Fragments and Connection

When war came, it arrived at Bletchley without fanfare. The great house stood behind hedges and wire on the edge of town — lawns, trees, a lake — ordinary at a glance, except for the guards at the gate and the blackout curtains. Soon, the grounds filled with wooden huts, thrown up in haste to house the ever-growing ranks, their corridors thick with the smell of damp coats and cigarette smoke. Typewriters clattered through the night.

Alan Turing arrived in September 1939 carrying a small suitcase and a head full of mathematics. In another life, he might have stayed in Cambridge, writing papers that few others understood. But the war had changed the utility of intellect. Britain needed minds that could see patterns others couldn’t.

The problem was Enigma — a cipher machine the Germans believed unbreakable. Each letter passed through rotors and plug-boards that scrambled the alphabet into trillions of combinations, resetting every midnight. To most, unlocking it would have seemed impossible.

Turing began, as he always did, with logic. He spent nights walking the grounds, thinking about permutations and probabilities — how to turn a problem of uncertainty into one of speed. The idea that came from those walks changed everything: a machine that could test combinations faster than any human. His design reimagined an earlier Polish concept, building a new generation of Bombes able to uncover Enigma’s secrets at industrial pace.

This story, the one we tell about Bletchley and the one brilliant mind cracking the code that saved civilisation, says as much about us and our desire for neat stories as it does about him. We prefer the heroic singular to the untidy plural.

Because the truth of what made Bletchley powerful was far more complex.

Next door, Gordon Welchman was sketching diagonal lines across a blackboard — a tangle that looked like chaos but changed everything. His “diagonal board” linked all twenty-six letters at once, allowing false solutions to be discarded in seconds. Without it, Turing’s machine might have worked, but far too slowly to matter.

Across the courtyard, Joan Clarke, one of the few women recognised as a cryptanalyst, was testing a probabilistic method she called Banburismus. Her team spent long days comparing letter patterns by hand, narrowing the odds through logic and patience. Their work broke the German Navy’s code, saving thousands of tons of Allied shipping and countless lives.

A few huts away, Bill Tutte was studying fragments of intercepted messages from an entirely different system — the Lorenz cipher used by Hitler’s own staff. No one in the Allies had even seen the machine that produced it, yet Tutte deduced its inner wiring from the patterns alone — a leap of reasoning as astonishing as anything Turing achieved.

Tommy Flowers, a Post Office engineer drafted in to help, built the machine to test Tutte’s theory. It used vacuum tubes instead of rotors — faster, temperamental, and entirely new. Flowers believed in it so much that he paid for parts himself when the Treasury refused. The result was Colossus, the first programmable electronic computer.

And around them all, thousands more: Wrens running Bombes in rotating shifts, clerks ferrying files from hut to hut, linguists translating intercepts, analysts searching for patterns. Mathematicians, engineers, typists, administrators — each holding a fragment that only made sense in combination.

We have to remember that Bletchley was run by the military, with all the rigidity and secrecy that implies. Orders came through the chain of command; information travelled on a need-to-know basis. Most people never saw the full picture — and that was by design. Yet necessity has a way of bending rules. Notes were slipped between huts. A casual word over tea might solve what weeks of process could not.

Understanding spread in the gaps. The official system kept the operation safe; the unofficial one made it smart. Order protected security; curiosity drove discovery. The miracle wasn’t the brilliance it contained, but that those fragments connected fast enough to matter.

As Hannah Arendt wrote, power comes from people acting in concert. Bletchley showed what that looks like when an organisation listens to itself — fragments combining into collective intelligence, a kind of polyphony: many voices, thinking together.

What happened at Bletchley wasn’t just an episode in wartime ingenuity.

It was a working example of something organisations still struggle to achieve: a system that can hear itself think.

This chapter is about what companies now call employee voice — the umbrella term for all the ways people are asked to share their views at work, and told they have a say: surveys, meetings, suggestion schemes, “speak-up” channels — the rituals that fill company calendars, so routine they’re almost invisible. We rarely stop to ask what they’re really for — or what they make possible.

At first glance, Bletchley doesn’t belong in that story at all. There were no engagement surveys, no town-hall Q&As, no programmes to encourage contribution. And yet what happened there — the movement of fragments, the combining of partial insights into shared understanding — is what voice, at its best, is supposed to do. It’s what most organisations are still trying to achieve, even if they go about it in very different ways.

The Fight to Be Heard

For most of the industrial age, being heard at work wasn’t a privilege or a programme — it was a fight.

Factories were dangerous places — hot, loud, and relentless. Hours were punishing, and losing your job could mean losing your home. There were no channels, no protections, no promise that anyone in power would listen. Voice wasn’t soft or symbolic; it was survival.

The first acts of what we’d now call voice weren’t about being heard — they were about being seen. Petitions, walk-outs, and strikes were acts of resistance born of necessity more than ideology. They forced power to notice what it preferred not to see. The fight was for dignity before it was for democracy.

Unions gave that resistance shape. By combining numbers into leverage, workers turned grievance into bargaining power. They didn’t ask to be heard; they made themselves unignorable. Through collective strength came the foundations of fairness we now take for granted: limits on working hours, safer workplaces, paid leave, pensions, redundancy pay. The things we get to enjoy as rights today were once demands shouted across picket lines.

As the twentieth century advanced, many of those demands were absorbed into law. Governments codified what solidarity had fought for — protection from exploitation, limits on child labour, a minimum wage, the right to organise. The law began to do what unions once had: act as a counterweight to corporate power. Voice, for the first time, could exist without revolt.

But the systems built around it didn’t evolve with it. They knew how to manage confrontation, not conversation. When the shouting stopped, silence looked like progress. Management mistook compliance for harmony, and the idea of voice began its long slide from necessity to nicety.

Order and Understanding

After the war, management turned its attention to morale. The Human Relations movement — inspired by the Hawthorne studies of the 1920s and 30s — showed that people worked better when they felt noticed. The idea took hold in the 1940s and 50s as companies sought stability after years of upheaval.

Listening became a management tool — not a way to share power, but a way to improve performance. It sounded humane, and in part it was. But it also gave management a new kind of control: softer, psychological, and harder to resist. It was progress, but on management’s terms.

Voice became something to be managed — an input to productivity, not a mechanism of influence.

Companies built consultation architecture. Joint committees appeared; in parts of Europe, works councils and co-determination gave employees formal standing in decisions. Where those structures carried weight, the system could still hear itself. Where they didn’t, the form remained but the force thinned. The meeting existed; the listening didn’t.

For a time, this balance held — an uneasy truce between authority and agency. Yet even in that apparent harmony, a new problem was forming. Voice had become procedural. It was no longer the messy, self-organising conversation that forced systems to adapt; it was a designed process, tidy and containable.

The energy that once came from necessity was redirected into rituals that signalled participation without ever threatening control. Factories gave way to offices, but not much else changed. People were still told how to feel — only now it came with a smile.

Where the industrial age treated voice as conflict, the managerial age treated it as cooperation. Both reduced the noise, but only one reduced the learning.

By the late 1960s, organisations had begun to believe they could domesticate dissent. The language of conflict was being replaced by the rituals of consultation. The next great question was no longer how to manage people, but what responsibility organisations owed to them — and to society itself.

That shift in responsibility reshaped the meaning of voice. It had begun as a demand for fairness — a way to make intolerable conditions visible. But as management theory took hold, that demand split in two. One version of voice remained oppositional, the domain of negotiation and dissent. The other became psychological: the search for motivation, cohesion, and belonging.

The two strands never reconciled. As legal protections expanded, the first lost urgency; as organisations turned voice into a tool for performance, the second grew louder. Yet the emotional equation barely changed. Speaking up still carried risk, even when the posters said it was safe. The language softened, but the stakes didn’t.

Exit or Voice

In 1970, the very same year that Milton Friedman published The Social Responsibility of Business Is to Increase Its Profits — the essay we explored in Primacy — another economist was asking a different kind of question.

Albert Hirschman, a German-born development economist, wanted to understand how people respond when the organisations or institutions they depend on begin to fail. In Exit, Voice and Loyalty, he argued that members of any system — a firm, a government, a community — have two options: to exit (leave) or to use voice (speak up and try to change things from within).

Friedman and Hirschman weren’t in dialogue, but they were both responding to the same turbulence: the late-1960s collision of social activism, economic uncertainty, and institutional fatigue. Where Friedman sought stability through focus and withdrawal, Hirschman searched for renewal through participation.

It was a neat pairing of ideas for its time. In a world of stable employment and local economies, both exit and voice still carried weight. Workers could, at least in theory, leave one job for another, and “speaking up” carried visible risks but also clear pathways — unions, councils, committees. Hirschman’s framework made sense in a world where both options were real and voice still had a collective form.

Half a century later, both options have thinned.

I once asked the CEO of a regional building society about his organisation’s remarkably high engagement score. “Eighty-nine per cent,” he said proudly. “Wow, that’s impressive,” I replied. “So, presumably, if you were to suggest a change in direction, people would be quick to move with you?” He smiled. “Oh no. Change here is like wading through treacle.” Then, he paused for a moment before offering this: “Perhaps it’s because they have very little alternative. We’re the biggest employer in most of our towns. In many of them, there really isn’t anywhere else to go.”

That, in a sentence, is the flaw in Hirschman’s logic — and in the feedback rituals that pretend to replace it. Exit depends on alternatives. When those vanish, voice becomes the only route to agency — but it bends around risk, finding safe words instead of true ones.

And while voice was softening, exit was thinning. Globalisation hollowed out local economies. Technology detached work from place. Exit became easier on paper and harder in practice. You can click apply anywhere, but mortgages, childcare, and fear hold people still. The psychological cost of leaving is high; the organisational cost of not listening is hidden. Voice, then, should have become more important. But it didn’t.

In Hirschman’s model, exit and voice worked in tension — one silent, one noisy, both corrective. Exit signalled failure through departure; voice signalled it through dissent. Loyalty held them in balance: it delayed exit just long enough for voice to matter.

But that balance has collapsed. As exit has become harder — economically, socially, psychologically — the need for voice has only grown. Yet dependence makes voice dangerous. What Hirschman saw as loyalty now often looks like necessity — people stay because they must, not because they choose to. At the very moment systems most need to hear their members, those members have most to lose by speaking.

This is also why retention is such a misleading comfort. It measures constraint as though it were commitment. People stay for many reasons — habit, fear, dependency — and the longer exit is constrained, the quieter those reasons become. What looks like loyalty may simply be the exhaustion of alternatives.

Hirschman had imagined something far more hopeful. He saw voice as a vital mechanism of renewal — a way for systems to hear that something was wrong before it was too late. When people spoke up, protested, or pushed back, it wasn’t disloyalty but devotion: the warning signal of those still invested enough to care. Voice, in his view, was the act of staying — the effort to repair what might otherwise collapse.

The Theatre of Participation

By the middle of the twentieth century, the study of work had turned inward. The language of conflict and fairness gave way to one of cooperation and participation. What had once been a struggle for agency was becoming a science of motivation.

Thinkers such as Douglas McGregor and Rensis Likert translated the messy, human work of cooperation into something that could be designed and managed. McGregor’s Theory X and Theory Y treated motivation as a reflection of managerial belief; Likert’s System 4 turned participation into structure and measurement. His name survives in the scales used to quantify experience — a fitting legacy for a discipline that learned to mistake measurement for understanding. Their ideas promised a more enlightened kind of management, but they also changed the meaning of involvement. The Human Relations pioneers replaced the moral claim of participation — we deserve a say — with the managerial technique of participation — we’ll create structures for you to have a say. It was the start of domestication.

Management science reframed cooperation in instrumental terms. Instead of asking How do we understand one another? it began asking How do we get more from people? Participation, which had once been a bottom-up, organic feature of collective work, became a top-down, managed, permissioned process. This is where “employee voice” began to be institutionalised — safe, formal, and controlled.

Management succeeded in pacifying conflict, but at the cost of vitality and truth. It was the birth of theatre: everything looked cooperative, but nothing really moved.

Chris Argyris saw what was happening behind the curtain. He turned his attention to a puzzle that still defines organisational life: why do companies that say they value openness and learning behave in ways that prevent both?

He noticed the gap between what people say they value and what they actually do — the same gap we explored in Distance, between the declared and the real.

Leaders espoused principles like empowerment and honesty, but their true theories-in-use — the unwritten rules that governed behaviour — demanded caution, deference, and control. To protect themselves and others from embarrassment or threat, people developed what he called defensive routines: habits of tact and politeness that made real learning impossible.

These routines weren’t malicious; they were professional. The more educated, well-intentioned, and high-performing the organisation, the better it became at avoiding the discomfort that genuine reflection requires. Argyris called this paradox skilled incompetence — the ability to look constructive while ensuring nothing truly changes.

It was the missing piece in the Human Relations story. McGregor and Likert had reimagined cooperation as something that could be designed; Argyris showed why that design never quite worked. You could measure morale, codify participation, and train managers to listen, but as long as people feared the consequences of honesty, the system would choose safety over truth.

He also gave language to what later scholars like Amy Edmondson and Megan Reitz would rediscover: that silence isn’t a moral failing or a lack of courage — it’s a predictable outcome of systems that reward compliance. Argyris revealed that the theatre of participation was already in rehearsal. Organisations didn’t just manage work; they managed the appearance of learning.

Argyris had shown how participation could become performance. The next generation made a business of it.

Just as Milton Friedman’s argument for shareholder returns became the doctrine of shareholder primacy, William Kahn’s work on engagement grew into an entire industry. It’s hard to imagine either man could’ve foreseen how dominant their ideas would become; both were amplified and flattened by others until their ideas hardened into management’s operating logic.

Kahn had described the psychological conditions that allow people to bring their full selves to work — presence, meaning, safety — but his insight was quickly distilled into a dashboard metric. Listening became an abstraction: something to be measured rather than experienced. Yet Kahn’s work had never been about metrics. He wasn’t trying to quantify feeling; he was trying to understand what makes people genuinely show up. When those conditions were met, people didn’t just perform; they inhabited their roles. Work became an act of expression, not endurance.

It was a deeply human insight — the opposite of the mechanised engagement industry that followed. Consulting firms such as Gallup, Hewitt, and Towers Perrin seized on the language of engagement and turned it into a management product. They reduced Kahn’s quiet psychology to survey instruments and indices: engagement as a percentage, belonging as a score. What began as a study of human presence became a marketplace of measurement. I know I’m part of this story, and we’ll come back to that.

The irony is hard to miss. Kahn had shown that people engage when they feel seen and safe; organisations responded by asking them how engaged they felt. The question replaced the condition.

Amy Edmondson later gave this dynamic its clearest form: psychological safety. Her research showed why even well-intentioned workplaces fall quiet — because the risks of saying something awkward or unwelcome still outweigh the rewards. She made explicit what Argyris had seen decades earlier: organisations create the conditions for silence and then call it culture.

Megan Reitz and John Higgins extended the point. Speaking up, they argued, isn’t a personal trait but a negotiated permission. Who talks, who listens, and what gets said depend less on bravery than on power — the subtle, everyday signals that grant or withdraw the right to speak and the duty to hear.

Courage, in that sense, is a symptom of design. In healthy systems, it shouldn’t be required.

Think of almost any company you’ve worked for. The pattern’s familiar. An employee forum is launched. Representatives gather, collect concerns, and feed them up. Minutes are tidy; actions logged. But the issues that matter most — workload, headcount, priorities — require trade-offs no one in the room can make. The forum becomes a place where people learn what won’t change. Or it sets up a hotline: invaluable for reporting wrongdoing, useless for learning. The nuclear button of organisational life — everyone knows it’s there, but only a few are willing to press it, and never without consequences.

In theory, that button protects integrity; in practice, it replaces it. When the only safe form of dissent comes after the damage is done, silence has already won.

The shift from confrontation to consultation made workplaces safer and conversations cleaner — but it also made it easy for leaders to mistake procedure for progress. The heatmap, the survey, the town hall: each creates the impression of dialogue without the movement.

We built systems that could display participation without ever having to experience it. And somewhere in that pursuit of order, voice became so sanitised, so safe, that the system stopped seeing itself. It stopped hearing its own contradictions, its own weak signals — the sounds that tell a living organisation where it really stands.

The Noise That Made Sense

In the early years of Harkn, we spent an extraordinary amount of time and money trying to isolate and refine our ability to produce signal — to take the messiness of what people said and turn it into something cleaner, neater, more directly actionable.

Partly, this was in response to pressure from our clients: tidy the data, benchmark us against others, tell us what to do. It wasn’t a surprise. Organisations like clarity — dashboards, trends, answers they can compare. If we could filter out the noise, we told ourselves, leaders would finally hear what mattered.

But Harkn wasn’t a survey. We didn’t ask people questions about their opinions or experiences. We invited them to share small, daily reflections about them — micro-stories about how they were feeling and what was shaping their day. Not reports, but moments. Over time, those snippets built a map of life at work, as it was actually lived.

At first, it was all too easy to consider what people were sharing as trivial, and often it looked like noise. People talking about breakfast, the weather, a bad meeting, and the walk they took at lunchtime. Benign commentary. But those small, ordinary posts were the fabric of working life. And we learned that when people shared these stories, it meant things were okay — that the organisation was breathing normally.

We learned that people would talk about almost anything — and that was the point. Small talk built fluency; the habit of sharing the ordinary made it easier to share what mattered. The daily flow of posts gave everything else its context. Variation in tone, energy, and attention showed what was normal and what wasn’t. When humour faded, when lightness turned to caution, when stories flattened into facts, something underneath was shifting. Because it was peer-to-peer, people weren’t performing for management; they were exploring together — comparing their experiences with others’, testing what could and couldn’t be said, and seeing what happened when they did. The noise began to carry meaning.

Anonymity made that exploration possible. It wasn’t disguise; it was protection — a shared assurance that honesty wouldn’t carry a cost. Because people didn’t have to manage impression or hierarchy, they could drop the performance and stay with the truth a little longer. Over time, that safety created something close to collective candour: people tested ideas, surfaced tensions, and admitted uncertainty. It wasn’t rebellion; it was reality, finally visible.

What at first looked like noise was often people working things out together — early, unpolished, essential. It was the organisation thinking.

Someone would describe an odd experience — “Is it normal that…?” Others would respond, gently at first, then with recognition. Those half-formed exchanges were the earliest signs of deeper issues: behaviours or dynamics that might otherwise have gone unseen until they became formal grievances.

The same pattern appeared when people were solving problems. Innovation rarely arrived as a clear proposal; it began with someone asking, “Why does this keep happening?” or “Is it just me or…?” The back-and-forth — the questioning, the testing — was the work. The noise was where issues were crystallised and the early steps of invention lived.

Even risk had a rhythm. Before anyone said something that challenged power, there were small tests — tentative posts, mild provocations, phrases designed to see if it was safe.

As people shared what was on their minds — the good days and the bad, the irritations, the small wins — something began to take shape. The same questions surfaced in different places; the same frustrations appeared in new words. People started to notice the echoes, and that noticing mattered. The organisation was beginning to understand itself — not through a survey or a meeting, but through the steady rhythm of ordinary conversation.

Dialogue, in this sense, was different from what most were used to. It wasn’t feedback, and it wasn’t upwards. It was speaking into the system — not asking someone else to fix things, but trusting that someone, somewhere, would pick it up.

And mostly, they did. Someone would spot a comment that resonated, offer a different view, or simply say, “I’ve seen this too.” They didn’t need the COO to fix a glitchy laptop — although, memorably, that happened once. What mattered was that people stepped forward where they could: solving problems in their own remit, adding clarity, connecting dots. Understanding moved to where it was needed, without waiting for permission.

We’d seen a similar kind of movement before — at Bletchley, where fragments of insight passed quietly between people until they formed something whole. Much of it happened off the record, across huts and hierarchies, outside the sanctioned lines of communication. The breakthroughs came not from authority but from exchange — from people thinking together, despite the system.

The lesson, then, is the same now: intelligence doesn’t live in one head or at one level. It happens between people, in motion, when the system gives itself permission to think aloud.

Many Voices, One System

Modern organisations are complex, adaptive systems — dispersed, interdependent, full of local variation. No single vantage point, certainly not the view from the top, can ever see how things really work. Leaders operate on abstractions: reports, dashboards, survey scores — the fossilised remains of past conversations.

Yet the people inside the system already know what those abstractions can’t show. They feel how policies land, how culture bends under pressure, where workarounds and contradictions live. Their everyday talk — the questions, irritations, hunches — is the organisation’s living intelligence. Voice, in that sense, isn’t feedback. It’s how awareness circulates.

But the systems built to capture that awareness were designed for a different age — one of stability, hierarchy, and control. They assumed that understanding could be collected, codified, and managed. In today’s world — fluid, distributed, fast-moving — that logic no longer works. The old rituals of listening can’t keep pace with the reality they claim to represent.

Voice hasn’t been lost; it’s outgrown its containment. The challenge now is to rebuild it for complexity: less as a channel for expression and more as a medium for sensemaking — a way for organisations to stay in live conversation with themselves.

To do that, we have to think in the plural. Voices — overlapping, contradictory, incomplete — are what keep a system awake to its own conditions. When they move freely, fragments connect, and collective understanding emerges.

That’s what we saw at Bletchley: intelligence forming not from control but from connection. The same is true today. Awareness doesn’t live in dashboards or individuals; it lives in the movement between them.

What it means to listen is what comes next.