Chapter 3. The legacy we're living

A working draft, November 2026. This book is being written while The Grove is being built; expect it to change. Checked against the build as it stood on 1 November 2026.

In this chapter

• How the org chart began on a railway in 1855, and the problem it solved
• What the tree of command was really doing: moving information up and decisions down
• Why that job has changed, and what matrix, flat, agile, sociocracy, Holacracy and Teal each learned
• How the habits of hierarchy live on in us after the boxes change
• What we kept from hierarchy, what we dropped, and why we started from a clean sheet

There's a phrase most of us use without thinking: who do you report to?

It's often the second question at a party, after what do you do? It tells you where someone sits. Above them, a manager. Above the manager, a director. Somewhere at the top, a chief executive, and above them a board. Draw it and you get a tree, branching downwards, every box connected to exactly one box above.

It feels so natural that it's hard to imagine an organisation drawn any other way. Bosses exist, surely, because they're necessary. Someone has to be accountable. Someone has to make the call. Without that, as the last chapter's sceptic put it, things fall apart.

This chapter takes a different angle on that belief. Not that bosses are bad, or that hierarchy was a mistake. Hierarchy was a brilliant answer to a real problem. The question is whether we still have that problem.

A drawing from 1855

One of the earliest organisation charts we know of was drawn in 1855 for the New York and Erie Railroad, under its general superintendent, Daniel McCallum.¹

Railways were the first businesses too big for one person to see. Trains ran over hundreds of miles, day and night, on single tracks where a mistake meant a head-on collision. Information moved at the speed of the telegraph, and only to stations that had one. McCallum's problem was simple to state and hard to solve: how do you get the right information to the people who need to decide, and their decisions back out to the people who need to act, fast enough to stop trains crashing?

His answer was a structure. Clear lines of authority, regular reports flowing up, orders flowing down, every person knowing exactly who they answered to. It looked, in fact, rather like a tree.

Over the next half century that shape spread. The historian Alfred Chandler called the rise of salaried managers the "visible hand": where markets coordinated through prices, the large firm coordinated through layers of people whose job was to watch, decide and direct.² The economist Ronald Coase asked why firms existed at all, when markets could in principle arrange everything, and answered that using the market has costs: finding people, negotiating, checking. Inside a firm, a manager can simply tell someone what to do.³ Hierarchy was cheaper.

Then came the engineers of management itself. Frederick Taylor, at the turn of the twentieth century, separated the thinking from the doing: managers would study the work and design the one best way, workers would carry it out.⁴ In France, Henri Fayol set down principles that still sit in management textbooks, among them unity of command: every person should take orders from one boss and one boss only.⁵ The sociologist Max Weber described bureaucracy as the most rational form of organisation ever devised: rules, offices, files, and a clear chain from top to bottom.⁶

None of these people were villains. They were solving the problems of their time, often brilliantly. In a world of slow information, scarce education and work that could be broken into simple repeatable steps, a tree of command was an excellent way to coordinate thousands of people.

What the tree was really doing

Strip away the history and the tree does one main thing: it moves information.

In the last chapter we put it as three jobs. The boss knows, because reports flow up to them. The boss decides, because only they can see enough of the whole. The boss tells, because orders have to flow back down. Each layer of management exists because one person can only hold so many conversations, so the tree branches, and every branch adds another person who knows a little more than those below and a little less than those above.

Hearing about things too late wasn't something that happened once. It was the norm. And when I looked to the people who'd built companies before me for guidance, the lessons were mostly about force. Rule by fear. Be right. My way or the highway. Find your leverage over people and use it. I did what they said, for a while. It never felt right.

The tree also does something quieter. It decides who matters. Information that fits the reporting lines travels; information that doesn't, gets lost. The person on the phone with an unhappy customer knows something important, but it has to survive three or four retellings before it reaches anyone who can act. Each retelling smooths it a little. By the time it arrives, it's a number on a dashboard.

What changed

Look at what the tree was built around: slow information, few people who could see the whole, and work that could be designed at the top and done at the bottom.

None of those holds any more.

Information now moves instantly, to anyone, anywhere. Most work in most organisations isn't simple repeatable steps; it's what the management writer Peter Drucker called knowledge work,⁷ where the person doing it usually knows more about it than anyone above them. And the people who see the customer, the problem and the opportunity first are almost always at the edge, not the centre.

So the tree is now trying to solve a problem we mostly don't have, and in doing so it creates new ones. It slows decisions to the speed of the busiest person. It filters out the signals that matter most. It burns out the people at the top and switches off the people at the bottom. It made sense when it was the fastest way to move information. Today it's often the slowest.

Even some of the people who run large companies have started to say this out loud. A recent essay from the founders of the payments company Block argued that the long line of experiments in running without hierarchy had never managed to replace hierarchy's real job, routing information, at scale, and that AI might.⁸ They're not alone. The question is no longer whether the tree is straining, but what takes over its job.

What we've tried since

None of this is new. People have been trying to get beyond the tree for at least half a century, and the attempts are worth taking seriously, because they're where the lessons are.

The matrix. Give people two bosses, one for their function and one for their project, so information can flow across as well as up and down. It helps a little and often makes things worse: two sets of priorities, two sets of reports, and a lot of negotiation over who gets which hours.

Flat. Take away the boss altogether. In the early 1970s the activist Jo Freeman wrote about what happened in groups that tried, in an essay called The Tyranny of Structurelessness.⁹ Her point still stings: there's no such thing as a group without structure, only groups where the structure is hidden. When nobody is formally in charge, the most confident, best connected or longest serving people end up in charge informally, and nobody can hold them to account.

Agile. Let small teams organise themselves around the work, in short cycles, close to the customer. It works well inside the team. But very often the team sits inside an unchanged tree, and the moment a decision crosses the team's edge, it goes back up the old lines.

Sociocracy. Developed by Gerard Endenburg in the Netherlands from the 1970s, building on earlier work by the educator Kees Boeke: circles that govern their own domain, decisions by consent rather than by vote or by boss, and links between circles so information flows both ways.¹⁰

Holacracy. Brian Robertson's practice, developed in his software company in the early 2000s: authority given to roles rather than people, a written constitution, and a way for anyone who senses a problem to propose a change to the structure.¹¹ It spread widely in the 2010s, most famously at the online retailer Zappos, where the move was difficult and a notable share of staff chose to leave.

Teal. In 2014 Frederic Laloux gathered stories of organisations that had moved beyond hierarchy, from a Dutch nursing organisation run by small self-managing teams to a tomato processing company where nobody has a boss, and argued that they shared a common shape: self-management, wholeness and an evolving purpose.¹²

Each of these got something important right. Each also ran into the same wall. The structure had to be held by people: meetings to run, records to keep, rules to remember, facilitators to keep everyone honest. When people got tired, or busy, or a founder quietly kept the final say, the old tree grew back underneath. Many practitioners carry scars from this, along with real insight.

Laloux's colours have an older root. From the 1950s the psychologist Clare W. Graves studied how people's values change as the conditions of their lives change. He found not a fixed ladder but ways of seeing that emerge in turn, each a response to problems the one before couldn't solve, and none of them final. Don Beck and Christopher Cowan carried his work into the world as Spiral Dynamics, and from there, by way of Ken Wilber's integral theory, it reached Laloux's colours.¹³ We hold it as lineage and as a lens: a way to understand why a person or an organisation sees the world as it does. Never as a scale to rank anyone on. Nothing in The Grove labels a person or an organisation by stage, and those who've worked longest with developmental models give the same warning: development is an arc, held lightly, never a league table.

The legacy in our heads

There's one more layer to this legacy, and it's the hardest to see.

Even when the boxes change, our habits don't. We still say line manager, report to, promotion, above my pay grade. We still look up for permission. We still feel, somewhere in the body, that speaking out of turn is risky. Generations of working inside trees have taught us how to be in one, and those lessons don't disappear because the org chart has been redrawn.

That's why so many attempts drift back. It isn't only that the tools were weak. It's that the tree lives in us as well as around us.

How we decided

Before The Grove existed, I'd already lived most of this. I'd built the tree myself and felt it strain. I'd seen flat fail.

My first encounter with self-management came from an unlikely place: the social media company Buffer, which published openly how it ran itself without traditional managers. I was shocked that it could work. That led me to Laloux's book, and then to Holacracy. I took a course with Nick Osborn of Evolving Organisation and later joined his organisation as a partner, picking up roles in marketing, the website and finding new business. I loved working that way, under his guidance. I also saw how much of it rested on people carrying the practice by hand.

Over more than five years I tried to build something better, again and again. A first rough version on WordPress. A node-based attempt. An organisational operating system inside pureDAO, an accelerator programme on the Internet Computer, in 2023. Then Lumenophore. Each one taught me something, and each one stalled for the same reason: it was too big for one person to hold.

When I took the idea to the "we", the first question was what to keep. We didn't want to throw out everything hierarchy does. It does some things well: clarity about who holds what, someone who carries a promise, work that actually gets finished on a day. We wanted those, without the tree. So we kept accountability and dropped command. We kept structure and moved it out of people and into the system itself, where anyone can see it and raise a tension about it, and the people who hold seats where a change is decided can propose one.

Accountability without command turned out to need a home of its own. In an early simulation, work reached exactly the right part of the organisation and then sat there. A small team shared it, nobody held any one piece, and it was everybody's, which meant nobody's. It was Freeman's hidden structure again, turning up inside an explicit one. So a promise in The Grove is held by a part of the organisation and carried by a named person, on a clock, and when it starts to slip it rises to the part above, where someone can step in. That's the tree's real job, done without the tree.

We also had to decide whether to build on one of the existing practices. Holacracy was the obvious candidate. It's clear, well documented, and openly licensed, and others have since taken its constitution and extended it for organisations with AI in them. We chose to start from a clean sheet instead, beginning with what Laloux described and what nature does, and only then reaching for practices that fitted. That meant more work and more mistakes. It also meant we weren't bound to a meeting format, a vocabulary or a set of rules designed for a world where everything had to be done by hand.

And we decided to take the legacy in our heads seriously. A system can't undo generations of habit on its own. But it can be built as means and guardrails: the means to work the new way made easy, and guardrails that stop an organisation quietly sliding back to the old one when people are tired or under pressure. It shouldn't dictate the organisation's shape, which belongs to the people in it. It should make the old moves hard and the new ones natural, and show people, gently and repeatedly, that their sense of what's wrong is welcome.

What you might take from this

Bosses aren't a law of nature. They were a technology, and a good one, for a time when information moved slowly and only a few people could see the whole. That time has passed.

The question isn't whether we can do without anyone being accountable. It's what takes over the job the tree used to do, now that the tree has become the bottleneck.

Notes and further reading