unusual. A regulation creates a condition nobody anticipated. A technical limitation collides with an opportunity important enough to justify a one-time trade. Five people need to bring different kinds of knowledge into the room because no policy could have settled the case in advance without becoming absurdly rigid. The meeting takes an hour, judgment happens, the decision is made, and the people involved return to work.

res.p.000525There is nothing inefficient about that hour merely because several calendars were involved.

res.p.000526Another meeting has the same outward form but a different history. Five people assemble because no one knows which of them can safely decide. The person who appears to own the decision believes another function has veto power. That function insists it is only advisory but has reversed similar decisions before. Nobody wants to discover the true boundary by making a choice and being publicly corrected, so everyone whose later objection could matter gets invited in advance.

res.p.000527The attendance list becomes a diagram of uncertainty about authority.

res.p.000528A large meeting can look collaborative for that reason. Everyone is included. Nobody can claim they were ignored. Consensus is patiently sought. The organization may even regard this as evidence of a healthy culture. Beneath the civility is a simpler fact: nobody knows whose yes is sufficient.

res.p.000529Other meetings form because authority and consequence have been separated. Someone can make the decision, but somebody else has to live with what the decision produces. A sales team is rewarded when a contract is signed; an implementation team inherits the complexity of what was promised. A business unit decides how quickly it wants something; a shared-service team absorbs the capacity collision created by the date. The person choosing the path experiences one set of incentives, while the person paying for the path experiences another. Neither side is behaving irrationally. Their recurring meeting functions as an improvised constitution, giving the downstream party influence that the formal allocation of authority failed to provide.

res.p.000530Measurement can do the same thing without anyone disagreeing about a single fact. One group owns volume, another quality. One records the event at signature, another at revenue recognition. One is rewarded for what enters a system, another for what survives six months later. Their dashboards are all accurate. The conflict arrives because the organization talks as though locally correct measures describe one shared object.

res.p.000531People spend hours reconciling numbers that were never actually answering the same question.

res.p.000532Once a meeting has become the place where these discrepancies are repaired, attacking the meeting itself can make the organization worse. Shorten it from an hour to thirty minutes and people simply negotiate faster. Require a pre-read and the contradiction arrives better formatted. Introduce a disciplined agenda and the unresolved authority gap now has action items. Cancel the meeting entirely and the work does not politely vanish. It moves into email, private messages, spreadsheets and hallway negotiations. Decisions rise upward because someone eventually needs enough authority to end the ambiguity. Informal brokers become more valuable because they know how to get through a system whose formal paths do not quite work.

res.p.000533The calendar was showing where the work surfaced, not necessarily where it was created.

res.p.000534Experienced organizations can become difficult to simplify. They have often built remarkably capable human interfaces around structural mismatches. Remove the person who has spent ten years reconciling finance and operations and everybody discovers that two systems thought to be integrated were connected largely through one person’s memory. Eliminate the weekly escalation meeting and teams discover that no ordinary decision right exists for the cases the meeting had been absorbing. Fire the spreadsheet and learn that the spreadsheet was quietly making two databases agree enough for the business to function.

res.p.000535The compensating mechanism has been present for so long that it looks like part of the underlying system.

res.p.000536Leaders can become compensating mechanisms too.

res.p.000537A difficult cross-functional issue enters a room and remains unresolved until a senior person joins. The leader hears both sides, understands enough of each function to translate, and possesses enough authority to make the trade. What would have taken two teams several days collapses in fifteen minutes. Everyone leaves relieved. The leader was useful.

res.p.000538The next difficult issue therefore goes to the leader sooner.

res.p.000539That is rational. Why spend three days negotiating if a person upstairs can settle the matter before lunch? Over time, functions lose incentive to develop the authority or shared understanding necessary to resolve certain problems themselves because escalation works better. The leader’s calendar fills with evidence that the organization requires the leader’s presence.

res.p.000540The evidence is real. The dependence may still have been trained into the system by the reliability of the rescue.

res.p.000541This is not an argument for leaders withholding judgment in order to teach everyone a lesson. Sometimes the senior person should make the decision because making consequential decisions is part of the reason senior roles exist. The question appears across repetition. Is the leader repeatedly supplying judgment that properly belongs at that level, or repeatedly substituting for a missing boundary the organization could actually establish?

res.p.000542Routine dependence often hides beneath flattering language. The executive is a great integrator. The manager sees across silos. The chief of staff knows how to get things done. The long-tenured operator understands all the personalities. These can be valuable capabilities. They can also describe people whose competence has become load-bearing around conditions nobody has repaired.

res.p.000543Meetings provide a good habitat for this because solving the immediate case creates genuine relief. A decision has been made. Work can continue. Everyone was heard. The emotional experience of progress is real enough that the absence of structural progress can remain unnoticed.

res.p.000544Some meetings conceal something harder than an authority gap.

res.p.000545Everybody in the room may understand the underlying problem and also understand that naming it would change the kind of conversation they are having.

res.p.000546A project has been underfunded from the beginning. The team meets every week to discuss sequencing, dependencies, productivity and delivery risk because discussing whether the promised outcome can still be achieved with the available resources would require the executive sponsor to revisit a public commitment. Problems that originate in the commitment are translated into execution problems because execution problems can be assigned to the people already in the meeting.

res.p.000547A senior employee is damaging a team but has enough status or sponsorship that confronting the behavior carries unusual risk. Meetings appear around communication norms, workflow, team health and roles. Each intervention may improve something. None needs to contain the sentence everyone privately believes would explain why so much intervention is necessary.

res.p.000548Two executives disagree about strategy. Their disagreement enters the organization one level down as a coordination problem. Teams are asked to align plans whose incompatibility originates above them. Meetings multiply. Smart people become extremely skilled at making both directions appear simultaneously possible.

res.p.000549The organization starts doing process work around a political fact.

res.p.000550This kind of meeting can be sophisticated. People are not necessarily cowards. Naming the source may genuinely carry consequences, and employees do not cease being rational merely because organizational theory would prefer perfect candor. There are situations in which a junior person should not be expected to solve a conflict that belongs to two executives. The important thing is what the organization believes the meeting is accomplishing. If everyone is repeatedly managing symptoms because the causal variable lies outside their authority, better facilitation cannot close the gap between the stated problem and the real one.

res.p.000551The room can become a highly civilized place to avoid the conflict that gives the room its agenda.

res.p.000552Consensus complicates this further because meetings produce records as well as decisions.

res.p.000553A group can enter with genuine disagreement and leave with one course of action. That is often the whole purpose of gathering. Collective action would be impossible if everyone retained a veto until their internal model became identical to everyone else’s. Someone eventually has to decide. Finance can prefer the smaller investment and still execute the larger one after the chief executive makes the call. Legal can continue to regard a contractual term as risky after the business accepts the risk. An engineer can believe a technical choice is unnecessarily aggressive and nevertheless implement it responsibly after the authority to choose has been exercised.

res.p.000554The danger begins when the necessity of acting rewrites the history of knowing.

res.p.000555A decision becomes “alignment.” Attendance becomes “sign-off.” An objection that was heard and overruled becomes a risk that “the team reviewed.” Weeks later, nobody is lying when they remember the meeting differently. The official record contains the action that followed. The dissenter remembers that the action proceeded despite their objection.

res.p.000556Both accounts can be accurate enough to create trouble.

res.p.000557The Challenger launch decision remains an extreme example of what can happen when unresolved technical disagreement changes form as it travels through an organization. The Rogers Commission’s investigation described the January 1986 launch process in which Morton Thiokol engineers raised concerns about the solid rocket booster’s O-rings under the unusually cold forecast conditions and initially recommended against launch below the temperature supported by previous flight experience. After NASA managers challenged the recommendation, Thiokol management caucused privately and reversed the company’s position. When the teleconference resumed, management recommended launch even though engineers who had opposed it had not suddenly acquired evidence that resolved their concern. Senior decision-makers received a management recommendation whose internal disagreement had largely disappeared from the form in which it moved upward.

res.p.000558The Commission placed that episode inside a much longer institutional history. Concerns around O-ring erosion and blow-by had accumulated over prior flights, and the organization had repeatedly found ways to treat behavior outside the original design expectation as an acceptable flight risk. The disaster cannot be reduced to one conference call, one manager or one meeting technique. That is what makes the meeting useful here. The relevant people can gather, engineers can speak, managers can ask questions, and information can still lose something essential as it changes organizational status.

res.p.000559The disagreement did not have to be censored to disappear.

res.p.000560It had to become a recommendation.

res.p.000561A less catastrophic version happens constantly. Someone says, “I think this date creates a serious implementation risk.” The group discusses it and chooses the date anyway. A month later the project is described as having received cross-functional approval. Nobody fabricated the meeting. The person attended. Their objection was considered. The action moved forward.

res.p.000562What vanished was the difference between participated in the decision and believed the decision was sound.

res.p.000563Preserving that difference does not require turning meeting notes into legal discovery. Most ordinary decisions do not need a transcript of every objection. But some uncertainty belongs to the decision and should remain attached to it. “We are launching Friday; operations believes Friday increases the risk of X, and Jordan is accepting that risk” contains information that “all teams aligned on Friday” destroys.

res.p.000564The first sentence leaves reality somewhere to return.

res.p.000565If X happens, the organization can examine a risk it knowingly accepted. If X does not happen, the dissent does not need to be treated as disloyal or foolish; uncertainty really was uncertainty. The record simply resists the temptation to make the past cleaner because the present requires coordinated action.

res.p.000566A meeting that allows disagreement to survive action can be healthier than one that manufactures unanimity.

res.p.000567That can feel culturally uncomfortable. Organizations often admire alignment because disagreement is expensive to hold in view. A dissenting view complicates communication upward. It can make a decision look weaker. Someone may worry that preserving the objection encourages passive resistance after the choice has been made. Those are legitimate concerns. There is a difference between keeping uncertainty visible and giving every participant permanent permission to relitigate the decision.

res.p.000568People can disagree and still owe the decision competent execution.

res.p.000569The integrity lies in refusing to confuse execution with retroactive belief.

res.p.000570Once meetings are viewed this way, the simplistic war on meetings becomes hard to sustain. Some human work is irreducibly joint. Distributed knowledge sometimes has to become conversation before it can become action. A written exchange allows each function to develop its own argument in isolation; a room lets one person ask a question whose answer changes what another person thought the disagreement was about. Tone carries information. Hesitation carries information. People revise mid-sentence. Somebody who looked obstructive from a distance explains the constraint they are protecting and the constraint turns out to be real.

res.p.000571A company composed entirely of clean handoffs could become extraordinarily efficient at preserving local misunderstandings.

res.p.000572The purpose is not fewer meetings as a moral achievement. It is to stop requiring meetings to carry contradictions that have become ordinary enough to deserve another form.

res.p.000573A recurring exception meeting should become strange over time.

res.p.000574If the group has reviewed forty cases and thirty-two of them share the same structure, the cases themselves have begun drawing a boundary the organization might be able to formalize. Perhaps a regional team can act independently inside a previously approved set of claims and technical conditions. Legal and brand define the envelope rather than returning for every instance. Operations makes the required inputs visible. Only work outside that envelope comes back to the group.

res.p.000575The meeting survives.

res.p.000576Its agenda gets weirder.

res.p.000577That is a healthy result. The humans remain available for judgment where judgment still earns its cost, while ordinary cases stop purchasing senior cognition at retail every week.

res.p.000578Pricing exceptions often reveal the opposite pattern. A company has rules, but so many commercially important deals require exceptions that senior sales, finance, legal and executive leadership meet repeatedly to negotiate them. Everyone gets excellent at the exercise. Sales explains strategic value. Finance protects margin. Legal identifies precedent. An executive decides.

res.p.000579At some point the meeting is no longer merely processing exceptions. It is functioning as the pricing strategy.

res.p.000580Perhaps that is appropriate because the deals really are too heterogeneous to encode. Perhaps the business has learned enough from the repetitions to establish clearer discretion by deal size, margin, contract term or strategic importance. The meeting itself cannot tell us which. Its history can.

res.p.000581Repetition is evidence.

res.p.000582If the participants keep resolving genuinely novel tensions, the repetition may reflect the kind of institution they are. Boards meet repeatedly because governance is recurrent. Investment committees keep encountering investments. A leadership team should not expect to automate away every conflict among growth, risk, capital and people.

res.p.000583If the same people keep having the same argument because nobody is permitted to resolve the condition underneath it, the recurrence says something else.

res.p.000584Meetings belong to the same feedback loops as the earlier chapters. A meeting produces an answer; the useful question is whether anything in that answer travels backward far enough to change the circumstances under which the next case will arrive. If every exception remains merely an exception, the system learns almost nothing from needing them. When a pattern of exceptions changes authority, process, measurement, or promises, the meeting begins reducing the amount of itself the future requires.

res.p.000585There is an appealing organizational fantasy in which all of these adjustments eventually produce a frictionless machine. I do not think that machine exists, and I would be suspicious of it if it did. Human institutions contain values that do not collapse onto a single axis. People have different information and different interests. New circumstances arrive. Exceptions are sometimes genuinely exceptional. An organization that never needs people to negotiate may have eliminated more than bureaucracy.

res.p.000586The aim is not to abolish friction.

res.p.000587It is to keep noticing what the friction is made of.

res.p.000588A meeting can be carrying legitimate plurality. It can be carrying an unresolved authority boundary. It can be carrying the cost of incentives that point in opposite directions. It can be carrying a political conflict the people in the room did not create. It can be carrying uncertainty that deserves to remain uncertainty until somebody with authority chooses. It can be carrying a problem nobody has yet learned enough to encode.

res.p.000589Those loads do not deserve the same remedy.

res.p.000590Sometimes the meeting is the institution through which the tension should continue to be held. Keep it. Give it enough time. Put the right people in it. Stop apologizing for the fact that judgment consumes calendars.

res.p.000591Sometimes the meeting has become a skilled group of people manually maintaining a contradiction no one has been willing to move.

res.p.000592Then the most productive moment may come when somebody stops discussing the case in front of them and says the sentence underneath the agenda.

res.p.000593We keep negotiating deadlines because the people responsible for them do not control them.

res.p.000594We keep escalating customer promises because commitments are being made before the teams carrying the work are involved.

res.p.000595We keep reviewing this project as an execution problem because asking whether the project should still exist belongs to somebody who is not in this room.

res.p.000596We keep calling this a coordination problem because the two executives above us want different things.

res.p.000597The sentence may make the meeting worse for a while. It does not magically reassign authority, repair incentives or make a politically difficult decision painless. It may expose that the people in the room cannot solve what they have been asked to solve.

res.p.000598That is still progress of a particular kind.

res.p.000599The meeting has stopped hiding what it is carrying.

Chapter 13 — The Clean No

res.p.000600When I took over a BDR team, one of the first things they told me was that the outbound volume expected of them was functionally impossible with the tools they had.

res.p.000601They had lost the outbounding platform the team depended on. This was not a matter of the software being slightly less convenient or the reps preferring one interface to another. The activity expectations had been built around a technical capability that no longer existed. Work that had previously been batched and automated had become manual. The arithmetic of the job had changed while the expectations surrounding the job had not.

res.p.000602The team knew this. They had been living inside the mismatch every day. For months, though, the organization had managed to possess both facts at once: these were the output expectations, and the team no longer had the machinery required to reach them. People continued working. Numbers continued being reviewed. The gap expressed itself as inconsistent performance, frustration and increasingly elaborate explanations of why the team was struggling.

res.p.000603When I began asking what was actually getting in the way, their answer was remarkably clean. Until the tool was replaced, the required outbound volume was not credible.

res.p.000604That sentence helped me more than another month of effort would have.

res.p.000605It gave me something I could carry upward without translating an operating constraint into a motivational problem. I did not need to argue that the team was tired enough, frustrated enough or deserving enough. The work and the capacity to perform it no longer matched. If leadership wanted the output, we needed to restore the capability. If restoring the capability was not worth the money or urgency, then the output expectation had to change.

res.p.000606Once the trade was visible, procurement accelerated.

res.p.000607The situation had existed for roughly eight months. The people doing the work had not been hiding it. What had been missing was an answer strong enough to force the organization to choose between the things it was pretending could coexist.

res.p.000608That is what I mean by a clean no.

res.p.000609The phrase can sound like another entry in the enormous literature of personal boundaries, and there is useful work in that literature. Plenty of people have spent years agreeing to things through fear, habit or compulsive accommodation and genuinely need permission to refuse. I am interested in something slightly different. The value of a no is not that refusal is inherently brave or self-protective. A no becomes useful when it prevents other people from building plans on a capacity, commitment or condition that does not exist.

res.p.000610A careless yes does the opposite.

res.p.000611When someone asks whether a team can take on another project, the question remains open until the answer arrives. The requester may have to reduce the scope, change the date, spend more money, find another team, disappoint a customer or decide that something already underway matters more. Those alternatives can be unpleasant, but they remain available while the capacity question is unresolved.

res.p.000612Say yes and most of them disappear.

res.p.000613The requester can move on. The work enters the plan. Another function accepts a dependency. A customer hears a date. An executive update now contains the initiative. People make perfectly reasonable decisions because they believe somebody has already absorbed the constraint.

res.p.000614If the capacity was never there, the choice has not vanished. It has only moved.

res.p.000615Somebody delays another project without announcing that the project has been deprioritized. Somebody compresses review. A person who already had a full week opens the laptop after dinner. The team accepts a little more risk because acknowledging the actual trade would require reopening the commitment. By the time the cost appears, the original yes has dissolved into dozens of local accommodations made by people trying to keep it true.

res.p.000616Apparently generous commitments can become strangely unfair. The person saying yes gets to experience willingness. Other people may get the accounting.

res.p.000617Managers can do this with especially good intentions. There is an executive request, the manager understands why it matters, and they do not want to become the leader whose first response to difficult work is a list of reasons it cannot be done. They say yes in the room and plan to protect the team by sorting out the details afterward.

res.p.000618The team then receives a priority decision the manager declined to expose upward.

res.p.000619Which deadline should move? Which quality check can be shortened? Who has enough slack to absorb the work? Which person can take one more evening? What work can quietly become less good without attracting attention? These are genuine management decisions, but the people making them downstream may have considerably less authority to change the commitment that created the conflict.

res.p.000620The leader has protected the requester from scarcity and made the team administer it.

res.p.000621I have done versions of this. Most managers who are rewarded for being responsive probably have. There is an appealing competence in being the person who finds a way. Teams value leaders who do not panic at surprises. Businesses need people willing to stretch when circumstances demand it. Some of the best work I have seen happened because people voluntarily absorbed something unreasonable for a few days and got an important outcome over the line.

res.p.000622The pathology is not effort beyond the plan. It is a system that learns to regard extraordinary absorption as ordinary capacity.

res.p.000623A team that repeatedly rescues the organization can make the organization worse at knowing what the team can actually do. Urgent work finds reliable people because reliable people are good places to put urgent work. Ambiguous ownership finds the person who will pick it up rather than let it fail. Two priorities that cannot both fit find the manager who will refrain from making anyone choose between them.

res.p.000624Eventually the capability to survive contradiction becomes part of the reason the contradiction keeps being produced.

res.p.000625The BDR problem was a particularly clear version because the missing capacity had a name and a purchase order. Many constraints are less tidy. A team may technically possess enough hours but lack a particular expertise. A deadline may be possible only if nothing goes wrong in work where things routinely go wrong. A person may be capable of performing another task but already carrying enough context switching that the additional work damages several things at once. The constraint is real without being as easy to display as an absent software license.

res.p.000626“No” becomes too crude if it is treated as a complete management philosophy.

res.p.000627Often the clean answer is conditional.

res.p.000628We can take this on if the regional launch moves.

res.p.000629We can hit Wednesday if these two pieces of scope come out.

res.p.000630We can own the work, but not until next month.

res.p.000631We can promise the outcome once the team that has to deliver it has validated the commitment.

res.p.000632The point of those sentences is not rhetorical firmness. Each one returns the decision to the place where the trade can still be made. The leader requesting both projects learns that both projects are competing for the same capacity. The customer asking for Wednesday receives a real relationship between time and scope. The person allocating work learns when capacity becomes available instead of receiving a vague agreement that will later transform into delay.

res.p.000633A priority becomes real only when something else is permitted to lose.

res.p.000634This sounds obvious until you look at how often organizations perform prioritization without permitting loss. Twelve initiatives are ranked from one to twelve. Everyone agrees that one is the most important. Number twelve retains a sponsor, a deadline, a working team and a place in the quarterly plan. Number eleven still has a steering committee. Number ten needs a decision on Thursday, and because its deadline remains real, it receives the decision.

res.p.000635The ranking altered the order of the list. It did not alter the portfolio.

res.p.000636Then execution becomes the place where the organization discovers which priorities were fictional. Teams delay something, under-resource something, reduce quality somewhere or work beyond planned capacity. Leadership experiences uneven delivery and responds with closer performance management. The actual prioritization has occurred, but late and mostly among people without the authority to say what has been deprioritized.

res.p.000637A clean no can make that choice visible before execution has to make it secretly.

res.p.000638That is part of why I resist treating this subject as confidence training. An employee saying no upward and an executive saying no downward are performing different acts even when they use exactly the same word. The executive can often refuse cheaply. The junior employee may be correct that refusal will affect how they are perceived. A contractor can understand perfectly that additional scope deserves another fee and still know that pushing the issue may cost the client. Someone financially dependent on a difficult job experiences the manager’s “optional” Saturday request differently from a colleague with a year of expenses in the bank.

res.p.000639The price of truth affects how much truth a system receives.

res.p.000640If every statement of constraint is answered as a test of commitment, people will learn to stop presenting constraints as statements. The no still occurs. It arrives as slippage, weaker work, attrition, missed details and people quietly choosing which instruction not to follow. The leader who made refusal socially unavailable has not created a more committed organization. They have chosen more expensive forms of refusal.

res.p.000641Receiving a no is a leadership capability.

res.p.000642“I don’t have room for this unless something moves” is not automatically correct. The employee may be protecting work that genuinely matters less. Their estimate may be padded. They may not understand the priority. The deadline may be hard rather than impossible. The manager should be able to interrogate the constraint.

res.p.000643But interrogation is different from teaching someone that only one answer is safe.

res.p.000644If every capacity conversation ends with a speech about urgency, ownership or being a team player, the team does not become more capable. It becomes better at translating its limitations into language leadership prefers. “We’ll make it happen” appears in the meeting. The real no occurs later where nobody can use it.

res.p.000645The people doing the work usually know this distinction immediately. They can tell when a question about feasibility is actually a request for reassurance. They know which leaders hear an inconvenient answer as useful information and which hear it as an attitude problem. Over time, the quality of capacity information reaching the leader adapts to that knowledge.

res.p.000646People become clearer when clarity survives contact with us.

res.p.000647This operates outside work as well, although the moral terrain changes enough that business language becomes dangerous if imported carelessly.

res.p.000648There are commitments I should accept even when I do not have convenient surplus capacity. Parenting contains many of them. Friendship contains some. Shared obligations do not dissolve because I happen to be tired when my part arrives. A theory of healthy boundaries in which every unwanted demand becomes somebody else’s problem would produce a very comfortable person and a fairly bleak set of relationships around them.

res.p.000649Sometimes the clean answer is yes.

res.p.000650The cleanliness comes from owning the cost instead of accepting it and later charging somebody else emotionally for the fact that it existed.

res.p.000651If I agree to help a friend move on Saturday while knowing the week has already exhausted me, I can decide that the friendship is worth a difficult Saturday. What I cannot reasonably do is agree, conceal from them that the commitment collides with three others, arrive rushed and resentful, and then experience their request as the thing that ruined my weekend. My yes entered their planning. They were entitled to treat it as information.

res.p.000652A person who says yes to everything can therefore become surprisingly unreliable despite experiencing themselves as generous. Every individual commitment is made from care. The aggregate system has no additional time in it. Eventually the collision decides. They arrive late, leave early, become resentful, cancel, work overnight or reduce the quality of several commitments at once.

res.p.000653Nobody they agreed to help was necessarily unreasonable. Each person received a yes and planned accordingly.

res.p.000654The reluctance to disappoint anyone at the moment of request created more distributed disappointment later.

res.p.000655That does not make the clean no morally superior in every case. “I don’t want to” is real information and sometimes sufficient information. It is not universal jurisdiction. People have duties, and a commitment we freely made earlier may deserve authority precisely when the present version of us wishes it did not. The question is whether the answer corresponds closely enough to the obligation, the capacity and the consequences we are actually willing to own.

res.p.000656The same distinction appears in systems that formalize scarcity before pressure arrives. Google’s Site Reliability Engineering practice uses the idea of an error budget to mediate the permanent tension between shipping changes and preserving reliability. A service has an agreed reliability objective, and the gap between perfect reliability and that objective creates a finite tolerance for failure. When failures consume enough of that tolerance, the organization’s published example policy stops most ordinary releases until reliability has recovered.

res.p.000657The mechanism interests me because the no has been partially decided before the most attractive exception appears.

res.p.000658Without some prior structure, every new release can recreate the whole argument. The customer is important. The feature is almost finished. This quarter matters. The outage was unusual. We have already invested the work. Could we make one exception?

res.p.000659Any one of those claims may be persuasive. The error budget does not make judgment unnecessary, and the policy preserves exceptions for things such as critical security fixes. What it does is prevent every request from approaching the system as though it were the first request ever to encounter finite reliability.

res.p.000660Most human life should not be managed with error budgets. Friendship would become intolerable if every favor consumed a visible quarterly allowance, and family life has enough spreadsheets without adding one for affection. The technical example works because the thing being constrained can be measured unusually well.

res.p.000661The broader idea survives without the arithmetic.

res.p.000662The hardest moment to discover a boundary is often the moment someone wants the thing beyond it.

res.p.000663That was the problem with the BDR team. The capacity model had effectively remained open to negotiation every day. Each rep could push harder. Each target could still be discussed as though enough execution might close the gap. The organization had not been forced to decide whether the tool was necessary because the human beings between the target and the missing capability were still absorbing enough of the difference to keep the question ambiguous.

res.p.000664Once the answer became “we cannot credibly produce this volume under the present conditions,” the conversation changed. It was no longer primarily about whether the team could squeeze out another burst of effort. The missing tool became an organizational choice.

res.p.000665That is the most useful function of a clean no: it can return ownership of a constraint to someone capable of changing it.

res.p.000666Sometimes the constraint is budget. Sometimes it is time. Sometimes another priority has to move. Sometimes the request itself is impossible. Sometimes the constraint is mine and I need to absorb it because I already made the promise. Sometimes what looked impossible becomes possible once someone challenges an assumption I was treating as fixed.

res.p.000667The no has to remain correctable because every other judgment in this book does.

res.p.000668What it cannot do is vanish merely because the preferred answer was yes.

res.p.000669When people are allowed to refuse, agreement becomes more informative. “Friday works” starts meaning that Friday is credible rather than that Friday was the least dangerous answer available in the meeting. A team accepting a project means more when it could have exposed the work that would be displaced. A customer recommendation means more when the company is willing to admit that some customers should not buy. A person saying yes to an invitation means more when no would have survived the relationship.

res.p.000670The value of the no finally appears inside the yes.

res.p.000671The BDRs did not need me to teach them a lesson about boundaries. They needed the condition they were already living inside to become admissible in the decision above them. Once it was, the organization could respond to the actual problem.

res.p.000672The tool got replaced.

res.p.000673The work did not become easy. Targets did not stop being targets. People still had to perform.

res.p.000674But for the first time in months, the system was asking them to perform inside conditions that had some relationship to what performance required.

Chapter 14 — Money Is Stored Optionality

res.p.000675When I was younger, money did not feel much like a score.

res.p.000676I had a growing family and a fairly immediate need to make the economics of our life work. I did not spend much time wondering whether income was an expression of personal worth or whether accumulating it was an interesting game in its own right. Money solved problems. More precisely, enough money kept several problems from becoming one problem at the same time.

res.p.000677There is a point of financial pressure at which almost nothing stays in its lane. A car repair is not only a car repair because transportation is also employment, childcare and the timing of the next paycheck. A sick day is not only a sick day when losing the day’s income affects groceries. Leaving a bad job is not merely a professional decision if rent is due before another employer will begin paying you. Something that would be irritating in a household with reserves can become a small systems failure in a household without them, reaching into credit, schedules, relationships and whatever other obligation happened to be standing nearby.

res.p.000678As our circumstances improved, the most noticeable change was not that each individual problem became cheaper. The relationship among problems changed. Bills could arrive without requiring choreography. A bad month did not automatically colonize the next one. Work decisions could contain somewhat more attention to the work itself because every decision was no longer being adjudicated under the same immediate financial threat.

res.p.000679That room is one of the things money buys.

res.p.000680Most of the money we spend has already been told what it is for. A mortgage payment is highly specialized money. So is the grocery bill, tuition, a plane ticket or the amount that leaves the account automatically for insurance. Money in reserve has a different kind of usefulness. Five thousand uncommitted dollars might later become a repair, rent during a job transition, medical care, a flight to someone who needs you, childcare, legal help or three weeks in which income disappears. It might remain untouched because none of those things happens.

res.p.000681Its value includes the futures it leaves open.

res.p.000682That quality is easy to overlook because money is usually represented as a number, and numbers invite comparison. This dinner costs this much. The job pays that much. One house costs more than another. The retirement account contains some amount and ought eventually to contain a larger one. The units are wonderfully consistent even though the things we ask them to represent are not.

res.p.000683Security is not the same thing as pleasure. Pleasure is not the same thing as status. Status is not the same thing as autonomy. Autonomy is not the same thing as generosity. Yet money can be converted into all of them, imperfectly and under different conditions, which makes it tempting to treat more money as though it were simply more of every desirable thing at once.

res.p.000684Early financial pressure gives money a very concrete argument. Another ten thousand dollars can pay off the credit card, build the emergency fund, repair the roof or make childcare possible. The connection between the number and the life is easy to name. Once those pressures recede, additional money may still have excellent uses, but the question becomes less automatic. Accumulation can inherit the moral force of all the emergencies it once solved.

res.p.000685A person who once worked harder because more income meant security can continue working harder long after security has ceased to be the marginal thing the next dollar is buying.

res.p.000686This is not a disguised argument for asceticism. I like things. Some expensive things are wonderful. A beautiful home can improve ordinary Wednesday mornings for decades. Travel can be worth far more than its price. Convenience can return hours to a life. Hobbies are allowed to be financially ridiculous. There is no special moral purity in dying with an immaculate savings rate and a long list of things you were too disciplined to enjoy.

res.p.000687The question is what job we are actually asking the money to do.

res.p.000688A purchase can perform several jobs simultaneously. A house can provide shelter, room for children, access to a neighborhood, beauty, privacy, an investment and status. The status does not make the shelter imaginary. An expensive watch can be beautifully made and communicate something about its wearer at the same time. A promotion may bring more income, more authority, harder problems, greater security and the satisfaction of having moved upward.

res.p.000689Confusion enters when one of those purposes is doing the emotional work while another is supplying the argument.

res.p.000690A couple can argue about whether they can afford an expensive purchase when the account balance is not really in dispute. One person is protecting a sense of security. The other is protecting the right to enjoy money they helped earn. They can become increasingly precise about the arithmetic because arithmetic is easier to defend than either of those feelings. Both people may be mathematically correct for the entire duration of an argument the mathematics cannot settle.

res.p.000691The same thing happens inside one person. Someone pursues another compensation increase in the name of freedom while the work required to obtain and maintain that compensation consumes more of the time they hoped freedom would restore. Someone turns down a promotion because they prefer the life they already have, then discovers that part of them had been using advancement as evidence that they were still ambitious. Someone says an expensive purchase is about craftsmanship when some portion of the pleasure plainly comes from what the object communicates.

res.p.000692Allowing the real purpose into the explanation does not force a particular decision. It simply lets the decision be made against the thing actually being purchased.

res.p.000693Status is allowed to be on the receipt.

res.p.000694So is convenience. So is beauty. So is showing off. So is giving your children opportunities you did not have. So is wanting enough money that no employer can frighten you into doing something you believe is wrong.

res.p.000695The ability to refuse is one of money’s least glamorous and most consequential products.

res.p.000696The previous chapter dealt with the cost of saying no when the person receiving the request has less power. Savings can change that cost without making anyone wiser or braver. An employee with six months of expenses in the bank encounters a terrible manager differently from an employee whose rent depends on next Friday’s deposit. Both can understand the situation perfectly. Both can know exactly what they would prefer to do. One has a larger set of actions available before clarity becomes catastrophe.

res.p.000697Money did not provide the judgment. It altered the conditions under which the judgment could be acted upon.

res.p.000698That difference gets flattened easily in stories about success. Someone leaves a bad job without another one lined up and eventually finds work that fits them beautifully. The story becomes courage, patience or unusual conviction. Perhaps all three were present. If the person also had enough money to search for six months, the runway belongs in the explanation.

res.p.000699A founder can make a riskier bet when family money protects the household from failure. A parent who can buy childcare has hours another parent has to supply personally. Someone recovering from illness with savings and paid help occupies a different recovery environment from someone trying to cook, clean, drive children and maintain income through the same illness. Resources change which decisions are survivable.

res.p.000700That does not make the person with resources less capable. It makes comparisons based only on the visible outcome much less informative.

res.p.000701Virginia Woolf built a famous argument around precisely this material fact in A Room of One’s Own. The room matters, but Woolf’s narrator also has an inherited income of five hundred pounds a year. The money changes more than purchasing power. It reduces the extent to which every thought has to pass through somebody else’s economic approval. Work no longer has to be selected entirely according to what another person will pay for immediately. Woolf does not ask women to achieve intellectual independence by developing a better attitude toward material dependence; the argument insists on the material conditions under which independence becomes easier to exercise.

res.p.000702The absence of those conditions affects what becomes visible as talent. A person who has privacy, education, time and enough economic security to continue working can produce evidence of ability. Someone whose attention is repeatedly interrupted by survival may possess no corresponding record. Looking only at the finished work can make the difference appear to reside entirely in the people when part of it resided in the worlds around them.

res.p.000703That should make us careful whenever financial outcomes are allowed to become moral evidence.

res.p.000704Wealth can reflect unusual skill. It can also reflect ownership, inheritance, timing, geography, leverage, institutional position, luck, appetite for risk and the fact that some markets reward certain abilities extraordinarily well. People can perform difficult, socially necessary work and remain poorly paid. Other people can become extraordinarily rich while significant costs of the activity that enriched them are carried elsewhere.

res.p.000705Financial struggle proves no inverse moral theorem either.

res.p.000706There is a particularly ugly genre of advice that looks at someone under material pressure and searches immediately for the belief causing the pressure. Scarcity mindset. Failure to value yourself. Fear of abundance. Lack of alignment. The language becomes obscene once it is being used to make arithmetic disappear.

res.p.000707If wages do not cover rent, wages and rent belong in the explanation.

res.p.000708If childcare consumes most of a paycheck, the price of childcare belongs there.

res.p.000709If someone chooses a more expensive short-term option because every cheaper long-term option requires cash, credit or time they do not possess, the choice should be compared with the alternatives they actually had rather than the alternatives available to the person offering advice.

res.p.000710Integration cannot manufacture margin where the arithmetic contains none.

res.p.000711Where discretion does exist, however, financial commitments can accumulate exactly the way calendar commitments do: separately at first, simultaneously later. The house was affordable. The car was affordable. So was the school, the annual trip, the membership, the second property, the convenience that saved an hour every week, the restaurant habit that became normal because nobody regretted any particular dinner.

res.p.000712Each decision can be reasonable on the day it is made.

res.p.000713The collection can still consume the optionality the higher income was expected to provide.

res.p.000714This can produce a strange reversal. Someone works for years toward greater financial freedom, succeeds, and builds a life whose recurring cost makes the high income increasingly non-negotiable. The large house, cars, tuition, travel and ordinary baseline of the household now require the salary that was once attractive partly because it promised a future in which salary would matter less.

res.p.000715The person may love the life they purchased. There is no contradiction in that. The difficulty begins when the resulting dependence is described as freedom because freedom was the reason the income was originally pursued.

res.p.000716A larger income and a larger margin are different things.

res.p.000717That distinction also explains why two people earning identical salaries can inhabit very different financial realities. One can survive a long job search, help someone through an emergency or take a year on work that may fail. The other may have nearly every future dollar assigned before it arrives. Neither condition tells us enough to judge the life that produced it. Housing, medical costs, care obligations, debt and geography can consume enormous amounts of income without extravagance ever entering the picture.

res.p.000718Sometimes there simply is not enough money.

res.p.000719Where there is enough, the accumulation of fixed claims deserves attention because optionality has to remain uncommitted in order to function as optionality. The emergency fund works because the emergency has not yet been selected. Runway works because the future job has not yet been chosen. Money capable of becoming ten different things tomorrow is performing a kind of work today even while it appears to be sitting still.

res.p.000720It is preserving routes.

res.p.000721There is no obvious point at which enough routes have been preserved. Another dollar can always defend against a rarer disaster, extend the period without income, enlarge an inheritance or protect a higher standard of living through retirement. These are genuine goods. The cost appears because the resources used to acquire the next increment of money have other possible lives.

res.p.000722The better-paying role may require travel. The promotion may consume evenings. Another five years in work you dislike may enlarge the reserve enough that you can finally stop doing work you dislike. At some point the circularity deserves to become visible.

res.p.000723Money cannot tell us how much life should be spent acquiring money.

res.p.000724The number has no internal answer.

res.p.000725This is one place where wealth can become as distorting as scarcity, though the distortions are obviously not equivalent in harm. Scarcity can force reality onto a person with brutal speed. Abundant resources can sometimes prevent reality from arriving.

res.p.000726A company with enough capital can sustain a product without customers for an impressively long time. Another management layer can be added when communication stops working. Consultants can absorb the consequences of an executive who never develops the capability they were hired to provide. An individual with enough money can make inconvenience vanish from their own day by paying someone else to perform the inconvenient part.

res.p.000727Often this is exactly what resources are for. I do not want a medical emergency to become educational because I lacked insurance. A family reserve should prevent children from experiencing every fluctuation in adult income. A good business keeps cash so that one difficult quarter does not force destructive short-term decisions.

res.p.000728Buffers make systems resilient because every shock does not need to become an emergency.

res.p.000729The same buffer can become a problem when it prevents useful consequences from making the return trip.

res.p.000730A leader can keep hiring people around a weakness that the organization would otherwise force them to confront. A wealthy person can leave every community after conflict and develop a remarkably stable theory that every community eventually becomes unreasonable. A company can interpret the ability to keep funding something as evidence that continuing to fund it remains sensible.

res.p.000731Money widens the range of responses. It does not select among them.

res.p.000732The distinction becomes particularly important in a book concerned with correctability. Optionality is valuable because it keeps more responses possible after new information arrives. If the new information cannot actually alter the use of the resources, the optionality has become mostly ceremonial.

res.p.000733A reserve that can never be spent because spending it produces intolerable anxiety is not functioning exactly like a reserve. Wealth pursued for freedom but protected from every use that would exercise freedom has started becoming its own objective. The person possesses options and experiences almost none of them as available.

res.p.000734The opposite mistake spends every reserve before uncertainty has finished being uncertain. Because money can become almost anything, almost anything can be supplied with an argument for why this is the moment the reserve was meant to serve. Security disappears one persuasive exception at a time.

res.p.000735Neither problem can be solved by naming a universal savings rate.

res.p.000736The number depends on obligations, temperament, age, income stability, health, family structure, the social safety net surrounding the person and the risks they are actually trying to survive. A useful question is less precise and more demanding: what freedom is this money meant to preserve, and does the way I am earning, holding or spending it still serve that freedom?

res.p.000737The answer can include enjoyment.

res.p.000738It can include a ridiculous guitar.

res.p.000739It can include more security than another person considers necessary. It can include leaving a large inheritance. It can include giving money away because increasing another person’s room around a problem matters more than increasing your own balance.

res.p.000740Money given well can produce optionality unusually directly. Rent during a crisis can keep someone from accepting the first bad solution available. A plane ticket can make presence possible. Legal help can turn a right that exists on paper into a right someone can exercise. Childcare can return attention. A month of expenses can let a person recover instead of choosing between recovery and income.

res.p.000741The money becomes room around somebody else’s problem.

res.p.000742Generosity complicates the transaction because money is unusually capable of carrying invisible claims with it. Families know this well. A gift can later become a vote. Help can become evidence in an argument. Money offered so that someone can make their own decision can quietly establish a new expectation that the giver be consulted about the decision.

res.p.000743The bank statement contains the transfer and none of the terms that mattered most.

res.p.000744One person remembers generosity. The other remembers the moment their independence became negotiable. Both can point to the same check.

res.p.000745This is another reason money causes conflicts that seem wildly disproportionate to the amount being discussed. The receipt gives people something solid enough to litigate. The actual dispute may concern security, control, gratitude, fairness, permission, fear or what each person believes the money entitles them to decide.

res.p.000746Eventually arithmetic reaches the edge of its jurisdiction.

res.p.000747Someone has to say that the expense is not frightening because of two hundred dollars; it is frightening because the spending makes the future feel less safe. Or that the repeated defense of the purchase is not really about affordability; it comes from feeling as though money jointly earned can only be enjoyed after permission has been granted.

res.p.000748The number remains real after the sentence.

res.p.000749It simply stops being required to carry the whole argument.

res.p.000750That is close to the relationship with money I want. Money is real. Material security is real. Poverty is real. A reserve can change which decisions are possible, and the absence of a reserve can make excellent judgment impossible to act upon. None of those facts requires money to become a measure of the person holding it.

res.p.000751Money is unusually powerful because it can wait.

res.p.000752Before it is committed, it can become many things. It can absorb a shock, purchase time, fund pleasure, protect a principle, support another person, make experimentation survivable or remain unused until a future we cannot currently describe.

res.p.000753That is stored optionality.

res.p.000754Its purpose is not to make the number large enough that life finally stops happening to us.

res.p.000755It is to leave enough room that, when life does happen, we retain some choice about how to answer.

Chapter 15 — The Cheapest Conflict Is the One You Actually Have

res.p.000756Some arguments survive their solutions. A couple agrees on the amount either person can spend without discussion and soon finds another purchase carrying exactly the same emotional charge. A manager clarifies what “be more strategic” means, watches an employee change the requested behaviors, and discovers that the unease remains. Two functions rewrite a handoff until ownership is unmistakable and continue accusing one another of making the work harder than it needs to be. None of this proves that the visible disagreement was fake. People really do overspend, communicate badly, miss deadlines and design terrible processes, and there is no special wisdom in treating ordinary failures as symbols of something deeper. The pattern becomes interesting when the ostensible problem is repeatedly repaired without producing the reduction in conflict that the repair ought reasonably to have produced. The vocabulary changes while the argument retains its shape. At that point, the recurrence itself becomes information. Something may be using the visible problem as a place to stand.

res.p.000757Concrete disputes are attractive because they can be prosecuted. There is a receipt, a missed deadline, a sentence that can be quoted, a timestamp, a workload that can be counted, a process step that did or did not occur. Larger conflicts are often harder to verify. Whether one person’s time receives less standing in a relationship, whether a manager trusts an employee’s judgment, whether a team feels supervised because another function genuinely sees dangerous mistakes or because authority has never really been surrendered—these are harder objects. They contain history, interpretation and facts that may be distributed across several people. A smaller dispute can therefore become a kind of handle on something much harder to grasp. The handle is not imaginary. The late text really was late. The missed deadline really was missed. The question is whether fixing the handle changes the thing whose weight it was carrying.

res.p.000758“Be more strategic” is almost perfectly designed to sustain this kind of confusion. It has enough real professional content that nobody has to be insincere when using it. A manager can reasonably mean that an employee needs to connect functional work to broader business outcomes, spend less time inside execution, anticipate what leaders will need and bring stronger recommendations. The employee can reasonably respond by doing exactly those things: delegate more, write at a higher level, frame recommendations around business tradeoffs, enter meetings with a view rather than a pile of information. If the feedback then returns three months later in slightly altered language—step back more, operate at the next level, think more strategically—both people can continue working on the word for a very long time. Competency models can become more detailed. Presentations can become more executive-friendly. The manager can produce additional examples, and the employee can become increasingly skilled at displaying the expected signals. Sometimes this eventually works because the problem really was strategic capability. Sometimes the harder sentence has been waiting underneath the entire development plan: the manager does not yet trust the employee’s judgment.

res.p.000759That sentence costs more because it creates obligations for the manager as well as the employee. Distrust sounds like an assessment until the other person asks what it rests on. Which decisions caused concern? What did the employee fail to see? What authority are they actually supposed to exercise independently? If their conclusion differs from the manager’s, is independent judgment still desired? What evidence would alter the manager’s confidence, and would improved confidence result in actual changes to the operating model? Those questions may vindicate the manager. Perhaps the employee is excellent at analysis and repeatedly weak when competing interests have to be traded against one another. Perhaps two decisions exposed a pattern the employee genuinely needs to correct. But the questions may also reveal that the manager has been requesting autonomy while overriding nearly every consequential decision, or that a piece of work associated with the manager’s own identity has never really become delegable. “Strategic” can absorb almost any dissatisfaction. Trust is harder to hide inside because trust immediately asks for evidence and some imaginable relationship between evidence and future authority.

res.p.000760Organizations can build large amounts of infrastructure around a conflict that remains at this level of blur. One team believes another needs more rigor, so new fields appear on the intake form. An approval is added. Documentation becomes more formal. A weekly review begins. None of those changes is necessarily foolish. If important errors have occurred, a new control may be exactly the correct response. Documentation created during a period of distrust may become genuinely useful institutional memory. The problem arises when the machinery has no defined relationship to the condition that produced it. If the real operating fact is that one group does not trust another to make a particular class of decision unsupervised, then improved performance should eventually be capable of changing the supervision. Otherwise process has become the medium through which distrust is administered indefinitely. The team can satisfy every new requirement without learning whether success will ever buy them more authority, while the supervising group can point to the continuing controls as evidence that the controls remain necessary. What began as a response to mistrust becomes an environment in which mistrust no longer needs to be stated.

res.p.000761The same thing happens with apparently neutral requests for accountability. Sometimes nobody knows who owns a decision, and a clearer assignment of responsibility solves the problem immediately. Sometimes ownership is already obvious and another leader dislikes the choices the owner keeps making. Sometimes the owner is held responsible for an outcome without control over the people, money or dependencies required to produce it. Sometimes two executives want incompatible things and everyone below them is being asked to solve the resulting conflict as though it were confusion about roles. All four problems can produce a workshop and a RACI chart. The first may be repaired by one. The others will simply become better documented versions of themselves. A process artifact becomes wasteful when we ask it to resolve a disagreement we have declined to put inside it.

res.p.000762Getting closer to the real conflict does not require announcing a theory of somebody else’s motives. In fact, one of the quickest ways to make this entire line of inquiry useless is to become the person who has recently learned psychology, systems thinking or therapeutic language and now possesses an explanation for everybody. You are not angry about the missed deadline; you have a control issue. You are not opposed to the reorganization; your status is threatened. You do not want more detail; uncertainty makes you anxious. Even when such an interpretation happens to be correct, accuracy does not grant ownership of another person’s interior life. Conflict becomes more tractable when people stay close to information they actually possess. “After the last two dates slipped, I am having trouble trusting this estimate” contains events and an inference. “You are unreliable” turns the inference into a property of the person. “When our plans change repeatedly and your work plans do not, I start feeling that my time has less standing” gives another person something they can dispute, contextualize or recognize. “You don’t respect my time” arrives with the adjudication already complete.

res.p.000763The difference is epistemic as much as interpersonal. If I describe what happened and what it is causing me to believe, the other person can show me information I do not have. They can point out the counterexamples my memory dropped, explain a constraint I interpreted as preference or agree that the pattern is real while disagreeing about its cause. I retain the right to report my experience without requiring them to accept my explanation of theirs. The cheapest conflict is not the one with the deepest psychological interpretation. It is the one stated close enough to the consequential disagreement that changing it would change what happens next.

res.p.000764That can take patience because the larger conflict may not exist in a cleanly formulated state inside anyone’s head. Imagine an employee who keeps objecting to a reorganization. First the reporting line is wrong. Leadership changes part of it. Then the rollout sequence creates problems. The schedule is adjusted. The names of roles are confusing. The FAQ expands. Each objection may be legitimate, and leadership can spend months responding conscientiously while becoming increasingly convinced that no response will ever satisfy this person. “Resistance to change” becomes an obvious diagnosis precisely because it can explain every future objection in advance. A more useful intervention begins from the pattern without pretending to know its cause: we have addressed several specific concerns you raised, and your overall concern seems largely unchanged; is there something about this reorganization that we have not actually discussed? The answer gets to remain literal. Perhaps the reporting structure really is the problem and the other changes never mattered. Reality does not owe us a more sophisticated explanation. Or the employee may finally say that they believe the incoming leader is incapable of running the function, that their own role is being reduced while everyone calls it development, that the new structure moves work away from the people who understand it, or that they believe a colleague was treated unfairly. Those statements are harder to deal with, which helps explain why the argument found safer objects first.

res.p.000765Proxy conflicts are not merely mistakes. They often perform useful social work. They let somebody object without making the full objection. A person can argue about dishes before admitting that they feel alone in maintaining the household. An executive can question execution before admitting that the strategy they publicly sponsored may be failing. A team can ask for clearer process instead of saying that they do not trust a senior leader’s decisions. The disguise creates distance from a more dangerous claim and sometimes preserves enough of the relationship for people to continue functioning while they decide what they actually think. The fashionable preference for radical directness becomes morally thin. Directness is not equally priced for everyone. A chief executive can tell a subordinate, “I don’t trust this team’s execution,” and expect an agenda to form around the concern. The subordinate who believes the executive team repeatedly changes direction and then blames execution may need months of evidence and a remarkably safe moment to say anything similar upward. Both claims may describe the same organization from different sides. Only one speaker can assume curiosity as the default response.

res.p.000766The asymmetry becomes much more serious in relationships where truth is punished. A person dealing with coercion or abuse is not failing a communication exercise because they raise the larger subject indirectly. If naming the real conflict produces retaliation, financial punishment, rage or violence, the smaller argument may be protective. Likewise, an employee who builds elaborate documentation because undocumented disagreement gets rewritten later may be performing unnecessary labor in some abstractly better organization and completely rational labor in the organization they actually inhabit. Telling the downstream person to “just be direct” can mistake the location where compensatory work appears for the location of the condition creating it. Some indirectness is expensive because people are afraid of ordinary discomfort. Some is expensive because the direct route has been made genuinely unsafe. Any account of conflict that treats those conditions as interchangeable will end up advising courage to people who need leverage, protection or an exit.

res.p.000767Where enough safety exists, though, proxy conflicts become expensive precisely because every disguise has to be maintained. If the disagreement is about trust while the conversation remains about process, nobody knows what successful process is supposed to permit them to stop doing. If the disagreement is about recognition but continues appearing as workload, tasks can be redistributed with great precision while the person who feels taken for granted remains taken for granted. If two teams actually disagree about who should possess authority, each imperfect handoff becomes another piece of evidence about the handoff rather than evidence about the unresolved authority boundary. People become extremely good at solving manifestations of the conflict because manifestations are what the system permits them to touch.

res.p.000768Household arguments show the same pattern with unnerving clarity because people can collect exquisite data about questions the data cannot settle. Who cooked? Who cleaned? Who made the appointments? Who remembered the birthday, drove the children, initiated the difficult conversation, took time off when somebody got sick? Counting can expose a real imbalance and sometimes the count is exactly what a couple has avoided seeing. But there are arguments in which every chore can be divided fairly and one person still experiences the relationship as nonreciprocal. The spreadsheet was not wrong. It simply reached the edge of what it could establish. Money behaves similarly. A couple may dispute a purchase as though affordability were uncertain even when both understand the balance perfectly. One person may be defending security; the other may be defending the right to enjoy money they helped earn without feeling supervised. A spending rule can prevent real financial damage without resolving either of those experiences. A proxy can be entirely real and still incomplete.

res.p.000769The useful question is not whether the surface demand should be discarded, but what victory over that demand is expected to change. If the dishes were divided exactly as requested, would the resentment plausibly end? If the team completed every new control for six months without a serious error, would authority move? If an employee performed the development plan exceptionally well, would the manager genuinely allow them to make decisions currently being reviewed? If the financial target were reached, would the anxiety become smaller, or would safety simply acquire a new number? A “no” to any of these does not make the requested change unnecessary. It tells us that the requested change is carrying more than one job.

res.p.000770Large negotiations sometimes make this distinction between position and concern easier to see because the consequences of fusing them become so visible. One narrow part of the negotiations that produced the 1978 Camp David framework concerned Egyptian sovereignty over Sinai and Israeli security concerns connected to the same territory. Egypt sought the return of land occupied by Israel since the 1967 war. Israel’s security concerns were partly bound to military control of that land. If sovereignty and security remained identical to physical possession, the bargaining space was extremely narrow. The eventual framework separated them enough to permit Israeli withdrawal and restored Egyptian sovereignty alongside limitations on forces, security zones and monitoring arrangements designed to reduce the risks Israel associated with withdrawal. The example has to remain narrow. Camp David did not resolve the wider Arab-Israeli conflict, and the unresolved Palestinian dimension makes any cheerful parable drawn from the accords grotesque. The structural point is simply that two positions which appeared fused contained concerns that could be treated differently without pretending either concern had disappeared.

res.p.000771Smaller conflicts often contain the same hidden fusion. “I need final approval” may contain “I will be held responsible if this fails.” “I need you to stop checking my work” may contain “I cannot develop judgment while every consequential decision is preempted.” “We need to spend less” may contain “I am frightened by how little room remains if something unexpected happens.” “I need you at this family event” may contain “I need to know that belonging to this family still makes a claim on you.” None of those underlying concerns invalidates the original demand. The accountable leader may genuinely need final approval. Spending may genuinely have to fall. Attending the family event may be part of a promise someone ought to keep. Naming the concern does not manufacture a compromise. It stops us from assuming that compliance with the position will necessarily satisfy the reason the position has acquired so much force.

res.p.000772What remains after a repair is often more informative than the initial dispute. If a manager institutes the review process they asked for and mistrust remains untouched, the organization has learned that the process and the mistrust were not identical. If a partner begins texting whenever plans change and every delay continues to be received as evidence of disregard, something besides notification may be active. That conclusion should not be rushed. Trust often lags behind changed behavior for good reasons. An apology does not instantly remove hurt. A new process may need months of successful repetition before anyone should be expected to loosen controls. The clue is not residue after one repair. It is a conflict that repeatedly selects new objects after previous objects have been addressed well enough that the emotional structure should have changed if those objects were carrying the whole problem.

res.p.000773By then people are usually building histories. An argument about this evening recruits the evening three months ago. A manager collects examples of an employee’s weak judgment while the employee collects examples of having sound decisions overridden. Two functions preserve screenshots, exception logs and stories establishing why the other cannot be trusted. Friends and colleagues become appellate courts. Each side’s archive can contain real facts, and the archive itself begins changing how new facts are perceived. The latest small event arrives already belonging to a much larger case. This is how apparently minor conflicts become difficult to settle even after people finally reach the issue underneath them: the proxy years generated consequences of their own.

res.p.000774Naming the actual conflict earlier would not necessarily have made it gentler. “I don’t think you are ready for this role” is harder to hear than another quarter of development language. “I no longer trust you” can threaten the relationship immediately. “We cannot support both priorities” creates a loser that another round of prioritization vocabulary might have kept temporarily invisible. “I do not want the life you want” may reveal an incompatibility that no amount of better communication can repair. The cheaper conflict is not cheaper because the sentence hurts less. It is cheaper because whatever pain follows is finally attached to something capable of changing the system. Trust can be rebuilt, bounded or discovered to be gone. Priorities can be chosen. An incompatible future can be faced. What becomes unnecessary is the expanding machinery required to keep producing substitutes.

res.p.000775This places a limit on the language of integration that matters for the whole book. Clarity does not guarantee reconciliation. Two interests can be accurately understood and remain incompatible. One party can be morally wrong. There are boundaries that should not be negotiated toward a midpoint, organizations in which one objective genuinely has to defeat another, relationships that cannot survive everything the people in them learn about what the other wants. Reality sometimes clarifies a conflict by making compromise harder. The value of accuracy is not that it creates harmony. It stops administrative debris from impersonating resolution.

res.p.000776That debris is easy to recognize once it has accumulated: meetings whose purpose nobody can state cleanly, rules built around behavior nobody wants to name, supervision with no imaginable exit condition, process revisions that cannot specify what success would permit people to stop doing, arguments repeatedly won on the facts while the relationship carrying the argument becomes worse. Every piece of the machinery may have a local justification. What makes me curious is recurrence without learning. If we implemented the requested change perfectly, what concern would still remain? What would have to become true for the review step to disappear? What claim are we trying to establish every time a new small event becomes evidence in the same old case? The answers do not have to be profound. “The deadline was missed repeatedly, and I need three successful cycles before I trust the new estimate” is excellent. The review now has a reason, evidence and some imaginable relationship to an ending.

res.p.000777The harder answers are harder precisely because they leave less compensatory work available. We keep revising the process because I do not trust you. We keep arguing about money because I am afraid. We keep treating this as an execution problem because admitting the strategy failed would require me to reverse something I publicly sponsored. We keep litigating small acts because I no longer experience the relationship as reciprocal. None of those sentences tells the people involved what to do next, and some may turn out to be wrong once spoken. What they do is return the work to the conflict producing it.

res.p.000778The conflict we actually have is usually cheaper, even when it is not easier, kinder, or more likely to end well. Fewer unrelated things have to carry it. Fewer people become interpreters. Fewer procedures administer what nobody has named, and fewer years of small disputes accumulate before the real disagreement can affect a decision. The surface problem can return to its proper size: a missed deadline is a missed deadline, a spending rule a spending rule, a process useful without being asked to restore trust by itself.

res.p.000779The disagreement may remain.

res.p.000780At least the work around it has stopped pretending to be about something else.

Chapter 16 — Toyota: Stop the Line

res.p.000781On a Toyota assembly line, a worker who encounters a problem does not have to wait for final inspection to prove that something went wrong. A part may not fit correctly. A tool may fail to complete its operation. The worker may see that the current unit is falling behind the production cycle and know that the next one will arrive before the work is finished. Toyota’s production system gives that person a way to make the abnormal condition visible while the evidence is still in front of them. An andon display can identify where help is needed, and equipment built around the same principle can stop automatically when it detects a problem. If the issue cannot be resolved within the available cycle, production can stop before additional work is completed around something already in doubt.

res.p.000782The stoppage has an immediate cost. People wait. Equipment sits idle. Fewer units move through the line during those minutes. Continuing can look cheaper because most of its cost has not happened yet. The questionable unit reaches the next station and another operation is completed. Then another. When somebody finally confirms the original defect, workers may have performed perfectly good work that now has to be removed merely to reach the first problem. The machine may no longer be in the state that produced it, and the operator’s immediate observation has become a memory. If several units passed through under the same conditions, inspection expands from one unit to a batch. A few minutes of stopped production have been avoided and replaced by diagnosis, rework, reconstruction and a larger population of things nobody is quite sure they can trust.

res.p.000783Toyota traces jidoka back to Sakichi Toyoda’s automatic looms, which were designed to stop when a thread broke rather than continue producing defective cloth until someone happened to notice. The later automotive system carried that design into a much more complicated environment. Workers and machines could signal abnormal conditions while the work, equipment and observation that exposed them were still together. There is nothing mystical hiding in the Japanese terminology. A known abnormality is allowed to interrupt production before additional operations make its consequences harder to contain.

res.p.000784The BDR call that opened Response was a knowledge-work version of the same problem. She had repeated observations from customer conversations while the campaign was still only copy, configuration and a consultant’s completed work. Her evidence was weaker than the performance data we could have collected after launch. It was also more useful because almost nothing depended on the original decision yet.

res.p.000785That asymmetry is easy to miss. A missed deadline proves more than a warning that the estimate looks optimistic. A churned customer provides cleaner evidence than an account manager saying the relationship feels wrong. Waiting can strengthen the case while shrinking the set of cheap responses. Early information earns attention partly because it is still allowed to be incomplete.

res.p.000786Organizations can produce this behavior while possessing all the expected machinery for candor. There is an escalation path. Projects can be marked red. Leaders ask for bad news early. The people using those systems are also watching what happens when somebody actually invokes them. If turning a project red immediately produces an inquiry into why the manager failed to keep it green, the status color reports something about the manager as well as the project. If someone raises a risk and spends the next hour defending why they do not already have the mitigation plan, they know how complete the next warning should be before they expose it.

res.p.000787Toyota’s workers operate much closer to the conditions they can interrupt. Knowledge work frequently separates observation from authority. An analyst can see that two numbers do not reconcile without having the authority to delay an executive presentation. An implementation team can recognize that a promise will create months of downstream work after another function has already presented the promise to a customer. Someone performing a recurring process can describe the manual steps that make it function while the budget required to change the process sits three levels away. People between the problem and the authority learn to compensate.

res.p.000788Toyota’s production vocabulary includes muri, or overburden, alongside muda, waste, and mura, unevenness. People can sustain expected output by absorbing demands that the operating design does not adequately account for. Continued success can make those conditions harder to see. The extra call gets made. The number gets checked twice. Somebody remembers the exception. Someone finishes the work at home. Output remains steady enough that plans built from output and headcount contain little evidence of what happened between them.

res.p.000789Heroics create particularly bad evidence. Sometimes the rescue is exactly what responsibility requires. Payroll still has to run. A customer may need help before the company has repaired the process that made helping them difficult. A healthcare worker cannot let a patient become the demonstration case for inadequate staffing. A parent does not withhold care from a child so that the household can observe the failure of its labor model. People often have to absorb the immediate problem first.

res.p.000790What happens afterward determines whether the extraordinary effort becomes information or precedent. A team works through a weekend and saves an important deadline. Everyone may deserve genuine thanks. If the next project receives the same scope, staffing and timeline, the organization has also treated the rescue as evidence that the original plan was feasible. The next group inherits an expectation built partly from labor that was never represented in the plan.

res.p.000791Human work cannot be treated as standardized production without producing nonsense. Drafts should contain bad sentences before editing. People need room to think aloud without every unfinished idea becoming an escalation. Creative work sometimes wanders for hours before the apparently irrelevant path produces anything worth keeping. Relationships contain periods of imbalance that cannot sensibly be classified as process defects. Children generate demands no resource plan will ever normalize. Grief lowers productivity without becoming waste.

res.p.000792The amount of later work that will depend on an uncertain thing gives us a better way to distinguish among these cases. A rough idea in a private notebook has almost no dependency attached to it. A questionable number about to enter a board presentation has much more. An uncertain product date discussed inside the product team can remain uncertain while everyone in the conversation understands its status. Give that date to a customer making a purchasing decision and another person can now organize real work around it.

res.p.000793Deadlines acquire the same weight gradually. A manager can privately believe Friday is feasible and be wrong at very little cost. Several teams schedule around Friday and the estimate has become part of their plans. Customer communications go out and changing it now requires expectation management. Leadership announces the date publicly and reversal acquires another cost because people have attached credibility to it. The original uncertainty may never have changed. The number of decisions depending on it has.

res.p.000794A sales organization can shorten the time required to close contracts while implementation spends months resolving commitments that made the shorter sales cycle possible. Product can increase release frequency while support absorbs the instability generated by each release. An executive can decide rapidly while several groups spend days determining how the latest decision interacts with everything already underway. Every local measure can improve while more labor accumulates elsewhere.

res.p.000795Maximum caution merely produces another failure mode. Toyota pairs jidoka with Just-in-Time, a system built around reliable flow and producing what is needed when it is needed. A production line that stops over every suspicion cannot produce efficiently. An organization that requires certainty before acting eventually stops learning through action at all. Some uncertainty should remain local and reversible. Other uncertainty is about to become an input into enough decisions that waiting for proof greatly increases the cost of being wrong.

res.p.000796Organizations need some stable account of ordinary conditions to make that judgment. A project described as taking six weeks cannot routinely be sold with a four-week deadline while the missing two weeks remain an exception. Either the six-week estimate is wrong or people are repeatedly supplying the difference. A role described as forty hours that reliably takes fifty-five hours is not generating extraordinary effort fifteen hours at a time. If an emergency workflow appears every month, the organization is using that workflow as part of normal operations regardless of what the process document calls it.

res.p.000797People can misuse any mechanism that grants concerns authority. Someone can call a preference a risk because risk language carries more weight. A team can escalate decisions it should be capable of making. A person opposed to a plan can continue finding new reasons to delay it after the relevant objections have been answered. Frequent false alarms can consume attention and eventually make real alarms easier to dismiss.

res.p.000798An employee who raises a concern and turns out to be wrong has still given management something different from an employee who acted irresponsibly. Collapse those categories and people will protect themselves by waiting for stronger evidence. Keep them separate and a manager can inspect the concern: what has the employee observed, what would become harder to reverse if work continued, what information would waiting provide, and what would be lost by pausing? The decision may still be to continue. The person closest to an emerging problem does not acquire a veto merely by being concerned.

res.p.000799People closest to work encounter many conditions in this early, ambiguous state. Most are not alarms. Some are ordinary variance, incomplete thinking, understandable discomfort or simply wrong judgments. A few are the first available evidence of assumptions, promises or capacity problems that other people are about to rely on. Whether those signals reach anyone while options remain cheap depends less on what the escalation policy says than on what people have learned will happen when they speak before they can prove the case.

res.p.000800The physical line at Toyota can stop while the worker, machine and questionable work remain together. Most organizations have to create that proximity deliberately. The people who notice something first need a practical route to people who can alter the plan before the rest of the organization has committed itself around the existing one.

res.p.000801They will sometimes be wrong.

res.p.000802Waiting until they cannot be wrong is how the cheaper problem disappears.

Chapter 17 — Havel: The Cost of Two Realities

res.p.000803Václav Havel’s greengrocer has vegetables to sell. In “The Power of the Powerless,” written in 1978, Havel imagines a shop manager who places the slogan “Workers of the world, unite!” in his window among the onions and carrots. The greengrocer is not presented as a committed ideologue eager to educate his customers about international proletarian solidarity. He displays what people in his position are expected to display. Refusing can attract attention, and attention from a one-party state with power over employment, education, travel, housing and family life is not an abstract inconvenience. The sign performs a social act quite apart from its literal words. The shopkeeper demonstrates that he knows what is expected of him and has no intention of making trouble.

res.p.000804People around him know how to read it. The customer buying potatoes does not need to believe the greengrocer spends his evenings contemplating Marxist internationalism. The sign belongs to the ordinary visual grammar of public compliance. Similar gestures appear elsewhere: attendance at expected events, participation in elections whose outcomes are not seriously in doubt, approved language in official settings, silence at the appropriate moment. Havel’s phrase “living within a lie” describes a society in which people can become skilled at performing public assent without privately possessing anything like the beliefs that assent appears to certify.

res.p.000805No single accommodation has to feel decisive. A person learns that one joke belongs among friends and another should not be completed after someone new enters the room. A writer knows which sentence can be published, which can circulate privately and which may attract the attention of the security services. Someone who privately despises an official ritual may help enforce it at work rather than risk being associated with the person who refuses. Fatigue can produce compliance as readily as conviction. So can ambition, concern for children, fear of losing a job or the reasonable desire not to turn every ordinary transaction into a confrontation with the state.

res.p.000806The visible record becomes misleading without requiring everyone producing it to be deceived. A person wondering whether anyone else shares their doubts sees slogans in windows, official ceremonies, newspapers repeating the authorized account and votes displaying overwhelming agreement. Private skepticism has no comparable street presence. Each individual act of caution can therefore become evidence, for somebody else, that caution is unnecessary because everyone appears to agree. A public consensus can contain large quantities of private disbelief and still look like consensus to anyone without access to the private conversations.

res.p.000807People find other ways to exchange what the official channels cannot carry. In Czechoslovakia, banned writing circulated in copied manuscripts. Apartments, kitchens and cottages became places for political and cultural conversation. Musicians developed audiences beyond official institutions. Relationships among people who trusted one another acquired functions that newspapers, publishing houses, unions and other public organizations could not safely perform. Havel’s own banned essays circulated through these networks, alongside work by people whose politics were far from identical. Catholics, former communists, liberals, artists and others could disagree about the society they wanted while sharing a more immediate experience: official institutions excluded too much of what they already knew about the society they inhabited.

res.p.000808Charter 77 made the discrepancy unusually concrete. Czechoslovakia had formally accepted international human-rights agreements and incorporated those commitments into its legal order. The Charter’s signatories compared those guarantees with censorship, prosecution, restrictions on employment and education, and the punishment of people attempting to exercise rights that officially existed. They did not need to begin by designing an entirely new political system. They asked the existing state to answer for the difference between its own declared standards and what people encountered when they tried to live under them.

res.p.000809The state treated the comparison as a threat. Signatories faced surveillance, detention, dismissal and prosecution. Havel was imprisoned. Jan Patočka, another early spokesperson for Charter 77, died after repeated police interrogation. Newspapers attacked a document most citizens had little practical opportunity to read for themselves. Under those conditions, keeping separate public and private accounts was not a quirky cultural habit. People had learned from experience that some descriptions could be spoken openly and others carried material penalties.

res.p.000810Nothing in an ordinary company deserves comparison with those political stakes. A manager softening a project update is not Havel’s greengrocer living under a Soviet-aligned police state, and calling an employee a dissident because they challenge a quarterly forecast trivializes what political dissent can cost. The resemblance appears only in the mechanics of maintaining different accounts for different audiences.

res.p.000811A project team can privately believe a date is unlikely and still present the project as green. Nobody has to falsify a number. The official plan may still contain a physically possible path to the date if two pieces of scope disappear, a vendor delivers on time and several reviews proceed without delay. The executive sponsor has already repeated the date to customers and has grown impatient with what they consider recurring execution noise. The team uses “green with manageable risks” in the executive materials while quietly making decisions against a different set of assumptions. Someone reserves additional capacity for the following week. Another person prepares a reduced-scope option. Sales is asked privately to avoid additional commitments. The formal plan remains intact while the people responsible for delivery prepare for the scenario they consider more probable.

res.p.000812From the executive level, the problem appears late. The status was “green” until it was suddenly “red.” From inside the team, people have been discussing the danger for weeks. Both accounts can be sincere. The early information existed, but the people holding it had not yet put it into the channel the executive reporting system treated as authoritative. They were working the problem, waiting for stronger evidence, protecting themselves from an escalation they expected to become an inquiry into their own competence, or hoping to solve the issue before exposing a commitment somebody senior had already made.

res.p.000813After enough repetition, official language stops carrying its ordinary meaning. Experienced employees know that “green” can mean the date remains theoretically achievable rather than likely. “On track” can mean nobody with enough authority has accepted a revised plan. A “risk” may refer to an outcome most of the team already expects but which has not yet received permission to become an assumption. New employees hear the words literally because nobody has yet taught them the local dialect.

res.p.000814They learn.

res.p.000815“Green” does not mean we are confident.

res.p.000816“Talk to Melissa before sending that to finance.”

res.p.000817“Legal owns the approval, but get product on board first.”

res.p.000818“The policy says thirty days, though nobody actually waits thirty days.”

res.p.000819The newcomer follows the published process and makes a mistake that an experienced employee would have avoided. The veteran fixes the immediate problem by explaining how things really work. Another person has now acquired the unwritten knowledge required to operate the institution successfully. The documentation remains unchanged.

res.p.000820This kind of knowledge can be genuinely valuable. Written procedures cannot encode every contingency, and organizations need people who understand history, personalities, exceptions and context. A company whose employees know only what appears in the policy manual would be extraordinarily brittle. The problem is narrower. People should be able to follow the institution’s own instructions without repeatedly creating failures that veterans prevent through oral tradition. If competent participation requires knowing when the official description should not be believed, the organization has made experience partly synonymous with fluency in contradiction.

res.p.000821The duplication can become extensive. A reporting system exists beside the spreadsheet people trust. The large meeting follows a smaller meeting in which the actual disagreement was settled. The documented process requires a sequence of unofficial conversations. Customer health looks one way in the dashboard and another way among the people who spend time with the account. Removing the unofficial systems abruptly can make things worse. Ban the private spreadsheet before correcting the official one and the accurate numbers may disappear. Force every disagreement into a room people have learned not to trust and they may stop voicing the disagreement altogether.

res.p.000822The existence of an unofficial system does not establish that the unofficial account is more accurate. Veterans can become attached to workarounds that give them status and influence. A private spreadsheet can contain wishful thinking. A pre-meeting can exclude the people most capable of challenging a comfortable conclusion. Employees who consider themselves sophisticated enough to “know how things really work” can become contemptuous of formal controls that were created to prevent exactly the behavior they prefer. Cynicism is not a privileged form of perception.

res.p.000823Havel’s language invites that mistake unusually easily. “Living in truth” is a flattering description to apply to oneself. Once somebody believes they are the person who sees through the official fiction, disagreement can become further proof of everybody else’s conformity. A dissenter can be courageous and wrong. If I insist a project cannot succeed under certain conditions and it succeeds under those conditions, the outcome belongs in my model. If I believe a leader suppresses dissent and later discover that several teams independently examined the issue and reached the same conclusion as the leader, I do not get to preserve my theory by declaring all of them captured by the system.

res.p.000824Private conviction requires correction too.

res.p.000825Havel’s own political history resists the satisfying story in which one man told the truth until a dictatorship collapsed. Charter 77 remained small. By 1989, Czechoslovakia was changing amid transformations throughout Eastern Europe: changes in Soviet policy, economic pressures, developments in neighboring states, student demonstrations, mass protests, long-established dissident networks and rapid political organization all entered the conditions under which the communist government lost power. Havel became a central figure in Civic Forum and was elected president in December. Compressing all of that into a story about the inevitable triumph of individual authenticity removes most of the causality and almost everybody else.

res.p.000826His first New Year’s address as president in 1990 did not place the entire moral burden safely on the former rulers either. He spoke about habits people had developed under the old order, including the distance between what they thought and what they had learned to say publicly. That did not make responsibility equal. A political prisoner, a frightened shopkeeper and a security official do not occupy equivalent moral positions merely because all participated in the same society. Power changes both the consequences a person can impose and the risks they incur by refusing participation.

res.p.000827Organizations distribute responsibility just as unevenly. An executive who punishes the first serious disagreement, the manager who begins softening every warning before carrying it upward and the employee who keeps the forecast they actually trust in a private spreadsheet can all contribute to the separation between formal and practical information. The executive controls much more of the environment in which the separation developed. The employee still changes who can see the accurate forecast by keeping it among trusted colleagues. Each adaptation solves a local problem while removing information from somebody else’s view.

res.p.000828Leaders sometimes try to repair this by changing the reporting system. Status definitions become more objective. Risk categories acquire thresholds. Dashboards gain fields intended to force earlier disclosure. Those changes can repair genuine defects in the old process. They cannot by themselves alter what employees have learned about the consequences of using it. If the first person to turn a project red loses standing, people will continue waiting regardless of how precisely red has been defined.

res.p.000829A new leader inherits the same difficulty. Announcing that candor is welcome does not erase the evidence accumulated under the previous leader. Employees watch what happens to the first person who brings unwelcome information. They watch whether an ugly forecast remains information about the business or becomes an inquiry into the forecaster. They notice whether dissent survives a meeting without reappearing later as a problem of attitude or alignment. People who spent years learning that the written process was ceremonial will continue relying on personal relationships until following the written process produces enough good outcomes to compete with what experience taught them before.

res.p.000830Trust in an official system is built from those repetitions. A bad forecast enters the executive deck and the forecaster keeps their standing. Someone challenges a sponsored strategy and the disagreement remains attached to the strategy rather than their loyalty. A team follows the published process and gets the result the process claims it will produce. A leader admits that a decision failed without constructing an explanation in which all causality flows toward execution below. The private spreadsheet loses part of its audience once the official one consistently contains numbers people can actually use.

res.p.000831The difference between privacy and duplicated reality should remain intact. People are entitled to private thought. A confidential draft is not a lie because it has not been published. A manager who knows layoffs are coming may have legal and ethical reasons not to answer a casual staffing question with everything they know. Friends are allowed to decline to publish every irritation they feel toward one another. Public and private speech will never correspond perfectly, nor should it.

res.p.000832The trouble begins when people must use one description to perform the institution and another to navigate it.

res.p.000833Then ordinary activity requires translation. People prepare the executive account and the working account. They explain the policy and then explain the exception required to function under it. New employees have to be taught which words do not mean what they appear to mean. Formal meetings become less informative as more disagreement migrates into pre-meetings. Leaders see surprising failures after the early evidence has been routed through channels they do not see. Experienced people become increasingly necessary, partly because they know how to bridge systems that would otherwise contradict one another.

res.p.000834The work can continue for a very long time that way. Good people are astonishingly capable of keeping incompatible descriptions operational at once.

res.p.000835Havel’s greengrocer does not remove the sign merely because he understands what it represents. Understanding the ritual can make him better at surviving it. Refusal carries consequences, and insight does not pay those consequences for him. In an organization, recognizing that the forecast is ceremonial or the process is broken likewise does not automatically give the person doing the recognizing enough authority to change either one.

res.p.000836Someone eventually has to alter what happens when the two accounts collide.

res.p.000837Until then, people keep both.

Chapter 18 — Jobs, Boyd and the Power of Compression

res.p.000838When Steve Jobs returned to Apple in 1997, he found a product line he later said he could not make sense of himself. Apple had accumulated roughly fifteen product platforms with numerous variants, each supported by some combination of customers, engineers, prior investment and internal history. Jobs spent weeks trying to understand what the company was selling and eventually reduced the roadmap to a grid simple enough to draw on a whiteboard: professional and consumer on one axis, desktop and portable on the other. Four product families. Much of the existing roadmap was cut, and engineering effort was concentrated on what remained.

res.p.000839Almost everything one could know about a computer was missing from that grid. Processor architecture did not appear. Neither did price, screen size, expansion, peripherals, industrial design, operating system or distribution. The four boxes were useful only if the distinctions Jobs retained organized Apple’s choices better than the larger taxonomy they replaced. A customer requesting a product no longer supplied enough reason by itself for the company to build one. An engineering team’s investment no longer guaranteed that its platform would survive. Jobs had chosen a small number of distinctions against which many other facts would be judged, and thousands of later decisions could now be made without reopening the entire question of what kind of company Apple intended to be.

res.p.000840Reduction of this kind is easy to mistake for simplification in the ordinary sense of making something less complicated. Apple’s products did not become technically simple because the roadmap had four boxes. The remaining engineers still had to solve complicated hardware, software, manufacturing and distribution problems. Customers still differed. Markets still changed. The grid moved complexity rather than abolishing it. People no longer had to preserve every product distinction at the level of portfolio strategy, which left them free to deal with enormous complexity inside the smaller set of products Apple had chosen to make.

res.p.000841By 1997 Jobs was applying the same kind of selectivity at a different scale. The long roadmap contained many legitimate facts about existing customers, markets and work already underway. He decided that some of those facts should no longer determine what Apple built. The four-box grid gave thousands of employees a smaller set of commitments they could use when deciding which projects, features and requests deserved attention. That concentration also had human consequences. Apple canceled products, including the Newton platform, reduced the roadmap dramatically and cut jobs. People lost work they had invested years in. A strategic reduction that looks clean on a whiteboard is experienced much less abstractly by the people whose work lands outside the boxes.

res.p.000842Jobs sometimes described the resulting clarity in terms that made the pain sound almost self-resolving. In a 1997 talk to Apple employees, he said the company had eliminated roughly seventy percent of its product roadmap and described even people whose projects had been canceled as newly excited because they finally understood the direction. Some probably were. The surviving record also contains an August message from Avie Tevanian, one of Jobs’s closest software leaders from NeXT and then Apple, saying that he felt exhausted and unmotivated and, for the first time in roughly a decade of working with Jobs, reluctant to challenge him when he disagreed. Jobs replied that Tevanian should keep challenging him because their disagreements helped produce better decisions. Both accounts describe part of the same working environment.

res.p.000843A leader who repeatedly makes strong reductions can save everyone around them enormous amounts of cognitive work. People learn what will survive the leader’s judgment and stop developing options that plainly will not. They learn which evidence belongs in a recommendation, which customer requests fit the company and which product compromises will be rejected even if they would save time. Eventually many decisions can be made without the leader present because other people understand the distinctions the leader would use. The same environment can make disagreement increasingly expensive if employees also learn that challenging the person who supplied the organizing model brings humiliation, exhaustion or career risk. The group can become faster partly by becoming better at anticipating the leader and partly by becoming less willing to transmit information that would force the leader to reconsider.

res.p.000844John Boyd spent much of his career studying situations in which people had much less time to discover that the model in their heads no longer matched events. Boyd was a U.S. Air Force fighter pilot, instructor and strategist. His early work included an Aerial Attack Study that formalized fighter tactics and, with mathematician Thomas Christie, Energy-Maneuverability theory, which provided a quantitative way to compare aircraft performance through relationships among speed, altitude, energy and maneuvering capability. Experienced pilots could sometimes perceive these relationships without being able to state them formally. Boyd’s work helped make them explicit enough to teach, analyze and use beyond one pilot’s intuition.

res.p.000845Boyd later developed his thinking far beyond air combat in the body of work associated with the Observe-Orient-Decide-Act loop. OODA is often reduced to a four-step instruction for acting faster than an opponent: observe, orient, decide, act, then repeat before the other side completes the same cycle. Boyd’s model of orientation was much denser. Previous experience, cultural traditions, analysis, synthesis and incoming information all influenced how a person interpreted what they observed and which actions appeared plausible. Feedback from action changed later observation and orientation. An Air University discussion of his work quotes Boyd calling orientation the focal point because it shapes the rest of the cycle.

res.p.000846The thirty-second decision therefore begins long before the thirty seconds. A pilot has spent years learning which changes in geometry deserve attention. A musician can hear an ensemble drifting sharp without examining every possible feature of the performance in sequence. An experienced negotiator can notice that a conversation has shifted from disagreement about terms to disagreement about trust while somebody newer is still trying to answer the last stated objection. The speed comes from distinctions already acquired, patterns already compared and prior judgments that have survived enough contact with reality to become available without full reconstruction every time.

res.p.000847Every such reduction discards information. A manager who has learned that a particular customer profile churns frequently may stop hearing what distinguishes the next customer who resembles it. An investor who recognizes the structure of a previous successful deal can give the similarities more weight than the differences. A physician’s pattern recognition can organize symptoms rapidly and still obscure an atypical presentation. An executive who has spent twenty years watching cautious teams underestimate themselves may become excellent at pushing ordinary caution aside and progressively worse at recognizing the person who has brought genuinely disconfirming evidence. Experience can preserve earlier errors along with earlier insight.

res.p.000848Boyd’s orientation never becomes a finished asset for that reason. New information has to retain some capacity to alter the patterns through which later information will be interpreted. A pilot who responds rapidly through a model that no longer fits the opponent is not helped by the speed. Boyd’s larger strategic work concerned both disrupting an adversary’s ability to make coherent sense of changing conditions and preserving one’s own capacity to adapt as those conditions changed. Two people can therefore make an equally fast decision from very different internal conditions. One recognizes a situation they understand well. The other stopped looking closely enough to notice that the situation had changed.

res.p.000849Executive work creates a slower version of the same problem. A product decision may arrive with abundant evidence and no shortage of intelligence. A major customer wants a feature. Broader research suggests most customers may never use it. Building the full version will delay two other projects. A narrower version might capture most of the immediate value. Sales expects delay to hurt the quarter. Product believes the larger danger is committing six months to something that may be peculiar to one account. Finance can add more scenarios and customer research can add more interviews, but no analysis will eliminate the eventual need to decide how much weight each kind of evidence should receive.

res.p.000850A useful recommendation exposes that weighting. “Build the narrower version” is not merely a shorter summary of the research. It can mean that the team believes one large customer does not yet constitute enough evidence of a broader segment, that delaying the other projects creates more risk than losing some value in the current deal, and that implementation cost should include the support burden that will continue after development ends. Someone who disagrees can now attack those assumptions rather than demand another round of analysis whose interpretation will still depend on them.

res.p.000851Shared models can produce something that looks almost like organizational instinct. A designer removes a shortcut without asking for executive review because everyone understands the kind of compromise the company rejects. A salesperson walks away from a prospect because the prospect falls outside a boundary the company has made clear enough to use locally. A product manager cuts a feature because the user problem has been defined well enough that the feature is recognizable as distraction rather than opportunity. People do not have to transmit the whole history behind every choice. They can coordinate through distinctions already held in common.

res.p.000852A false premise travels just as efficiently. If everyone has learned the wrong account of the customer, distributed decision-making allows hundreds of people to make locally coherent mistakes without waiting for centralized approval. A strategy can change in an executive presentation while sales continues qualifying opportunities according to the previous ideal customer, managers continue promoting people who resemble yesterday’s successful leaders and finance continues requesting the evidence the previous investment model required. People cannot replace years of learned distinctions at the same speed that somebody replaces a slide. Apple’s four-box grid was visually simple; rebuilding the company around it required canceled projects, changes to teams, manufacturing and distribution work, new hiring and repeated product decisions made under the new constraints.

res.p.000853Revision becomes harder as a model gets socially embedded. Jobs’s grid selected professional versus consumer and desktop versus portable as the distinctions worth preserving. Another strategist might have preserved price, distribution or operating system instead. Neither the raw market data nor the whiteboard itself determines which distinctions deserve priority. Two functions can review the same customer request and compress it differently. Sales sees evidence that a segment wants the feature. Product sees an unusually large customer pulling the company toward custom development. Finance sees uncertain return on six months of investment. Implementation remembers the last small exception that produced years of service work. Nobody has to be irrational for the disagreement to survive.

res.p.000854The argument improves once those reductions are stated with enough specificity that other people can test them. Treating one customer as evidence of a segment can be challenged by asking whether comparable customers show the same need. Treating delay to existing projects as the larger cost can be tested against the value and likelihood of the current deal. Treating implementation complexity as a one-time development expense can be compared with the support burden created by similar features. People may continue to disagree after all three questions. They will at least be disagreeing about the model assigning weight to the facts rather than competing to add more facts to a pile whose governing assumptions remain implicit.

res.p.000855A strong model therefore has to be held in an awkward state. People need enough confidence in it to act without rebuilding their worldview at every decision, and enough distance from it to notice when exceptions have become common enough to challenge the category itself. The professional/consumer distinction changed as computing changed. A fighter pilot trained against one aircraft encounters another with different capabilities. A management practice that works beautifully with ten employees can become ridiculous with ten thousand. People who reopen every model at every decision never gain much from experience. People who refuse to reopen any model eventually spend their experience defending yesterday.

res.p.000856The difficult cases often present as a choice between those two failures. One team keeps every branch alive because choosing a frame feels premature; analysis grows and action recedes. Another team can explain the situation in three sentences and no longer notices evidence that does not fit those sentences. The first carries too much complexity into every decision. The second has discarded the wrong complexity and will discover it later through consequences.

res.p.000857Good compression does not make a subject small. It makes judgment visible. Jobs’s four boxes did not contain Apple’s technical complexity. Boyd’s orientation did not contain every fact in a cockpit. A recommendation does not need to reproduce the research that supports it. Each leaves enormous amounts of detail elsewhere while exposing which distinctions will organize action now. Other people can then bring detail back when it challenges the reduction.

res.p.000858That last condition separates a working model from a doctrine. The sales forecast people privately believed at Apple needed a route back into the official number. A pilot’s new observation has to be capable of changing the pattern through which the next observation will be understood. An employee labeled “not strategic” needs some conceivable sequence of evidence that can retire the label rather than merely generate new interpretations of it. A strategy that cannot be revised by information from the people using it has ceased to provide cheaper judgment. It has begun charging for every fact that no longer fits.

res.p.000859We cannot avoid reduction by being more thorough. The attempt to preserve everything is itself a decision to leave the weighting for later, often to whoever has the most authority in the room. Nor can we make reduction safe by finding the perfect model. Every useful compression will eventually encounter something it handles poorly. People choose what to preserve, act through that choice, observe what follows and revise when the discarded material starts affecting outcomes.

res.p.000860Jobs could make a complicated product problem legible enough that thousands of people could move in roughly the same direction. Boyd studied how people could act under rapidly changing conditions without deliberating from zero each time. Both cases show how much speed becomes available once somebody has decided which distinctions deserve to survive.

res.p.000861They also leave no reason to confuse speed with wisdom.

res.p.000862A con artist can read a room, discard irrelevant detail and select the right cue with extraordinary accuracy. A predatory institution can build a clear strategy and distribute it so effectively that employees barely need explicit instructions. Coordination can become remarkably efficient before anyone has asked whether the thing being coordinated should be done.

Chapter 19 — Alignment Can Make You Dangerous

res.p.000863A manager who has worked with the same people for several years knows things about them that no competency model will contain. One employee enjoys being handed the impossible problem. Another needs time before answering a request. Someone challenges dates immediately. Someone else will agree to almost anything if the request arrives wrapped in trust and gratitude. A manager does not need manipulation training to acquire this knowledge. Paying attention to people is part of managing them.

res.p.000864The knowledge does not change character when it is used badly. A manager facing an ugly deadline can choose the person least likely to refuse. The request itself can be completely truthful. The customer really is waiting. The work really is important. The employee really is unusually reliable. Nothing requires a threat. “You were the first person I thought of because I know it will be done right” may be sincere praise and an extremely effective way to place obligation on somebody who experiences trust as a debt. If the manager already knows that about the employee, the request contains more than the words spoken aloud.

res.p.000865The same attentiveness can support care. Parents learn when a child can hear criticism and when hunger has made the conversation pointless. Partners learn when the other person needs twenty quiet minutes after work. A good manager learns who wants blunt feedback, who thinks aloud, who needs to sit with criticism before responding and who will interpret silence as disapproval. Familiarity reduces the number of interactions that have to begin from zero.

res.p.000866An abusive partner can know the other person just as well. They can learn exactly how much anger produces compliance, which insecurity can be activated without looking like a threat and which subject will end an argument quickly. A salesperson can notice that a prospect is frightened of appearing indecisive and exploit that fear rather than help them evaluate the purchase. A charismatic leader can recognize who wants proximity, who needs praise and who has begun to doubt. Accuracy about another person does not tell us what we are entitled to do with the accuracy.

res.p.000867Wells Fargo spent years producing a much larger version of the same problem. Cross-selling additional banking products to existing customers was not inherently corrupt. A customer who genuinely needs checking, savings, credit and other financial services may prefer one institution, and a bank has an obvious commercial interest in deepening that relationship. Inside Wells Fargo’s Community Bank, however, aggressive sales targets and managerial pressure increasingly rewarded employees for producing accounts whether or not customers had asked for them. The Consumer Financial Protection Bureau’s 2016 action described widespread unauthorized deposit and credit-card accounts. Wells Fargo’s own board investigation later traced misconduct through years of sales pressure, distorted incentives, weak escalation and management failures.

res.p.000868Nobody needed to distribute an instruction reading create accounts customers do not want. Employees had targets. Managers asked about those targets. Branches were compared. People watched colleagues succeed and watched others struggle when numbers remained low. Managers had numbers of their own and transmitted the pressure downward. Some employees refused, left or were fired for sales-integrity violations. Others opened unauthorized accounts, moved funds, created records and found additional ways to generate the activity their managers were measuring.

res.p.000869There was real skill inside the system. A branch manager could identify the employees who reliably produced accounts. Employees learned how internal procedures worked and which actions would move the sales numbers. Regional leaders could compare branches and coach toward whatever the higher performers were doing. The organization could gather accurate local information, distribute techniques, establish goals, observe results and tighten performance. Customers’ actual desire for another product had less influence over those local decisions than the bank’s public description of the relationship implied.

res.p.000870Nothing about effective feedback loops prevents this. Nothing about clear principles prevents it. Recognition can make the employee faster at spotting the customer most likely to accept another product. A tightly defined metric can remove ambiguity from the workday. Shared language can make coaching more precise. A manager can align an entire branch around the same target and reduce the need to restate priorities constantly. People can become exceptionally efficient at work they should not be doing.

res.p.000871The bank eventually paid enormous regulatory and legal costs, changed leadership and sales practices, and spent years dealing with the consequences. Those later events make the failure easy to classify. Much harmful work never produces a public record large enough to settle the classification for everyone. A manager who repeatedly selects the employee least likely to refuse can instead develop a reputation for getting difficult projects over the line. Their team delivers. Senior leaders give them harder assignments. Employees who dislike the operating environment transfer or stop receiving the assignments through which advancement occurs. After several years, the remaining group may genuinely contain an unusual concentration of people who enjoy the pace and identify strongly with the team.

res.p.000872Output cannot tell an observer how that population formed. Some people may love the work. Some may value rapid advancement enough to accept the hours. Some have lives in which intense bursts of effort are manageable. Others may have learned that visible reluctance changes how they are perceived. The team can post extraordinary results under all four conditions.

res.p.000873Successful outcomes make these questions harder to keep separate. A founder works constantly and becomes enormously wealthy, and the hours begin to look like an explanation. An executive imposes severe discipline on a failing division and the turnaround makes every objection to the method look soft in retrospect. A political leader defeats opponents who repeatedly underestimated them, and ruthlessness acquires some of the prestige of the victories. The capability that helped produce an outcome may be real without supplying a moral defense of every behavior nearby.

res.p.000874Steve Jobs belongs uneasily in exactly this territory. The previous chapter can credit his ability to narrow a product strategy and his unusual eye for product quality without requiring us to turn the working conditions around him into part of the recipe. The Macintosh does not become less consequential because members of the team worked punishing hours. The existence of a consequential Macintosh does not make those hours intrinsically good. If one judgment is used to certify the other, biography has replaced analysis.

res.p.000875Havel’s dissent does not establish dissent as a guarantee of truth. Toyota’s production system can teach us something about surfacing abnormal conditions without settling every moral question involving labor or corporate power. Boyd’s work on orientation remains available to anyone capable of learning it, including someone whose objectives we would oppose. Mechanical advantage does not arrive with an ethical use license.

res.p.000876A company trying to exploit customers can read every chapter in this book. It can define principles so employees stop reopening decisions management wants closed. It can simplify communication until everyone knows which outcome takes precedence. It can build fast feedback around conversion and retention. It can train employees to recognize which customers are vulnerable to which appeals. It can reduce the hesitation that once made a salesperson stop before pushing someone into a bad purchase.

res.p.000877Companies already do milder and harsher versions of this constantly. Call centers learn which language retains people who are trying to cancel. Sales organizations test messages and sequences against response rates. Gambling products measure which rewards and prompts keep people engaged. Social platforms learn which material holds attention. Political campaigns test messages for fear, anger, donation and turnout. People observe behavior, alter the system and observe again.

res.p.000878The resulting metric can describe the improvement perfectly. Retention rose. More subscribers stayed. Session length increased. Donations arrived. The customer who spent another ten minutes trying to cancel appears only if somebody includes those ten minutes in the account. The person who kept gambling appears differently depending on whether the record is built around engagement or around the life in which the engagement occurred. A campaign can accurately report that an angry message outperformed three alternatives without recording what millions of recipients experienced while supplying the performance data.

res.p.000879People choose the boundary of the measurement before the number appears. A sales process looks excellent when evaluation ends at signature. A warehouse can look faster when the clock stops at the loading dock. An executive appears extraordinarily decisive if nobody adds the time six teams spend translating each decision into something executable. A product remains cheap when pollution, dangerous labor or disposal occurs outside the part of the transaction its price was asked to represent.

res.p.000880People with greater authority can experience these omissions more easily because other people perform the accommodation. An executive fixes a launch date. Teams rearrange work, postpone less visible projects and compress reviews. From the executive’s position, the sequence can look like a decision followed by execution. A parent announces that the family is leaving at eight. Somebody else turns that sentence into breakfast, clothing, backpacks, missing shoes and the child who reveals at 7:54 that school requires a poster board. Setting a direction may be completely legitimate. Announcing it and carrying it out are still different quantities of work.

res.p.000881Other people can supply information about the difference. An employee says the weekend requests have stopped feeling exceptional. A partner says the household keeps reorganizing itself around one person’s career. Someone executing a beautifully simple strategy explains the three undocumented processes required to make it work. A customer leaves after the sales dashboard recorded a successful conversion. Leaders can alter what they are doing after hearing this, or preserve the existing account by classifying the objection.

res.p.000882High standards always produce complaints.

res.p.000883People who cannot handle the culture leave.

res.p.000884The employee lacks commitment.

res.p.000885The customer was never a good fit.

res.p.000886The critic does not understand the full context.

res.p.000887The explanations may sometimes be correct. Their convenience appears when each possible contradiction already has a category that prevents it from changing the underlying judgment. A manager who interprets every objection as resistance can retain an extraordinarily coherent view of the organization. Fewer and fewer events require revision because disagreement itself has become confirming evidence.

res.p.000888Internal discomfort can interrupt that closure before anybody else does. A manager knows the reliable employee has already covered too many weekends. A salesperson notices that the customer does not really understand the agreement they are about to sign. A leader realizes that the only people still describing the strategy enthusiastically report directly to them. A parent hears themselves using a tone they hated when they were young.

res.p.000889The discomfort may be misleading. Guilt can attach itself to perfectly responsible decisions. Anxiety can make ordinary authority feel abusive. Managers sometimes need to ask for extraordinary effort. Parents sometimes need to carry a screaming child out of a store. Governments cannot wait for unanimous consent. A company can make a decision that will cost people jobs and still be choosing the least destructive available option.

res.p.000890Someone can make the right decision and retain awareness of who pays for it.

res.p.000891A manager asks an employee to work Saturday. The customer need may be real, the deadline consequential and the request reasonable under the circumstances. The employee’s plans do not vanish when those facts become true. If the manager knows this employee has difficulty refusing them, that knowledge belongs in how the request is made. “Adults are responsible for their own boundaries” contains a real principle. It becomes convenient when invoked only after deliberately selecting the adult whose boundary is easiest to cross.

res.p.000892The employee’s yes is not meaningless. Neither is the manager’s influence over the conditions in which it was given. Human beings persuade one another constantly. We appeal to affection, loyalty, ambition, shared standards and identity. A rule treating influence itself as suspect would make ordinary relationships impossible. The history around the request tells the employee what the words alone cannot. What happened the last time they said no? Did the relationship remain warm? Did the next opportunity still come? Was the refusal later remembered as a lack of commitment? Have other people declined requests without acquiring an invisible mark against them?

res.p.000893Employees answer those questions while deciding how optional an “optional” request actually is.

res.p.000894Written policy works the same way. A company can guarantee protection for people who raise concerns. Employees look at what happened to the last three people who did. A family can insist that everyone is free to speak honestly while preserving years of evidence about what follows unwelcome honesty. Governments can guarantee rights whose exercise remains dangerous. People do not need cynical personalities to trust consequences more than declarations.

res.p.000895An organization can become extremely smooth once consequences are predictable enough. Employees know which behaviors advance careers and which produce trouble. Managers learn what senior leaders want before being asked. Meetings become shorter. Dissent rarely delays execution. People who cannot or will not adapt either leave or lose influence. New managers are promoted from the population that remains and begin teaching the same expectations without needing explicit instruction.

res.p.000896Smoothness alone tells us almost nothing.

res.p.000897Bureaucracy can consume human lives with pointless procedure. A consensus rule can let one unreasonable person immobilize everyone else. Leaders who refuse to decide create uncertainty that everyone below them has to administer. Meetings, approvals and duplicated work should not be preserved merely so an organization can boast that it has retained friction.

res.p.000898The source of the friction changes the decision. Duplicate data entry can be removed. An approval that outlived the risk it controlled can disappear. A meeting that no longer performs useful joint work can be canceled. Another person’s ability to say that a decision will harm them belongs to a different category. Their objection may not prevail. A company sometimes has to close a plant. A manager sometimes denies a request. A government can enact policy that millions oppose. A parent eventually puts the screaming child in the car.

res.p.000899Other people do not acquire veto power merely by becoming visible.

res.p.000900Their costs also do not stop existing when someone else possesses the authority to choose.

res.p.000901The manager with the Saturday deadline still has several choices. Miss the date. Reduce scope. Do part of the work personally. Ask someone else. Offer compensation and recovery time. Ask the reliable employee after making refusal genuinely survivable. Perhaps the employee still says yes. Perhaps they want the opportunity, care about the customer or simply decide that this weekend is worth sacrificing.

res.p.000902On Monday, the project record will show that the deadline was met.

res.p.000903The employee’s family may have spent Saturday without them.

res.p.000904Both statements can be true without forcing a verdict about whether the decision was right.

res.p.000905Only one of them fits naturally in the status report.

Chapter 20 — Ubuntu and the System Boundary

res.p.000906On Monday, the project record is accurate. The deadline was met. The customer received the work. Whatever risk had threatened delivery can be closed. The employee who worked Saturday may also have canceled plans, left a partner to absorb the day alone or simply spent an afternoon doing work they would rather not have done. Perhaps they chose it willingly. Perhaps nobody at home was especially upset. None of that changes the project status. It does change the amount of human activity required to produce the status.

res.p.000907No useful record can contain everything. A project tracker should not become a biography of every person who touches the work. A business cannot calculate every remote consequence of every decision before acting. Families do not conduct causal audits before ordering dinner. Models become useful partly by excluding most of reality and retaining what the people using them need to know. Trouble begins when the excluded portion contains work or cost that would change the judgment if the people making the decision could see it.

res.p.000908A finance team can improve its expense process by requiring better information at submission. The old reports arrive with missing project codes, inconsistent categories and receipts that somebody in finance has to chase. A new form makes the fields mandatory and catches errors before submission. Finance saves two hours every Friday. Four hundred employees now spend another two minutes on each report. Nobody has manipulated a metric. The process may still be better. The company has to decide whether it removed work, redistributed work or did some of each. Measured inside finance, two hours disappeared. Measured across the people completing the form, the arithmetic is different before anyone has argued about which arrangement they prefer.

res.p.000909Organization charts encourage this kind of accounting. Sales can improve conversion while implementation receives customers whose commitments require more work. Support can reduce average handle time by asking customers to perform more troubleshooting themselves. Procurement can lower component cost while failures increase later in the product’s life. An executive can remove a management layer and discover months afterward that much of the work performed by those managers was still necessary; directors now spend more time coordinating, employees spend longer finding decisions, and somebody else has inherited the conversations the old structure used to hold. Sometimes that redistribution is exactly what people wanted. Customers may prefer self-service. Employees may have been unnecessarily dependent on a manager for decisions they were capable of making. A specialist may perform something faster and better than the generalist who used to do it. Following the work beyond the organizational box is the only way to know which happened.

res.p.000910A restaurant kiosk makes the distinction easy to see. For many customers, ordering becomes faster. The order reaches the kitchen without another person retyping it. People can inspect the menu at their own pace. Fewer employees have to stand behind a register during the rush. Another customer cannot find the ingredient modification they need. A parent coordinates three children through an interface that does not notice which child is currently changing their mind. Someone with a visual or motor impairment encounters a screen designed around different assumptions. An employee who used to take orders now spends part of the shift helping the smaller population for whom the kiosk does not work. The kiosk can be an excellent system. “Fewer labor minutes at the register” is simply too small an account to establish that by itself.

res.p.000911Modern convenience is full of legitimate transfers. I can tap a screen and have dinner brought to my house because other people receive the order, cook the meal, package it, travel to the restaurant, wait, drive to me and find my door. I pay partly to move that work out of my evening. The service has created real value for me. Division of labor lets all of us live inside arrangements no individual could reproduce alone. I do not grow wheat, mill flour and build an oven every time I want bread. The baker does not manufacture the truck that supplies the bakery. Families, markets, governments, professions and friendships continually move work toward people who can or will perform it.

res.p.000912The driver still has to find parking.

res.p.000913A price can represent some of that transfer remarkably well. A delivery fee reminds me that another person’s time is involved. Wages tell a company something about what labor costs. Supplier bids allow strangers to coordinate complicated production without needing intimate knowledge of one another’s circumstances. A price contains only what the people setting it have enough power and opportunity to put there. An employee who has difficulty refusing a Saturday request may never price the lost afternoon into their salary. A town downstream from a factory cannot necessarily add future health costs to the manufacturer’s invoice. Unpaid care for children or aging parents creates time and stability that appear all over other people’s economic lives without generating a corresponding line item. A customer who spends forty minutes resetting an account has contributed labor to the company’s service process even though the company will never classify them as service staff.

res.p.000914There is no defensible boundary large enough to include every causal consequence. Follow a sandwich far enough and eventually the analysis contains agriculture, land, water, refrigeration, transport, fuel, packaging, finance, immigration policy and the lives of thousands of people. Nothing gets decided if moral seriousness requires a complete map of the planet before lunch. The boundary can still include the people whose participation substantially produces the result and the costs likely to alter the choice if they become visible. The implementation team belongs in the evaluation of a sale whose terms they will have to deliver. Childcare belongs in a household decision about a promotion that changes when a parent will be home. The customer’s hour spent repairing a defect belongs somewhere in an account of the service experience. The employee’s Saturday belongs in the manager’s understanding of what it took to meet the deadline.

res.p.000915South Africa’s transition out of apartheid placed the Southern African term ubuntu into a constitutional setting where the boundaries around human standing had been enforced brutally for generations. The postamble to the 1993 Interim Constitution invoked ubuntu alongside reconstruction, reparation, understanding and reconciliation after apartheid. The term does not have one uncontested philosophical definition. Scholars have described different traditions of its use: as an ethical quality, a worldview and an account of personhood formed through relations with other people. Desmond Tutu rendered one aspect of that tradition by describing his own humanity as bound up with the humanity of others.

res.p.000916That formulation does not require two people to become one. The manager and employee remain different people with different authority, obligations, families and interests. The manager’s deadline enters the employee’s Saturday. The employee’s decision to work becomes part of the manager’s ability to meet the commitment. A description that stops at “the manager delivered the project” has cut the relation out of an outcome produced partly through the relation.

res.p.000917Ubuntu appeared again in South African constitutional reasoning during S v Makwanyane, the 1995 Constitutional Court case that invalidated the death penalty. The unanimous judgment rested on constitutional protections including life and dignity and the prohibition on cruel, inhuman or degrading punishment; several justices also drew on ubuntu while considering dignity, community and the treatment of offenders. This history makes the word difficult to shrink into corporate vocabulary without losing most of what gave it weight. It entered a legal and political conversation about the society that would follow apartheid, about dignity in the treatment even of people convicted of terrible crimes, and about relations among people after a state had spent decades assigning radically unequal human standing.

res.p.000918Management culture is very good at finding language like this and sanding off everything capable of interfering with management. Ubuntu becomes “we’re all connected.” A photograph of diverse coworkers appears underneath. The company talks about collective success and mutual dependence. The word has been imported without changing who can impose obligations on whom.

res.p.000919A manager can invoke community while deciding whose weekend will be used to protect a deadline. A company can call itself a family and ask employees to draw on the moral vocabulary of kinship when the organization needs sacrifice. Actual families contain obligations that ordinary contracts cannot describe: people care through illness, childhood, disability, old age and periods when repayment is impossible. An employer that adopts those claims selectively gains a useful arrangement. The employee is encouraged to treat the company like family when something needs to happen on Saturday. The company retains the ordinary legal and financial freedom to treat the employee as a position that can disappear when circumstances change.

res.p.000920Families themselves can perform the same trick. One sibling becomes the reliable person caring for an aging parent. Everyone sincerely believes that family members should care for one another. The reliable sibling schedules the appointments, handles the calls, rearranges work and notices what needs doing next. Other siblings may visit, contribute and express genuine gratitude. They may also construct their weeks around the assumption that the reliable sibling will cover whatever remains. The caregiver may value the role deeply and reject any description that reduces it to unpaid labor. Care would become grotesque if every hour had to be translated into a wage before we could recognize it as real. The family can honor the gift without pretending that gratitude distributes the work evenly.

res.p.000921Relational language becomes dangerous when the person with the greater capacity to impose an obligation also gets to define what belonging requires. “We are all in this together” means something different when everyone can make comparable claims on the group. If togetherness always explains why one person should absorb the cost and never constrains the person assigning it, the language has become decoration around an asymmetry.

res.p.000922Individual standing survives the relation. A person’s exhaustion, dissent, privacy and desire do not disappear when we acknowledge that people depend on one another. South Africa’s Interim Constitution invoked ubuntu alongside enumerated protections for individual equality, dignity, privacy, expression, association, movement and labor rights rather than replacing those protections with a general instruction to submit to the community. A relational account that allows “the whole” to consume individual people reproduces the problem from the opposite direction.

res.p.000923Ordinary executive decisions live inside that tension. A company considering a return-to-office policy may have serious reasons for doing so. New employees may be learning more slowly. Apprenticeship may have weakened. Some forms of collaboration may work poorly through scheduled calls. Office space may sit largely unused. Employees have commute times, childcare, disabilities, elder care, housing decisions, friendships at work, concentration needs and several years of experience with another arrangement. Leadership eventually has to choose. Some employees may discover that returning improves their work. Some may leave. Others may spend ten additional hours each week commuting. A company can decide that the organizational gains justify those hours without pretending the hours ceased to exist when the policy was announced.

res.p.000924The same accounting can preserve friction that should remain. A hospital adds another check before a high-risk medication is administered. Nurses spend more time and throughput falls slightly. Removing the check would improve a local speed measure while transferring additional risk to the patient whose body will receive the drug. A decision to retain the step does not reveal an institution hostile to efficiency. People have decided that the labor being removed and the risk being added do not belong to equivalent categories.

res.p.000925“Stakeholder” can make all of this strangely bloodless. Employees, customers and communities become boxes on a diagram with influence and interest scores. The diagram may be useful. A person goes home after the meeting. They depend on wages. They become tired. They use the product with a body that may differ from the one the designer imagined. They breathe the air around the facility. They wait on hold. They care for the employee who returned home depleted. Organizational categories help us reason about groups; they become dangerous when the category is treated as a complete substitute for the people from whom it was abstracted.

res.p.000926Households can build remarkably efficient systems on the same omission. One person decides meals will be planned Sunday, groceries ordered once, lunches prepared ahead and the kitchen cleaned during cooking. The routine works beautifully. Someone still notices that a child has stopped eating the scheduled lunch, checks what is actually in the pantry, remembers that Thursday is different because of practice, washes the container needed tomorrow and notices the milk is low despite what the inventory says. A household can become much more organized because one member has become an excellent operations department. Everyone else experiences the smooth routine. The person maintaining it experiences all the little observations that keep the routine smooth.

res.p.000927Workplaces produce similar illusions when they measure artifacts rather than coordination. Cancel recurring meetings and a calendar report shows fewer meetings. The people who had used those meetings to resolve dependencies may replace them with one-off calls, longer message threads and private coordination. That could be an improvement: two people exchanging information for ten minutes may be far better than twelve people meeting for an hour. Counting canceled meetings cannot tell us. Someone has to follow the coordination and see whether it became smaller or simply fragmented into activities that no longer carry the label “meeting.”

res.p.000928The person receiving transferred work often has less institutional authority than the person whose workflow became cleaner. Executive assistants absorb schedule volatility. Customer service teams absorb ambiguity produced elsewhere. Nurses compensate for software built around requirements that do not correspond neatly to clinical work. Teachers complete administrative work after students leave. Junior employees reconcile numbers, repair slides and chase approvals before senior people encounter the output. Competence can make somebody else’s system look much cleaner than it is.

res.p.000929The same problem extends beyond employees a company can see. Procurement may move production to a supplier offering the required quality at a much lower cost. The ordinary management report contains unit price, defect rate, lead time and reliability. Workers at the supplier have bodies, and the place around the factory has water and air. No company can investigate every supplier relationship without limit. Global production would become impossible under a standard of complete causal knowledge. Companies still decide which labor and environmental conditions suppliers must report, what gets audited, which conditions disqualify a source and how skeptical people should become when a price seems possible only because something important is absent from the quote.

res.p.000930Individuals face the same limit every time we buy something. I cannot reconstruct the production history of every shirt, phone, tomato and bottle of shampoo I use. A moral theory that requires complete knowledge before ordinary participation becomes a theory for paralysis or hypocrisy. There is still a difference between accepting a reasonable limit to what I can know and arranging my convenience around systematically not knowing who has to bear the costs.

res.p.000931Neither individual freedom nor communal obligation resolves these problems alone. “They agreed.” “They took the job.” “They clicked accept.” “They can leave.” Sometimes one of those facts settles the question. Sometimes it merely identifies the formal mechanism through which a person with fewer alternatives entered the arrangement. “The company needs this.” “The family needs this.” “The country needs this.” Those claims can also describe real collective necessities while giving the person speaking for the collective permission to consume individuals without sufficient constraint.

res.p.000932The manager needs the project delivered. The employee has a Saturday.

res.p.000933The company may need to reduce headcount. The people losing their jobs will still have rent, families and professional identities after the meeting.

res.p.000934The customer wants an inexpensive product. The people who manufacture it are not themselves components of the product.

res.p.000935A parent needs the family out the door. The other parent is not merely the mechanism through which everyone becomes ready.

res.p.000936Holding both facts in view does not guarantee a kinder decision. A plant may still close. The employee may still work Saturday. The cheaper supplier may still be used. The family may still leave at eight while one person is unhappy about it. Decisions do not become ethical only when nobody loses anything.

res.p.000937Some improvements become easier to recognize once transferred burden stops counting as elimination. A software change that saves finance two hours and saves employees twelve is excellent. Removing a management layer can genuinely reduce work when the people below it receive clearer authority rather than inheriting orphaned coordination. A delivery service can return an evening to a customer while giving a driver predictable, well-compensated work. A family routine can make mornings easier for everyone rather than making one person wake forty-five minutes earlier to manufacture everyone else’s ease.

res.p.000938On Monday afternoon, the manager can thank the employee who worked Saturday and leave the matter there. They can praise the rescue so enthusiastically that both people understand it as a career investment. They can ask whether the weekend caused trouble, hear “no, it was fine” and accept the answer. Adults are allowed to mean that. People are allowed to love their work, volunteer for difficult things and decide that a particular sacrifice was worth making.

res.p.000939The manager can believe them and still look at the previous three months.

res.p.000940If the same person has rescued several deadlines, their willingness has become part of the planning model whether anybody intended it to. The next plan can reduce scope earlier, rotate emergency work, preserve recovery time, change the estimate or admit that staffing does not support the promises being made. The employee does not have to become a symbol of exploitation in order for management to stop treating their generosity as ordinary capacity.

res.p.000941The status report can remain green. The customer received what was promised.

res.p.000942Later that week, somebody can also put three recent weekend rescues into the planning conversation, including the fact that the same person supplied two of them.

res.p.000943The next deadline no longer gets to assume Saturday for free.

Chapter 21 — Outlier Results Are Not the Test

res.p.000944I was born in 1990, which turned out to be an absurdly good year to be born if you were going to become interested in computers.

res.p.000945They were already present when I was a child, but they had not yet disappeared into the walls. You could still feel them changing. The family computer was a thing you approached deliberately. The internet made noises before it arrived. Websites had seams. You could right-click, view the source and see that whatever appeared on the screen had been made out of text somebody else had written. The web and I grew up together at roughly the right distance: close enough that none of it felt alien, early enough that much of it was still available to be taken apart.

res.p.000946I built my first webpage before I could write a compelling paragraph. It was a Pokémon fan site, and “built” deserves some qualification. I shamelessly pilfered images, information and probably entire chunks of markup from other sites. Intellectual-property doctrine was not high on my list of concerns. I wanted a page on the internet, so I figured out enough HTML to make one exist. The page taught me something school assignments rarely did: if I wanted a thing I did not yet know how to make, I could often inspect another thing that worked, copy what I understood, break it, repair it and slowly acquire the missing language.

res.p.000947At fourteen I built my first LAMP server out of a scavenged Apple Power Mac G3 tower in the basement. LAMP stood for Linux, Apache, MySQL and PHP, a combination of free software that powered an enormous amount of the early web. The G3 used Apple’s old PowerPC processor architecture, years before the company moved its Macs to Intel chips and later to its own processors. None of this was remotely exotic in the world of teenage computer obsessives, but it was unusual enough that there was no adult standing beside me with a curriculum. I wanted a machine in my house that could serve webpages and run database-backed software. I had access to an old computer. The rest was documentation, forums, failed installations and trying again.

res.p.000948School gave me every reason to expect a conventional academic future. I attended a prestigious high school and was near the top of my class in many subjects. My SAT score was close to perfect. I could do the thing American education is structured to reward: absorb material quickly, perform well on tests and produce evidence that I belonged on the next rung.

res.p.000949Then I entered college during an acute mental-health crisis.

res.p.000950By the time I emerged from the worst of it, the ladder looked very different. I was a college dropout. I married at twenty. We had a child, with another on the way. The person who had once seemed built for the clean academic route now had rent, groceries, diapers and a résumé with a conspicuous hole where the expected credential was supposed to be.

res.p.000951I spent those years first in a busy fast-casual kitchen and later in the onsite cafeteria of a large pharmaceutical manufacturer. In the second job, I did not actually work for the company whose building I entered every day. I worked for the food-service contractor operating inside it, in a unionized environment with another layer of rules, responsibilities and authority wrapped around the ordinary work of feeding people. A request could be simple in the kitchen and complicated everywhere around it.

res.p.000952The first kitchen had already taught me the physical logic of coordinated work. A lunch rush is a project with an unforgiving deadline. Prep creates capacity before demand arrives. Stations depend on one another. Several components have to finish at roughly the same time or the customer experiences the delay as a single failure regardless of which station caused it. Someone can be individually fast and still make the whole kitchen slower by working in the wrong sequence. When one station goes under, everybody nearby starts making decisions around the constraint.

res.p.000953The corporate cafeteria added bureaucracy to the same physical system. My employer had its management structure. The pharmaceutical company had its own. The union created another set of boundaries around roles, assignments and what could be asked of whom. Corporate policies, contractual responsibilities, client expectations and the practical requirements of getting lunch served did not always line up neatly. Sometimes the person who knew how to solve the immediate problem was not the person authorized to make the decision. Sometimes a procedure that made perfect sense from one layer created extra work in another. You learned quickly that the organization chart and the route through which something actually gets done are related documents, not identical ones.

res.p.000954I did not have language like systems thinking or organizational design for any of this. I knew that kitchens had rhythms, that bottlenecks propagated, that people compensated for weak processes, and that large institutions could create astonishing amounts of work in the spaces between formal responsibilities. I also knew that cooking felt strangely familiar to me. I had already spent years writing songs, and both involved form, sequence, contrast and expectation. Some elements carried structure while others changed the experience without being able to carry the whole thing. More was not automatically better. Timing could change the effect of material that had not changed at all.

res.p.000955Songwriting already felt related to writing. A verse prepared something. A chorus had to justify returning. Information had sequence whether it was sung or read. Writing, in turn, had begun to feel like information architecture: deciding what belonged together, what had to come first, what could remain implicit, what needed hierarchy and how much a person could hold before the structure stopped helping.

res.p.000956None of this felt like professional development while I was doing it. I was working food-service jobs because I needed the money. The useful connections accumulated anyway.

res.p.000957By twenty-three, I knew the arrangement could not continue indefinitely.

res.p.000958I had a family to support, no degree and a set of abilities that did not fit neatly on the résumé I would have produced for a traditional professional job. I could build websites. I understood enough design to know when something looked wrong. I knew how servers worked because I had been making them work since adolescence. The kitchens had taught me more about deadlines, dependencies, unhappy customers and practical coordination than I would have known to call project management at the time. I could write. I could learn quickly. None of these things, individually, solved the credential problem.

res.p.000959The internet had also changed since my Pokémon site.

res.p.000960Businesses needed websites in a way they had not when I was a child, and an enormous freelance economy had grown around that need. I began finding work through oDesk, an online marketplace where companies and individuals posted jobs for freelancers. A client might need a WordPress site fixed, a landing page built, a database problem solved or a larger project completed by someone they would never meet in person. Freelancers created profiles, applied for jobs and accumulated ratings and work history as they completed projects. oDesk later merged with Elance; the combined company became Upwork, but at the time oDesk gave somebody like me an unusually direct route to strangers willing to pay for work.

res.p.000961Some of the jobs came through agencies that had sold more work than their internal teams could complete. They would subcontract part of it: the agency kept the relationship with its client and paid someone outside the agency to perform a portion of the delivery. The end customer might know I was involved or might barely know I existed. I was useful because I could take a defined piece of work and return something the agency could use.

res.p.000962This arrangement was good for me in several ways I could not have designed in advance. The client did not have to decide whether I deserved a career. They had a website problem. If I could solve it, the question of my degree became smaller for the duration of the transaction. I did not have to persuade a hiring committee to take a broad bet on my future potential. I had to persuade someone that I could finish this thing.

res.p.000963Then I had to finish it.

res.p.000964That last part did more for me than the permission to start. A tutorial allows you to stop when you understand the concept. A paying client wants the broken form to work. The site still has to function in the browser they use. The deadline remains Friday even when the plugin behaves differently on the production server. A design that looks elegant on my screen can still fail if the client cannot edit it. Real work kept introducing requirements I had not known enough to include in my own exercises.

res.p.000965I learned faster because somebody was waiting for the result.

res.p.000966The projects also accumulated evidence. After the first job, I no longer had to say only that I thought I could do web work. There was a client who had paid me and received something. Then there were several. Agencies came back with more work. One successful project sometimes led directly to another without forcing the entire question of my employability to reopen. My résumé still lacked the degree. It had acquired another column of information.

res.p.000967Eventually one of those relationships became a full-time job.

res.p.000968From here, the sequence practically volunteers to become a motivational story. Gifted kid discovers computers early. Crisis derails college. Young father refuses to quit. Food-service work secretly teaches him operations. Freelancing turns adversity into opportunity. He enters technology through the side door and builds a career without waiting for permission.

res.p.000969Every sentence contains something true.

res.p.000970Put them together carefully enough and they begin lying.

res.p.000971Being born in 1990 had a lot to do with what became possible for me. Computers entered my childhood at exactly the right stage in their development. I was young enough to absorb them as ordinary objects and old enough to remember when building for the web still exposed much of the machinery directly. If I had been born ten years earlier, the same childhood curiosity would have encountered a much smaller consumer internet. Ten years later, I might have grown up with extraordinary software that hid more of the underlying structure and with a web-development market that was already more mature and crowded.

res.p.000972I did not earn 1990.

res.p.000973I also had access to computers. Somebody could discard a Power Mac G3 into a world where I could scavenge it. Internet access, documentation and communities of other people solving technical problems were available to me. English was my first language in an industry whose technical material overwhelmingly used English. By the time I needed to turn those skills into income, companies were spending rapidly increasing amounts of money on digital work.

res.p.000974The obstacle I usually mention—the missing college degree—was real. So were the conditions that made it possible to route around it.

res.p.000975I had also been unusually strong academically before college. Near the top of my class. Near-perfect SAT. The dropout story can obscure that foundation if I let it. I had already spent years receiving evidence that learning came quickly to me. A crisis interrupted the conventional credentialing process; it did not erase the reading, mathematics, writing and pattern recognition that had made me successful in school before the interruption.

res.p.000976Even the food-service years change depending on how the story is framed. They taught me things. They also paid too little for the life I was trying to support. I learned project management because kitchens and contractor systems contain project management, but I was not enrolled in an innovative experiential leadership program. I was working jobs I eventually needed to escape.

res.p.000977I would never advise someone to engineer my conditions deliberately. Have a mental-health crisis in college. Drop out. Marry at twenty. Have children before you have established a career. Spend several years under financial pressure so necessity sharpens your focus.

res.p.000978That would be insane advice.

res.p.000979Yet once the outcome is known, people are remarkably willing to award obstacles some of the credit for what survivors built around them. The brutal manager taught resilience. Poverty created drive. Scarcity created innovation. An unreliable childhood produced independence. The degree I did not have forced me to become resourceful, so perhaps not having it was secretly an advantage.

res.p.000980I am glad I became resourceful.

res.p.000981I would still have preferred to acquire the resourcefulness without the crisis.

res.p.000982Frederick Douglass gives us the morally enormous version of the same causal mistake. He learned to read while enslaved and used literacy with astonishing force across writing, teaching, publishing, public speaking and political life. Slavery does not get to appear later in the sentence as the forge that produced Frederick Douglass. It restricted education, controlled movement and consumed human possibility. Douglass salvaged capability from conditions designed to deny it.

res.p.000983Even the label “self-taught” can hide part of the field that made an achievement possible. Douglass pieced literacy together from lessons begun by Sophia Auld, white boys in Baltimore, letters marked on timber, discarded copybooks, books and newspapers. The ingenuity was his. So was the risk. The knowledge still arrived through a world containing other people, tools and traces of learning that Hugh Auld could not fully control. Later, Anna Murray helped finance his escape. Reading enlarged what Douglass could understand and navigate; it did not make money, timing, assistance or a workable route unnecessary.

res.p.000984A society does not discover ability simply by waiting to see who manages to display it under hostile conditions. People need time, access, safety, teachers, tools, audiences and enough freedom to use what they know. Douglass’s achievement tells us what he accomplished amid severe constraints. It does not turn those constraints into an examination everyone else fairly failed.

res.p.000985Humans salvage constantly. We make humor out of terrible periods, friendships out of jobs we hated, competence out of emergencies and entire lives out of circumstances we would never select for someone we love. Whatever we carry out belongs to us. The wreck does not become good because we found something worth keeping in it.

res.p.000986Success stories perform another distortion by showing us only the person who reached the visible end.

res.p.000987I can tell you that taking small web projects was a good decision for me. I cannot show you the parallel version of me who took the same projects in a slightly different economy and never found the client who eventually hired him. I cannot show you another twenty-three-year-old in 2013 with equal intelligence and greater web ability whose family experienced an illness that consumed the time mine allowed me to work. I cannot tell you how my life changes if the agency that gave me a second project decided not to. Those people are absent from my evidence.

res.p.000988The surviving version gets the microphone.

res.p.000989Business culture does this constantly. Put the top ten salespeople on a slide and everyone immediately wants to know what they do differently. Sometimes the differences are worth teaching. One follows up with extraordinary discipline. Another asks much better questions. A third understands the industry deeply enough to recognize the customer’s problem before the customer has fully explained it.

res.p.000990Keep looking.

res.p.000991One inherited two major accounts when a colleague left. Another works in a territory where the category is growing much faster. Someone landed a large opportunity after a former colleague became the buyer. One has a manager who keeps internal meetings out of prime selling hours. Below them on the leaderboard may be a representative who spent six months building an excellent deal before the customer was acquired and froze purchasing.

res.p.000992The ranking still tells us who sold the most.

res.p.000993That is exactly what it measured.

res.p.000994It is much less capable of telling us why.

res.p.000995Outliers attract imitation anyway. The best salesperson writes extremely short emails, so short emails enter the playbook. A successful company has very few managers, so another company starts deleting management layers. Toyota develops a production system over decades and another organization installs an andon board. A young founder becomes a billionaire and youth itself begins to look entrepreneurial.

res.p.000996Steve Jobs was twenty-one when Apple was founded. Bill Gates was nineteen when Microsoft began. Mark Zuckerberg was nineteen when Facebook started. Their companies became large enough that a tiny handful of young men supplied a cultural picture of what a founder looked like. Researchers using U.S. Census Bureau administrative data found that among the fastest-growing one in a thousand new businesses, the average founder was forty-five. Relevant industry experience also predicted stronger outcomes.

res.p.000997Jobs did not become less remarkable when the larger sample arrived.

res.p.000998His age became less useful as advice.

res.p.000999The same problem appears whenever visible behavior is easier to copy than the conditions surrounding it. An andon board cannot make an employee safe enough to report bad news. A four-box product grid cannot provide the judgment that determines which four distinctions belong on the axes. Working ninety hours a week cannot make an ordinary product excellent. A sales sequence cannot reproduce the territory, timing, inherited account base or customer relationships surrounding the person whose sequence we copied.

res.p.001000Outcome also changes the names we give behavior. Persistence becomes stubbornness if the company dies and conviction if it survives. Confidence becomes delusion or vision. Concentration becomes tunnel vision or focus. A large bet becomes reckless or courageous.

res.p.001001Sometimes the result teaches us something real. A theory that predicts well deserves more confidence than one that repeatedly fails.

res.p.001002A single outcome contains luck too.

res.p.001003Poker players know this almost physically. Make the statistically correct decision and you can lose the hand. Make an appalling decision and the river card can save you. Enough hands eventually tell us something about how a person plays. The quality of the individual decision cannot be reconstructed from the chips that happened to move after it.

res.p.001004Medicine would become dangerous under a different standard. A physician can choose the treatment best supported by available evidence and lose the patient. Another can take a reckless chance and get lucky. Sorting the two doctors by that single result and teaching everyone to imitate the survivor would train worse doctors.

res.p.001005We do the business version every day.

res.p.001006The company grew, so the strategy worked.

res.p.001007The company failed, so the founders were wrong.

res.p.001008The executive was promoted, so their habits become a leadership lesson.

res.p.001009The author sold millions of books, so someone asks what time they wake up.

res.p.001010My own career makes this problem especially inconvenient for the book I am writing.

res.p.001011I have done well. Better, in some respects, than the circumstances around twenty-three-year-old me would have made obvious. I have spent years noticing that certain ways of thinking reduce repeated work for me. Principles can close decisions I do not need to reconsider every week. Skills learned in one domain can become strangely useful elsewhere. Work can create both an output and the capability required for later work. Saying the true thing early often prevents a remarkable amount of repair later.

res.p.001012It would be very satisfying if my career proved all of that.

res.p.001013It does not.

res.p.001014If these ideas are any good, they should work before anyone gets promoted.

res.p.001015A manager can make capacity visible before accepting another deadline and find out whether the week contains fewer emergency repairs. A couple can identify the disagreement they keep relocating into arguments about smaller things and see whether the smaller arguments stop recurring. A team can change a process and follow the work far enough to discover whether they removed labor or merely moved it to someone with less visibility. Someone can settle a recurring decision once and discover that the fifth version no longer consumes the same amount of thought.

res.p.001016None of those people needs an exceptional outcome.

res.p.001017Someone can become much less divided inside a life that looks less successful from the outside. They may spend ten years pursuing a career and finally admit that they hate what the next promotion requires of them. They step sideways. Income falls. Their title stops advancing. Other people quietly assume something went wrong.

res.p.001018Monday becomes better.

res.p.001019Another person can continue rising while becoming increasingly dependent on everyone around them absorbing the consequences. Their spouse becomes the default parent. Their employees learn to repair contradictions before those contradictions reach them. Money buys enough assistance that inconvenient structures remain tolerable. Authority means more people adapt before asking the leader to.

res.p.001020From the outside, the second life is winning.

res.p.001021Success can afford a lot of friction.

res.p.001022The executive with the immaculate calendar may have three people helping create it. That can be a perfectly sensible allocation of work. Those assistants may be skilled, respected and well compensated. The distortion begins when the calendar appears in a profile as evidence that the executive has achieved extraordinary personal mastery over time.

res.p.001023The same thing happens in stories about balancing work and family. Someone protects family dinner despite an extremely demanding career. Wonderful. Somewhere outside the interview, another person may be handling pickup, appointments, school forms, groceries, birthdays, meals, laundry and all the small acts that make family dinner available as something another person can protect.

res.p.001024The arrangement can be fair.

res.p.001025The mechanism still contains both people.

res.p.001026I can see the same thing in my own story. “College dropout teaches himself web development and breaks into the industry through freelance work” is true and incomplete. I was also the kid building Pokémon pages before adolescence had finished introducing itself. The kid with the PowerPC server in the basement. The kid who tested exceptionally well. The food-service worker who already had songwriting, writing, web development and systems thinking bumping into one another in his head long before anybody paid him to call any of it strategy.

res.p.001027When I finally saw an opening at twenty-three, I was not beginning from zero.

res.p.001028I was carrying more than a decade of accumulated accidents, interests, advantages, failures and skills into a market that happened to have room for them.

res.p.001029I worked very hard after that.

res.p.001030Both sentences stay.

res.p.001031I do not know what my career would look like if we could change one of those conditions while holding every other variable still. Nobody does. I am comfortable claiming narrower things. Paid project work made me better at delivering real projects. It gave future clients evidence they could inspect. Subcontracting let me enter through smaller commitments than a full-time employer would have needed to make. Years of fiddling with computers reduced the cost of learning the next technical thing. Food-service work gave me practical experience coordinating dependent work under time pressure and navigating systems with overlapping lines of authority. Writing and music gave me structures I later recognized in information and systems.

res.p.001032Those mechanisms do not need my title as proof.

res.p.001033Someone else can use them and earn less.

res.p.001034Someone can use them while caring for a disabled parent and merely keep the household functioning.

res.p.001035Someone can make much better decisions during the same year their industry collapses.

res.p.001036Someone can remove years of pointless conflict from a relationship and still lose the relationship.

res.p.001037If the eventual outcome becomes the test of whether the process was good, every unlucky person starts looking defective. Poverty becomes evidence of poor internal alignment. A stalled career reveals a thinking problem. The founder whose company fails must have missed the correct principle. The caregiver who gives up professional momentum apparently failed to build enough leverage. Illness becomes evidence that someone neglected resilience.

res.p.001038That theory would explain everything and therefore nothing.

res.p.001039The failures described throughout this book are smaller and easier to observe. A principle that once saved thought starts producing exceptions people must constantly manage. A feedback system teaches people to game the measure. A manager’s clarity forces employees to translate the plan privately before they can execute it. A household routine remains beautifully efficient as long as one person remembers everything the routine forgot. A clean no becomes reflexive refusal. Financial optionality becomes hoarding. Directness becomes cruelty. Stopping the line becomes an inability to move.

res.p.001040The improvements are just as ordinary. A promise finally matches the capacity behind it. Bad news reaches someone before the expensive version arrives. A workaround disappears. The same apology stops being necessary. A meeting is canceled and nobody has to recreate its function through six private conversations.

res.p.001041I still like telling the story of how I got into this work, but I know how much hindsight has done to it. The Pokémon site, the scavenged Power Mac G3 running in the basement, the food-service years, the freelance projects and the client who eventually hired me now sit close together in memory because I know which parts became useful later. At the time they did not feel like stages of the same process. The server was something I built because I wanted a server. The kitchens were where I worked because I needed to earn money. The freelance jobs were individual chances to make a few hundred dollars doing something I knew how to do.

res.p.001042At twenty-three, I did not know that any of it would become a career. I knew I had two kids to support, a degree I was not going back to finish anytime soon, years of technical curiosity I had never treated as a professional asset and a growing conviction that continuing exactly as I was would eventually become its own decision.

res.p.001043So I started taking work.

res.p.001044Enough of it led somewhere.

Chapter 22 — Find the Rework

res.p.001045Every Tuesday morning, an analyst opens a spreadsheet that is not broken in any ordinary sense. It has been producing the report people expect for years. Three systems feed it, each carrying some history the others do not quite share. One records an event when it is created, another when the work is completed, and a third still uses pieces of a business definition that was replaced eighteen months ago without ever disappearing completely from the older tables. Dates arrive in two formats. A source field occasionally picks up trailing spaces that cause a lookup to fail. A product line was renamed in most places but not all of them. One tab contains a table everyone has been warned not to touch, although the warning becomes meaningful only after somebody explains which formulas depend on it.

res.p.001046The analyst knows all of this so well that most weeks it hardly feels like troubleshooting. They refresh the feeds, compare the total with the previous week, repair the dates, filter the exceptions and check the second source. One familiar discrepancy can be ignored until the next morning because the overnight load will resolve it. Another requires the old product mapping to be corrected manually. By ten o’clock, the numbers are clean enough to circulate. What once took an hour now takes about thirty-five minutes, partly because the analyst has built formulas and shortcuts and partly because years of repetition have taught them which abnormalities deserve attention. If they are out, another person can complete the report from notes, although the notes have the peculiar quality of many documents written by experts: extremely detailed around the parts the expert once had to learn and almost silent around the judgments that have become too obvious to remember explaining.

res.p.001047The people receiving the report encounter none of this. They see a reliable number arriving from a reliable person. That is not a deception. The analyst has done exactly what expertise is supposed to do. Experienced people should know where a process bends, which discrepancy will clear on its own, which one indicates a genuine failure, and how to keep ordinary variation from becoming an emergency. A person who treated every exception as evidence that the whole process needed to be redesigned would be exhausting to work with. Many systems function as well as they do because somebody has learned their texture deeply enough to distinguish the harmless irregularity from the thing that actually deserves interruption.

res.p.001048What can become difficult to see is how much of the apparent stability belongs to the system and how much belongs to the person repairing it. The report arrives every Tuesday because the analyst knows how to reconcile the sources. The customer remains calm because someone knows that the standard email should not be sent in this particular case. A team receives an executable request because a manager has already translated an executive instruction that made sense at one altitude and very little sense where the work actually happens. A meeting begins at nine, but the people who understand its history speak at 8:40 because they know which disagreement needs to be settled before twelve people spend an hour discovering it together. None of these are necessarily failures. They are ways people make organizations habitable.

res.p.001049I have always been drawn to that kind of work. An ugly process bothers me almost physically. If the handoff fails in the same place twice, I want a checklist. If people keep forgetting the same information, I start thinking about how to capture it earlier. If a report takes half a day, I want to know which forty minutes actually require a person and what happened to the rest. There is a satisfying point in this work when a thing that used to require constant attention becomes almost boring. The columns line up. The exceptions stop surprising anyone. The reminder arrives before somebody has to chase it. People who used to ask what happens next simply know.

res.p.001050I have also made bad arrangements easier to tolerate by becoming good at repairing them.

res.p.001051The distinction was not obvious while I was doing it. A reminder system improved, and everyone enjoyed no longer having to remember the reminder manually. Only later did it become apparent that nobody had asked why the same information needed to be chased every week. A report became much faster, which was genuinely useful, while three systems continued producing incompatible numbers because my reconciliation reduced the pressure to settle what each number was supposed to mean. Better documentation allowed a project to carry an ambiguity for months that someone with more authority eventually had to resolve anyway. The downstream work became cleaner, and because it became cleaner, the upstream condition stopped hurting enough to attract attention.

res.p.001052Teams can live inside these accommodations for years without anyone behaving irrationally. A weekly forecast meeting might bring together sales, finance and marketing, each with a number that answers a slightly different question. Early versions of the meeting spend ten minutes explaining why the numbers do not match. Eventually somebody makes a slide showing the definitions side by side. Footnotes appear. A few people reconcile the discrepancies beforehand so senior leaders can spend the meeting discussing the business instead of definitions. This is an improvement. It is also possible, several months later, to have a very efficient meeting in which nobody is especially troubled that an apparently singular commercial process requires three accounts before the people running the company can agree on what they are looking at.

res.p.001053Language acquires the same kind of history. “Pipeline” means something slightly different to one group than another. “Committed” has a formal definition and then the judgment people actually apply when they make decisions. A launch date is the date in the executive material until the conversation becomes private enough for the team to discuss the date they believe they can actually hit. People become fluent in both versions without experiencing themselves as dishonest. They are doing what humans do in complicated institutions: learning the practical meanings that accumulate around formal ones.

res.p.001054New employees have a talent for making these arrangements visible because they have not yet learned which instructions are literal. They follow the process as written and discover that an experienced colleague has to stop them. Finance should be consulted before the form that automatically routes to finance is submitted. Product needs to see the draft before legal even though the workflow says legal comes first. The system says “thirty days,” but everyone plans around ten. The customer technically has permission to change a configuration that experienced employees know they should never be allowed to touch without supervision. The newcomer writes these things down. Six months later the notes become unnecessary. A year later, they are explaining the exceptions to somebody else.

res.p.001055That transfer of knowledge is not evidence of some grand organizational pathology. Institutions have sediment. A process built for one software system survives the replacement of that system because part of it still serves a purpose. A form retains a field nobody wants because removing it safely requires discovering all the places downstream that still depend on it. Two departments develop different definitions because each definition was built around a different decision. A manager becomes the interpreter between functions because the organization does not have enough work of that kind to justify building a formal role around the translation. Excavating all of this can be expensive, and the person who knows how to navigate it may be the most economical solution available.

res.p.001056Households work the same way. During a difficult month, with children, jobs and perhaps a sick parent in the mix, nobody needs to stop family life and redesign the entire distribution of cognitive labor. One person may keep handling the school forms because they already know the teachers, the portals, the passwords and which emails require action. Another person may take care of medical appointments because the history lives mostly in their head. The arrangement can be uneven without being currently foolish. The family may have more urgent things to survive than the inefficiency of its information architecture.

res.p.001057Thirty-five minutes of spreadsheet repair can be a perfectly reasonable price.

res.p.001058The recurrence is still worth knowing about. Preparing a weekly report is recurring work because another week produced new information. Cleaning the same malformed source data before preparing it is different. Explaining again why the number generated by the system cannot be used as reported has accumulated more history. Quietly making the number usable before an executive sees it because nobody wants to reopen the source problem contains more than reporting. Something was already defined, decided, entered, promised or built, and a person has to restore the connection between that earlier act and what the situation now requires.

res.p.001059The difference can be easier to recognize emotionally than analytically. Someone accepts another request, completes it well and spends the evening angrier than the request seems to justify. A person types the same explanation for the fourth time and notices irritation beginning to attach itself to the recipient. One partner reminds the other about another appointment and dislikes the tone they hear coming out of their own mouth. A team responds to an ordinary request for project status with wildly disproportionate hostility. Resentment is unreliable as diagnosis; people resent duties they freely chose, consequences they helped create, obligations no reasonable arrangement could eliminate, and sometimes whatever happens to arrive after a bad night’s sleep. But repeated resentment tends to have a geography. It gathers around particular exchanges, particular kinds of work, particular relationships between responsibility and authority.

res.p.001060Getting close to the arrangement usually produces a less flattering and more useful account than deciding that somebody simply needs better boundaries. An employee may resent repeated requests because a no has never been explicitly forbidden but has acquired a predictable career cost. They may be angry at the requester while continuing to say yes and secretly wishing the requester would notice what the yes requires. Someone may enjoy being indispensable and resent the life built around everybody depending on them. A team may be understaffed badly enough that coworkers have begun treating one another as the cause of a problem produced mostly by arithmetic. A spouse may be carrying work the other person cannot value because the work is precisely what prevents the other person from having to notice it.

res.p.001061Tools often arrive too early. The couple fighting about money gets a budgeting app. The team exhausted by requests for updates gets a dashboard. An organization with too many meetings adopts a meeting policy. Someone overwhelmed by commitments downloads another task manager. Any one of these can be useful; I use tools constantly and have little nostalgia for work a computer can perform more reliably. The danger appears when the new tool becomes a better interface for the same contradiction. The budgeting app categorizes with exquisite precision the purchases through which two people continue disagreeing about what money is for. The dashboard gives leadership a cleaner view of projects while leaving them insulated from the conversations in which uncertainty is expressed honestly. Meetings disappear from the calendar while their coordination work fragments into messages, small calls and private follow-ups. The task manager becomes another surface on which a person reorganizes commitments that cannot all fit inside the week.

res.p.001062My own instinct makes this especially easy to miss. I enjoy the point when a messy thing becomes orderly, and order is persuasive. Once a process runs smoothly, asking whether it should exist in its present form can feel almost perverse. Somebody just solved the problem. Why are we reopening it? The discomfort becomes stronger when a person’s value has grown around the repair itself. The analyst is valuable partly because they know how to reconcile the systems. The manager who can translate a difficult executive has developed a real and useful skill. The employee who rescues deadlines becomes the person senior leaders trust with the most consequential work. The family member who remembers everything becomes, in a very literal sense, somebody everyone relies on.

res.p.001063Removing the underlying problem can alter the role through which a person has become competent and needed. That does not imply that people secretly want dysfunction, although sometimes we do become attached to the forms of importance it gives us. More often, the immediate consequences of changing the arrangement simply look worse than continuing it. Transfer school communication from the parent who has managed it for five years to the parent who has not, and the first few weeks may be objectively less efficient. Passwords are missing. Context has to be explained. An email that the experienced parent would have classified correctly in fifteen seconds requires a discussion. Field day turns out to require a particular shirt, because apparently field day has acquired a shirt. The person who already knows all this can complete the work faster and may reasonably wonder what has been gained by watching someone else stumble through it.

res.p.001064If they reclaim the task every time the handoff becomes clumsy, the family will continue producing excellent evidence that they are the only person capable of doing it.

res.p.001065Organizations recreate dependency the same way. Ownership is transferred. The new person moves more slowly, misses a nuance or asks questions that the previous expert stopped having years ago. The expert steps back in because the deadline is real and there is no virtue in allowing avoidable failure. Everyone feels immediate relief. After several rescues, the organization concludes that the work simply requires the expert. Sometimes it does. Expertise is real, and distributing every capability evenly would be a bizarre objective. The harder judgment is whether the temporary inefficiency belongs to learning or demonstrates that the transfer itself was poorly conceived. The answer does not appear in the first bad week.

res.p.001066Following rework upstream can produce an equally inconvenient result. Imagine the analyst finally stops correcting one recurring discrepancy silently and brings it to the people responsible for the two source systems. One group explains that its definition is necessary for operational reporting. Finance explains that its definition has been used for years and preserves historical comparability. After enough discussion, everyone realizes that neither group made a mistake. The systems disagree because they answer different questions. The analyst has been reconciling two legitimate measures into a single number because the executive report expects one.

res.p.001067The thirty-five minutes may not disappear. Perhaps the report should show both measures. Perhaps a third definition should be created for the specific decision the report supports. Perhaps someone really does have to reconcile the two every Tuesday. What looked like a poorly designed spreadsheet turns out to contain a question about what the organization wants to know, and there may be no answer that reduces the question to one universal number without throwing away information.

res.p.001068Other investigations are less dignified. A forecast keeps being rebuilt because the assumptions changed months ago and leadership will not change the target. Customer support maintains an elaborate exception guide because sales repeatedly promises configurations the product does not handle well. A project manager spends Friday evening rewriting status reports because uncertainty receives such a hostile response that everybody learns to soften it before information travels upward. One spouse continues carrying every household reminder because both people believe, with some justification, that they already do more than the other understands. In each case the downstream repair is easier to modify than the relationship, incentive or decision producing it.

res.p.001069People doing the repair frequently know this. They are not too unsophisticated to find the root cause. The analyst may have raised the source discrepancy six times. The employee working late can understand the staffing problem perfectly and still have a customer who needs an answer tomorrow morning. A spouse may have had the larger argument several times and decided that keeping the household operating through an unusually bad month matters more than forcing the issue again tonight. A manager may be translating an executive request because the team still needs something executable regardless of whether the executive ever learns to make the request better.

res.p.001070Downstream work can be necessary while its source remains absurd.

res.p.001071The person repairing it may be the reason everyone else can afford to postpone the source.

res.p.001072Authority changes what can happen after the pattern becomes visible. A manager receives the finished report and can become interested in what occurred before it looked finished. “I fixed it again” sounds different if again is allowed to remain part of the sentence. One employee has saved three deadlines in eight weeks. Every important request is being informally pre-cleared before entering the formal approval process. A customer exception has become common enough to have a nickname. Those details are often more revealing than the polished output because they show where people have already begun building unofficial infrastructure.

res.p.001073The first question does not have to be how to remove it. If nobody performs the workaround, what breaks? The answer may be payroll, a customer commitment, the accuracy of an executive report or a child getting where they need to be. The workaround survived because it is protecting something. Any attempt to change the surrounding process has to preserve whatever protection the repair currently supplies. Otherwise the organization has confused the person compensating for a defect with the defect itself.

res.p.001074I have watched process improvement become destructive when people get excited about compensatory work and start treating recurrence as proof of waste. Meetings are suddenly presumed guilty. Informal conversations look inefficient. Redundancy appears to be an obvious target. Anything undocumented becomes a failure of process maturity. Real human systems need slack, memory, judgment and overlapping forms of care. A nurse who checks something the software already checked may have years of experience with the failure the software does not make visible. Two colleagues talking for five minutes before a formal review may be using trust to surface an uncertainty that would disappear in the larger room. Family members remind one another of things. Experienced employees carry distinctions no procedure can anticipate completely.

res.p.001075The repeated work I distrust is narrower than repetition itself. I notice it when people keep repairing a known mismatch while the rest of the arrangement continues to behave as though the mismatch does not exist. Someone apologizes for being late every week while continuing to schedule travel according to a drive time that has never once described rush hour. A person reorganizes the same impossible task list every Sunday while protecting every commitment that makes it impossible. Someone rewrites feedback five times because the sentence is not what makes the conversation frightening. A company creates another status artifact because leaders say they need visibility even though project owners already understand the risks and have learned that making those risks fully visible creates trouble.

res.p.001076Sometimes the correction is wonderfully ordinary. Change the travel assumption. Fix the source field. Move the meeting. Make the phone call you have avoided for a year. Give the person the authority they have been exercising informally. Admit that the “temporary” condition has lasted long enough to become the planning baseline.

res.p.001077Sometimes the investigation ends with the repeated work intact. The weekly coordination meeting exists because several people genuinely need one another’s current information. Caregiving contains tasks that return because human bodies continue having needs. Two numbers remain different because they represent two distinctions the business needs to preserve. There is no merit in pretending all recurrence is evidence of bad design. Some work remains expensive after close examination because the thing being accomplished is genuinely expensive.

res.p.001078The analyst eventually opens the documentation for the two source systems on an afternoon when there are plenty of more urgent things to do. Thirty-five minutes every Tuesday is not a crisis. They could leave the process alone for another year and nobody would accuse them of negligence. They put the two definitions beside each other anyway. One counts the event when the customer commits. The other counts it when the work is completed. For years, the spreadsheet has been translating between them so that an executive report can present a single number.

res.p.001079The analyst sends the definitions to the people who own the report and asks which event they are actually trying to see.

res.p.001080The answer may leave the spreadsheet exactly as complicated as it was.

res.p.001081At least the complication will no longer be mistaken for something the analyst simply has not optimized well enough.

Chapter 23 — Decide Once, Stay Correctable

res.p.001082At work, one question has saved me from a remarkable amount of bad arithmetic: if we add this, what should move?

res.p.001083The question closes a recurring argument before urgency gets to reinvent arithmetic. Its usefulness depends on something less comfortable: I also have to notice when the facts underneath the settled answer have changed. A principle that cannot survive pressure is not much of a principle. A principle that cannot be revised by new reality has become something else.

res.p.001084Many principles quietly become preferences. Early warning is excellent unless the customer is especially important. People should speak candidly unless the risk is still uncertain. Capacity should be visible unless this quarter is unusually consequential. Employees should challenge assumptions unless the assumption belongs to the person evaluating them. The abstract rule survives while its application keeps exempting whatever has enough emotional or political force to make the rule inconvenient.

res.p.001085The opposite danger arrives after a principle survives long enough to become part of somebody’s identity. A person learns early in a career that being unusually available creates opportunities. Difficult work appears, and they are nearby. They learn faster because they see more problems. People remember that they answered, stayed late, figured it out and could be trusted when something broke. The behavior produces real returns. Promotions reinforce it. Praise reinforces it. Eventually “I make myself available when useful” becomes “I am the person who can always be counted on.”

res.p.001086Twenty years later, the phone is still at dinner.

res.p.001087The environment may have changed completely. The person has a family. Their work now requires sustained thought more than instant responsiveness. The people reporting to them are experienced enough to make decisions locally, except the leader keeps answering quickly enough that local judgment rarely has to develop. They complain that the team needs more ownership while providing such efficient access to their own judgment that ownership remains a slower option for everyone involved. Each response can still look responsible in isolation. The accumulated pattern is no longer producing what the original behavior produced.

res.p.001088The intelligence inside an old adaptation is what makes changing it difficult. Someone who learned early that depending on other people was dangerous may have learned something true about the environment they were in. They became resourceful, difficult to trap and capable of leaving situations in which dependence would have been used against them. Those capacities can remain valuable for a lifetime. Later they may build relationships with people who are not trying to control them and discover that the old response still arrives first. Asking for help feels humiliating. Accepting care feels like surrendering leverage. Self-sufficiency continues protecting against an environment that is no longer present, and because the protection was once intelligent, calling it irrational does not help much.

res.p.001089I would not want to solve this by treating every stable pattern as suspicious. A person who reopens every principle whenever it causes discomfort has not become flexible; they have created an internal legislature that never adjourns. There are things I want to remain settled for a long time. I do not need to conduct a new investigation into whether lying tends to create downstream work whenever telling the truth becomes awkward. I do not want to rank my family against my job every time work presents an impressive emergency. A decision that cannot exert any inertia over future circumstances was barely a decision.

res.p.001090The useful pressure for revision usually comes from recurrence.

res.p.001091An organization says “weekend work is unusual.” A customer changes scope and people work Saturday. A dependency slips the next month and produces another Saturday. Quarter-end creates a third. Someone leaves, and the team covers. A system breaks. Each event has a real explanation, and each explanation may justify the response to that particular event. Six Saturdays later, the question has changed. Management can continue explaining why each weekend was exceptional while the people working them experience something that has acquired a cadence.

res.p.001092I have become suspicious of “temporary” conditions that survive long enough to develop seasons.

res.p.001093Perhaps staffing is now too thin for the promises the organization makes. Perhaps deadlines are being committed before the work is understood. Perhaps the business has genuinely entered a difficult period in which extraordinary effort is necessary and worth asking for. If that is the case, people deserve an account of the period they are actually living through rather than a permanent sequence of exceptions to a normality that no longer exists. Repeated exceptions preserve information that the explanation for any one incident can hide. One Saturday contains reasons. Six months contains frequency.

res.p.001094A principle itself can create the recurrence that eventually forces revision. Imagine a team that has been taught to escalate risks early. People respond enthusiastically, and soon ordinary fluctuations are repeatedly entering senior conversations as emerging threats. Leaders investigate issues that rarely alter a decision. After enough false alarms, they begin discounting warnings, which makes the whole organization less able to recognize the warning that finally deserves attention. Early escalation was not a foolish principle. The way people learned to apply it degraded the information it was supposed to improve. Perhaps the threshold needs refinement. Perhaps employees need better models of normal variance. Perhaps the incentive rewards announcing risks but imposes no consequence for continually passing undigested uncertainty upward. The principle should be capable of learning from the behavior it helped create.

res.p.001095Consistency becomes especially dangerous when other people can use a change of mind as evidence of weakness. Public figures have two statements placed beside one another as proof that they cannot be trusted. Executives reverse a strategy and are asked how they could have spoken with such certainty six months earlier. Children discover precedent almost immediately after a parent changes a rule. Organizations respond with a predictable vocabulary. Strategies “evolve.” Priorities are “refined.” The new position is described as completely continuous with the old one even when everybody in the room knows that something substantial changed.

res.p.001096Sometimes continuity is real. A company can change a tactic while preserving the same purpose. New evidence can change which action best serves a stable principle. The more expensive maneuver is keeping a bad decision alive because too much reputation has been invested in the person who made it.

res.p.001097I have seen reports survive this way. Somebody important asks for one, and producing it is sensible at the time. Months pass. The leader stops using it or leaves. Another leader inherits the meeting where the report appears. The analyst assumes somebody must still need the information because surely nobody would knowingly request the same monthly report without a reason. Eventually a newcomer asks which decision it supports and discovers that the question itself has become difficult to answer.

res.p.001098The first report had an owner and a purpose.

res.p.001099The hundredth had tradition.

res.p.001100Personal rules lose their expiration dates in much the same way. A financial practice created under real scarcity can become hoarding after the person has enough security to behave differently. A boundary created during a harmful relationship can survive years into a relationship whose behavior no longer resembles the one that formed it. A career strategy organized around proving competence can keep driving the person after their standing is well established, leaving them to win arguments and accumulate responsibilities that no longer serve anything they actually want.

res.p.001101I often notice the decay first in the amount of explanation required to preserve the old answer. A principle that once fit into one sentence accumulates clauses. Someone raises an exception, and the explanation becomes longer. Another person notices the same problem, and more context is required. Soon a rule that was supposed to simplify judgment needs a small legal apparatus around it. The growing defense does not prove the rule is wrong. People routinely resist good constraints. Customers want exceptions. Teams sometimes dislike processes precisely because the process makes necessary work visible. But increasing amounts of compensatory reasoning can indicate that the principle is being protected from the evidence it was supposed to organize.

res.p.001102Correctability, at least as I experience it, does not feel like walking around eager to change my mind. I have beliefs I expect to survive scrutiny and decisions I do not want to reconsider every week. I am interested in preserving routes through which certain kinds of evidence can still reach them. The route differs with the decision. At work, a strategy can keep missing while the explanations for why each miss should not count become increasingly polished. In a relationship, another person can describe the same experience repeatedly while I continue answering with an explanation of what I intended. In a household, one person can keep saying they are overwhelmed while an apparently equitable division of labor survives on paper. In my own systems, compensatory work can reappear after a principle was supposed to remove it.

res.p.001103If I keep saying weekend work is no longer part of the normal plan and weekends keep disappearing, something has become more authoritative than the sentence.

res.p.001104If a budget was supposed to reduce anxiety and now every purchase requires a negotiation, the system is producing evidence about itself.

res.p.001105If a boundary was supposed to protect a relationship from resentment and the boundary has become the relationship’s main source of resentment, the original reason for having it deserves another look.

res.p.001106None of these consequences supplies the answer automatically. A sound strategy can have a bad quarter. A good product can have an angry customer. A worthwhile marriage can contain a miserable month. A career can become unusually difficult without becoming the wrong career. Decisions would be impossible if every negative outcome forced immediate reconstruction.

res.p.001107Twelve bad quarters can also be described as patience.

res.p.001108Years of complaints can become edge cases.

res.p.001109A life spent waiting for the rough patch to end can become the rough patch.

res.p.001110Experience gives us no universal number at which variance becomes pattern. It does give us comparisons across time, and other people are often useful precisely because they have not been living inside the explanations we developed for each event. A colleague can say that the supposedly unusual deadline has happened every month. A spouse can remember that work has been “temporarily crazy” every winter for five years. A friend can notice that three different relationships have eventually produced the same story. They can be wrong. Outsourcing judgment to whoever has enough distance would simply replace one bad model with another. But correction does not have to originate inside the same mind that constructed the thing being corrected.

res.p.001111Music gives me a physical version of that experience. Practice something incorrectly for long enough and the error stops feeling incorrect. The hand expects it. The ear predicts it. A rushed phrase begins to sound natural because familiarity and correctness have become entangled. Someone else listens and says I am rushing. I play it again and cannot hear the problem. They clap the pulse while I play.

res.p.001112Then I hear it.

res.p.001113For a while the corrected version feels worse. The hands resist. A passage that had been fluent now requires attention. The old version remains smoother because it has had more practice. Smoothness tells me how deeply the pattern was learned, not whether it deserved to be learned.

res.p.001114Organizations often evaluate change at exactly this unfair moment. The old process has years of local intelligence wrapped around it. People know whom to ask, what not to trust, which field can be ignored and how to recover from familiar failures. The new process is awkward because nobody has built those accommodations yet. People cannot find things. Questions increase. Work slows. A comparison made during this period pits a mature set of adaptations against a system that has barely begun acquiring them.

res.p.001115The old system may genuinely have been better. The new design may be bad. It may also need enough time to become competent before anyone can tell what portion of the new work belongs to learning and what portion is structural. That period is politically difficult because everyone involved can convert the inconvenience into evidence about themselves. Leaders who introduced the change become defensive. Employees who opposed it collect every failure. Soon the system is serving as a referendum on whether the people who chose it were right.

res.p.001116Evaluation becomes more useful when nobody has to win. I want to know what the change was supposed to improve, whether that improved, what new work appeared, what seems temporary, and what remains after people have learned the system. Implementation also reveals things nobody could have known when the decision was made. Most correction happens at this scale: a threshold changes, a rule narrows, a recurring meeting disappears, a manager asks a slightly different question, or a confident sentence acquires a qualifier because the qualifier describes reality more accurately.

res.p.001117The wording of a rule is less durable than the thing it was protecting. A capacity rule protects the fact that added work displaces something. An escalation rule protects the ability to respond while choices remain cheap. A boundary may protect a relationship from a pattern that repeatedly converts care into resentment. When circumstances genuinely change, preserving the reason can require changing the sentence. The relevant question is whether the new case differs along the dimension that produced the principle or merely carries more emotional weight than the ordinary case.

res.p.001118I do not want to ask that question every time. Doing so would recreate the deliberation I was trying to preserve. I tend to ask when the pattern begins pushing back: exceptions become regular, resentment keeps attaching to the same arrangement, the defense grows much more elaborate than the principle, outcomes drift away from what the rule was meant to produce, several people independently notice the same problem, or the rework that supposedly disappeared begins appearing somewhere else.

res.p.001119The return of rework is especially hard to ignore after spending this much time looking for it. Perhaps I aimed the principle at the wrong mismatch. Perhaps the environment changed. Perhaps I obeyed the wording so faithfully that I defeated the reason. Perhaps the work I was trying to eliminate was never compensatory at all and belonged to the irreducible difficulty of maintaining a relationship, caring for someone, exercising judgment or doing work that is genuinely complex.

res.p.001120Looking is more useful than defending the elegance of the framework.

res.p.001121There is a version of self-improvement in which all of this eventually disappears. Food is settled. Exercise is settled. Money is settled. Boundaries are settled. The calendar perfectly reflects priorities. A disciplined person has eliminated so much internal negotiation that life begins to resemble an airport operating according to an unusually competent schedule.

res.p.001122I can understand the fantasy. Repeated self-negotiation is exhausting.

res.p.001123I do not want the airport.

res.p.001124A living person encounters information that was not available when the principle was formed. Bodies change. Children change. Work changes. Relationships become safer or worse. People act in ways we did not predict. We discover that something we thought we valued no longer has much life in it, or that something we regarded as peripheral has become central. Obligations arrive that deserve consideration despite having failed to secure calendar space in advance.

res.p.001125I want to stop spending attention on questions whose relevant facts have not changed so that I have attention when they do.

res.p.001126A late feature request that brings no new information does not require me to rediscover finite capacity; a regulatory problem that makes the launch dangerous does. A pleasant conversation does not erase repeated lies; a year of different behavior should eventually be allowed to become evidence that the person in front of me has changed. One bad week does not decide a career; a career whose structure no longer contains the reasons I chose it deserves more than another attempt to optimize the week.

res.p.001127No checklist can tell me exactly when one situation has become the other. I am increasingly glad of that. If correctability could be reduced to a procedure, I would eventually learn how to complete the procedure while preserving whatever conclusion I preferred.

res.p.001128Mostly I watch how much work the old answer now requires. An explanation that was once clear begins needing paragraphs. Exceptions begin forming a schedule. Several people encounter the same cost while I continue calling each instance unusual. Something I describe as temporary survives long enough to acquire anniversaries.

res.p.001129Sometimes I look again and the old answer still holds. The evidence does not require a change, and the principle leaves with more support than it had before.

res.p.001130Sometimes the answer changes.

res.p.001131Both outcomes have to remain possible.

res.p.001132Then another good request arrives. It contains no information that changes what has already been promised. I ask what should move, and there is nothing worth moving, so the work waits. Later, something genuinely changes the plan, and we change the plan.

res.p.001133There is no contradiction between those decisions.

res.p.001134The consistency is in what was allowed to change them.

Conclusion — Less of You in the Way

res.p.001135A surprising amount of adult competence consists of becoming extremely good at making imperfect arrangements work.

res.p.001136Somebody fixes the number before the report goes out. Somebody knows the official deadline is implausible and quietly builds the schedule that can actually be delivered. A manager understands that an executive request will confuse the team if it arrives intact, so they translate it first. A salesperson knows which promise needs a private warning to implementation. One person in a household has learned the school calendar well enough that nobody else needs to think about it very often. Someone accepts work they do not really have room for and solves the capacity problem later, privately, with an evening or a weekend.

res.p.001137These acts are often generous. They can be intelligent, skilled and necessary. Most of the people doing them are trying to reduce trouble for everyone around them. They are the reason the report arrives, the customer remains calm, the children get where they need to go and the project stays green.

res.p.001138Competent people can keep a bad description functioning for a very long time.

res.p.001139Once they do, the repair starts looking like the work. We optimize the spreadsheet. We make the translation faster. We create the pre-meeting where the disagreement can be settled before the meeting in which everyone is supposed to agree. We document the workaround carefully enough that the next person can inherit it. We become proud of being the one who can handle the impossible request, remember the invisible dependency or rescue the plan.

res.p.001140Sometimes the repair deserves to survive. A complicated world produces genuinely complicated work. Children need to be fed again tomorrow. Bodies get sick. Relationships require repeated attention. A musical passage that was learned last year still lives in a body that has to play it tonight. Several people working together will sometimes need a meeting simply because the information required for a decision resides in several people. Love, justice and responsibility can all ask us to accept costs that would disappear if we cared less.

res.p.001141I do not want a life engineered until none of those things can inconvenience me.

res.p.001142The work I have become suspicious of is the additional work required to keep a mismatch from becoming authoritative. The promise exceeds capacity, so people repair the schedule. The public account differs from private knowledge, so people maintain both. Authority sits somewhere other than responsibility, so somebody negotiates informally before using the formal process. A metric drifts away from the thing it was created to represent, and people learn how to succeed according to the metric while privately compensating for what it misses. A principle that once reflected experience stops fitting, and increasingly elaborate explanations keep it from having to change.

res.p.001143Eventually the repair itself begins carrying information.

res.p.001144The same apology returns. The same weekend disappears. The same argument finds another subject. The same number requires correction. The same person has to remember. A process has an official version and an unofficial version known to everyone competent enough to use it. A condition described as temporary survives long enough that people can predict when it will return.

res.p.001145None of these observations tells us automatically what to do. The recurring meeting may be necessary. The difficult person may be right. The rule may still be sound. The employee who keeps volunteering may genuinely want the work. A family arrangement can be unequal and still be the arrangement everyone involved currently prefers. Reality rarely supplies moral conclusions in the convenient form of a dashboard.

res.p.001146It does keep supplying particulars.

res.p.001147Signal began there: with the difficulty of receiving those particulars without immediately forcing them through what we already believe, want, fear or need other people to think about us. Other people are part of that signal. So are consequences. So is the irritation we would rather explain away, the observation that does not fit the model, the person whose experience changes the apparent success of the system once we allow them inside its boundary. We cannot become perfectly transparent to reality. We can become easier to reach.

res.p.001148That sounds modest until the thing reaching us threatens something we have built.

res.p.001149A leader can say they want candor and become fascinated with evidentiary standards when the bad news concerns their plan. A person can value correction until the correction touches an identity they are proud of. An organization can celebrate customer obsession while quietly teaching employees which customer information cannot survive contact with the quarterly target. We can spend years becoming more perceptive and use the resulting perception to defend ourselves more skillfully.

res.p.001150Receiving the signal does not complete the work.

res.p.001151The BDR who interrupted the campaign near the beginning of Response had heard something the people farther from the customer had missed. She could have been thanked warmly while the campaign proceeded unchanged. Her observation could have entered the organization as information and left no trace on its behavior. Instead the material changed before it reached the market. The signal altered the response.

res.p.001152That movement can be almost embarrassingly small. A deadline changes. A promise becomes narrower. A manager asks what should move before accepting more work. Someone admits that the number everyone has been using measures the wrong event. A standing meeting disappears and, for once, its work does not quietly reappear elsewhere. A person who has spent years defending a decision says, eventually, that the facts now support a different one.

res.p.001153Nothing about these acts requires a reinvention of the self. They often require less performance from it.

res.p.001154I know what happens in me when correction reaches a subject I already regard as revisable. A weak paragraph can be a weak paragraph. An argument can contain a break. Once I see it, I can work on the break. The machinery becomes much more impressive when the threatened object is attached to who I think I am. Context rushes forward. The critic’s flaws become newly vivid. Every alternative interpretation presents itself for consideration. The defense may contain true things. Its extraordinary speed tells me that something besides the merits of the argument has entered the room.

res.p.001155The same machinery scales.

res.p.001156A company can protect a strategy long after the evidence has weakened because reversal would implicate the people who chose it. A team can preserve an impossible target and rebuild the forecast every month rather than admit that the assumptions underneath the target have changed. A household can keep calling one person “better at this stuff” while their competence quietly becomes the infrastructure that lets everyone else remain worse at it.

res.p.001157Once identity, authority and habit gather around an arrangement, truth has to do more than be true. It has to push through everything built around the old account.

res.p.001158Much of the work in this book has been an attempt to reduce that distance before the pushing becomes expensive. A clean no keeps a false capacity promise from entering the plan. An early warning reaches people while there are still choices. A principle preserves a decision that does not need to be rebuilt under pressure. A feedback loop lets an outcome revise the assumptions that produced it. Direct conflict keeps disagreement from migrating into safer but less useful arguments. Money can preserve enough room to refuse a choice that would otherwise become compulsory. Work can build another person’s capability instead of ensuring the expert remains permanently necessary. A system boundary can include enough of the people carrying its costs that transferred burden does not masquerade as efficiency.

res.p.001159None of these techniques knows what a good life is.

res.p.001160A predatory company can clarify its strategy. A manipulative person can become very good at reading people. A political movement organized around cruelty can reduce internal friction and coordinate efficiently. Someone can make excellent decisions in service of an ugly purpose. The chapter on dangerous alignment had to leave that problem open because mechanics cannot answer it for us.

res.p.001161The book cannot give the word “alignment” moral authority simply because alignment feels clean.

res.p.001162The world contains people who are coherent and wrong.

res.p.001163It contains people who are conflicted and right.

res.p.001164A system may become slower when people who had been excluded acquire enough power to be heard. A decision may become more expensive when its environmental cost enters the calculation. A manager may lose some speed when an employee’s life stops being treated as spare capacity. Justice can add friction to the experience of the person who previously enjoyed the benefits of an unjust arrangement.

res.p.001165Smoothness is not the test.

res.p.001166Neither is difficulty.

res.p.001167The test cannot be success either. My own career would make a particularly seductive proof if I allowed it. I was born into an unusually favorable moment for the skills that happened to fascinate me. I worked hard, learned quickly, made useful choices and found ways around constraints that could have ended the story differently. I also encountered a market with room for what I knew, people willing to give me work and circumstances that allowed one project to survive long enough to produce the next one. Another person can make excellent decisions and arrive somewhere less impressive. Someone can make terrible ones and get rich.

res.p.001168Outcomes still teach. They simply refuse to tell us everything we would like to know about merit, wisdom or causation.

res.p.001169The quieter evidence has become more persuasive to me. Does the work keep returning after we supposedly solved it? Does the same person keep absorbing the difference between the plan and reality? Can bad news travel while it is still ambiguous enough to be useful? When another person describes the cost of my decision, does their experience enter the account even if I ultimately make the same decision? Can a principle be strong enough to settle a question and permeable enough to new information that genuinely changes it?

res.p.001170These are things we can observe before the biography has an ending.

res.p.001171They also leave room for failure.

res.p.001172The truthful conversation can go badly. The company can make the better-supported decision and lose. The boundary can cost a relationship. A person can decline the wrong opportunity. The team can surface the risk early and still miss the deadline. Correctability would be meaningless if it were merely another name for always producing the outcome we wanted.

res.p.001173Reality does not promise to reward accurate contact with it on our preferred schedule.

res.p.001174Sometimes better contact makes the next hour worse. The conflict happens now instead of spreading itself across six months of proxy arguments. The customer hears the limitation before signing. A project turns yellow while there is still time to act rather than remaining comfortably green until nobody can. Someone says no and another person is disappointed. A rule changes and the new process becomes awkward before anyone learns how to use it.

res.p.001175The savings are downstream, and some of them are not savings at all. They are simply costs finally being borne where we can see them.

res.p.001176Music remains the closest thing I have to an image for what I want instead.

res.p.001177When I have learned a passage badly, an absurd amount of attention goes into surviving what should already be settled. The difficult measure approaches and my body prepares too early. The hand tightens. Part of my attention leaves the phrase and begins supervising mechanics. A mistake makes the field narrower still. I can continue producing the notes while becoming progressively less available to the music.

res.p.001178Practice does not eliminate the work. It changes where the work lives.

res.p.001179The fingering becomes ordinary. The rhythm no longer has to be rebuilt consciously. The hand has enough memory that attention can move outward again. I can hear the person playing beside me. The room can affect how long a phrase wants to be. Somebody breathes differently and I can answer instead of merely completing my part.

res.p.001180Then I make another mistake.

res.p.001181Perhaps someone else hears it first.

res.p.001182The old version may feel more natural because I have practiced it longer. I can slow down, listen, change the motion and make the corrected version awkward for a while.

res.p.001183Eventually attention becomes available again.

res.p.001184I do not know a final state beyond that. I do not expect to settle enough principles, repair enough systems or become perceptive enough that life stops producing information capable of disturbing me. Children grow into people no model of them can hold still. Work changes. Bodies change. Relationships reveal things. Institutions drift. We discover that an adaptation which once protected us has begun restricting us. We find out that something we considered an efficiency depended on another person doing invisible work.

res.p.001185A living system keeps generating signal.

res.p.001186The response cannot be permanently finished.

res.p.001187What can become smaller is the machinery required to keep us from hearing it, and the machinery required to preserve an answer after the signal has changed.

res.p.001188Less distance between what I know and what I say. Less distance between what I promise and what can be delivered. Less distance between authority and responsibility, between a metric and the thing it claims to describe, between the success I experience and the person carrying part of its cost. Less distance between the identity I prefer and the evidence that would require me to become slightly different.

res.p.001189Not zero distance. Human beings do not get zero distance. Language loses information. Memory edits. Incentives work on us before we notice them. We misunderstand ourselves and one another. We make decisions with facts that will look incomplete from the future.

res.p.001190Perfection would leave no room for creatures like us.

res.p.001191A smaller distance leaves more attention.

res.p.001192Not attention waiting to be filled with another productivity system. Attention capable of noticing the person beside us, the disagreement inside the sentence, the sound of the room, the information that arrived without permission. Attention for the difficulty that remains after the compensatory work has been removed because the difficulty belongs to the thing itself.

res.p.001193Less of me in the way does not mean less personality, ambition, attachment, judgment or desire. I am not trying to become a transparent pipe through which reality flows uncontested. Judgment is part of the response. Desire is part of the response. Love is part of the response.

res.p.001194I want fewer parts of myself employed full time preventing experience from revising the account.

res.p.001195The rest can stay.

res.p.001196Reality will keep talking.

res.p.001197I would like to remain reachable.

res.p.001198And when what reaches me deserves an answer, I would like my life to be capable of giving one.