# The Gap *On the fragility of power and the space where it fails* By [Lynne Heartwing](https://paragraph.com/@lynne-heartwing) · 2026-04-19 powerisanevent, theinterval, refusal, stateviolence, politicaltheory --- On the morning of 25 April 1974, a Portuguese soldier stood at a checkpoint in Lisbon, rifle in hand. The orders were clear. The regime he served had instructed its forces to maintain control. If necessary, they were to fire. And yet, when civilians approached, something unusual happened. He did not shoot. Instead, someone placed a carnation into the barrel of his gun. Across the city, similar moments unfolded. Orders were issued, but not carried out. Soldiers hesitated, delayed, or quietly refused. The machinery of authority did not collapse in a single confrontation. It simply failed, piece by piece, to function. By the end of the day, the regime was gone. We often describe such events as revolutions, as if a structure of power were overthrown by an opposing force. But this description obscures something more fundamental. Power did not collapse all at once. It failed to _happen_. Fifty-two years later, half a world away, a different scene unfolded. On the morning of 24 January 2026, Alexzandria Prettey, a 37-year-old ICU nurse and US Army veteran, stood on a street corner in Minneapolis. She was filming an operation by Immigration and Customs Enforcement agents. She was not a migrant. She was a citizen. She was not resisting. She was recording. Within five seconds, she was on the ground, subdued, and shot ten times. The first shots came after she was already kneeling, hands controlled. The final shots came as she lay motionless. The agents completed their actions. The interval between command and execution closed without hesitation. Power did not fail. It succeeded—in its most terminal form. Between these two moments lies the question this essay will pursue. Not what power _is_, but when it _happens_. Not where it resides, but how it succeeds or fails in the smallest unit of political life: the space between a command and the act that realises it. * * * **I.** ------ We tend to think of power as something that exists: a structure, an institution, a system that persists over time. Governments, bureaucracies, hierarchies appear solid, continuous, durable. They are treated as objects in the world, things that can be seized, held or dismantled. Political thought has long attempted to locate power within stable forms. Max Weber described it as the probability that one actor can carry out their will despite resistance. Hannah Arendt distinguished power from violence, grounding it in collective support rather than coercion. Michel Foucault dispersed power across networks of knowledge, discipline and discourse. Each of these accounts moves away from crude notions of domination. Yet they share a deeper assumption: that power, however contingent, is something that _exists_. But existence, in this sense, is misleading. Power is not a thing. It is an event. To see this, we must look not at institutions but at moments. A law is written. A command is issued. A directive is communicated. None of these, by themselves, constitute power. They are conditions, not outcomes. Power occurs only when someone acts. A police officer enforces a rule. A clerk processes a document. A soldier pulls a trigger. In each case, power does not reside in the system. It emerges at the instant the action is performed. Without that action, the system remains inert. This brings us close to the insight of Gene Sharp, who argued that political power depends on obedience and cooperation. When people withdraw that cooperation, regimes can unravel with surprising speed. Yet this formulation leaves something unresolved. It tells us that power depends on obedience, but not _how_ obedience unfolds. * * * **II.** ------- Every act of obedience contains a gap. Between receiving a command and carrying it out, there is a brief interval in which the action has not yet been completed. Within that interval, the outcome remains open. The command has been given, but not yet realised. In that space, obedience is still reversible. This reversibility is not accidental. It is structural. To obey is not to exist in a state of obedience, but to _perform_ an act. And every act, before it is completed, can still be interrupted, redirected, or abandoned. Within that interval, we can identify three distinct layers: * **The cognitive gap**: the command must be understood, but understanding is never automatic. Orders arrive with ambiguity, contradiction, plausible deniability. The actor must interpret. At this level, the act can still fail through confusion, through the quiet decision to read a directive differently than intended. * **The volitional gap**: the command must be accepted. Here conscience, loyalty, fear, and conviction converge. The actor experiences the command as something that can be embraced, resented, or refused. Even when understanding is clear, willingness remains uncertain. At this level, the act can still fail through hesitation, through doubt, through a refusal that never announces itself but simply never resolves into consent. * **The execution gap**: the action must be initiated. Here the body enters. The hand must move, the voice must speak, the trigger must be pulled. Between intention and action lies a final threshold that can always be crossed—or not. At this level, the act can still fail through delay, through the strange inertia that sometimes seizes the body when it is asked to complete what the mind has already accepted. Power depends not on obedience in general, but on the successful alignment of all three. The soldier can lower his weapon. The official can delay the signature. The worker can hesitate before executing a task. Power depends on the _closure_ of this interval. But that closure is never guaranteed. * * * **III.** -------- The political significance of this interval becomes visible in moments of rupture. Consider another recent event, this one from the other side of the world. In September 2025, thousands of young people took to the streets of Kathmandu. The immediate trigger was a government decision to ban 26 social media platforms, including Facebook, Instagram and WhatsApp. Nepal had 14.3 million social media users—nearly half its population—and for the generation born after the monarchy was abolished in 2008, these platforms were not entertainment but infrastructure. They were how work was found, how news was shared, how the world was known. But the deeper current beneath the protest was something else: a rising anger at corruption, at nepotism, at the political dynasties whose children posted lavishly on the very platforms the government now claimed to regulate. The protesters called them “Nepo Kids.” The government called the protesters a threat. When police moved to disperse the crowds, the interval began to fail. Officers were given orders. Some followed them. Others hesitated. Some lowered their batons. By the second day, protesters had breached the parliamentary complex, the prime minister’s residence had been set ablaze, and nearly 900 prisoners had escaped from a collapsed detention facility. Within 48 hours, the prime minister resigned. A new government was formed. The bans were reversed. What happened in Kathmandu was not a revolution in the classical sense—no organised vanguard, no declared ideology, no seizure of state apparatus. What happened was something quieter and, in its way, more revealing. Across thousands of small moments—a policeman stepping aside, a bureaucrat refusing to process an order, a soldier declining to advance—the interval between command and execution failed to close. The system did not collapse because it was defeated. It collapsed because, at enough points along its distributed chain, the actions that would have sustained it were simply not performed. * * * **IV.** ------- This is the distinctive logic of modern power. It does not reside in a single throne or a single command. It is distributed across chains of delegation, across bureaucracies, across systems of management and administration. No single actor appears fully responsible for the final outcome. Responsibility is fragmented across the chain. And yet the system still depends, at every stage, on the same condition: that someone, somewhere, completes the act. This structure resembles what might be called a system of loaded magazines. Multiple actors each hold the potential for action. Each step prepares the next. Each component is ready. But the final outcome depends on a sequence of completed moments. * **At the structural level**, this distribution ensures that no single point of failure exists. The system can absorb hesitation at one node because other nodes will compensate. Authority is not concentrated but dispersed. * **At the behavioural level**, each actor operates within a limited scope. The clerk does not see the full chain. The engineer does not witness the final consequence. Each performs a local action that appears, from within, to carry no decisive weight. * **At the ethical level**, this produces a peculiar condition. Responsibility is diffused to the point of disappearance. No one decides the outcome. Yet the outcome still occurs. Power requires execution. Harm requires the same condition. No one decides. Everyone contributes. * * * **V.** ------ If the interval’s failure to close explains how regimes can dissolve without being overthrown, its closure explains something darker: how violence becomes routine, how the state kills without appearing to decide, how the machinery of harm operates without any single hand on the lever. The events in Minneapolis in January 2026 are a case in point. The agents who killed Alexzandria Prettey were not rogue operators. They were acting within a structure that had, months earlier, been flagged by internal oversight as showing a sharp increase in use-of-force incidents. They were trained in protocols that blurred the line between legal requirements and operational shortcuts. They operated under a legal doctrine known as qualified immunity, established by the Supreme Court in 1982, which shields officers from liability unless they violate “clearly established” law—a standard so high that it often makes accountability impossible. And they were protected by a system of fragmented responsibility that mirrors the loaded magazine structure: ICE has its own Office of Professional Responsibility, which shares oversight with the Department of Homeland Security’s Office of Inspector General. In practice, these bodies rarely produce public accountability. They investigate, they report, they recommend—but the officers remain on the job, the patterns continue, and the next interval closes as seamlessly as the last. What makes such violence possible is not simply malice or brutality, though those exist. It is the systematic elimination of the interval. * The **cognitive gap** is closed by training that scripts responses until they become automatic. * The **volitional gap** is closed by a culture that treats hesitation as weakness. * The **execution gap** is closed by the physical momentum of the moment—the body moving before the mind can intervene. The agents who killed Prettey did not, in all likelihood, experience themselves as making a decision. They experienced themselves as _responding_. The interval collapsed before they could inhabit it. This is the deepest function of modern power: not to coerce, but to compress the interval until it disappears from conscious experience. When the interval vanishes, so does the possibility of refusal. The soldier does not choose to fire; he fires. The agent does not choose to shoot; she shoots. The system achieves its perfect state: action without agency, obedience without decision, harm without responsibility. * * * **VI.** ------- Yet the interval cannot be eliminated. It can only be displaced. Consider the technological systems that increasingly mediate authority. Automation and artificial intelligence are often presented as ways to reduce uncertainty, to eliminate hesitation, to ensure consistent execution. In effect, they attempt to close the interval permanently. But the interval does not disappear. It relocates. It moves into system design, into parameter choices, into oversight and intervention. The question is no longer whether a single agent will comply, but whether the system as a whole will continue to operate as intended. The engineer who sets the parameters operates within the same interval as the clerk who signed the document. The user who accepts the system’s output stands where the soldier once stood. The gap is never eliminated. It is only displaced. And with each displacement, it becomes harder to see—and therefore harder to interrupt. This is what makes the contemporary landscape of power so deceptive. On one side, we have regimes that dissolve in days, as Nepal did, because the actions that sustained them are withdrawn. On the other, we have structures that kill with impunity, as ICE did in Minneapolis, because the actions that constitute them are completed without hesitation. Both are expressions of the same mechanism. Both depend on the interval. Both reveal that power is not a possession but a performance. The difference lies in whether the interval is inhabited or bypassed. In Nepal, enough actors stepped into the space between command and action to allow refusal. They hesitated, they delayed, they quietly failed to comply. The result was not a heroic revolution but a distributed failure of execution. In Minneapolis, the interval was compressed to the point of invisibility. The agents did not inhabit the space where refusal might have occurred. They moved through it without pause, and a woman died. * * * **VII.** -------- If power is an event that occurs only when an action is completed, then the most fundamental political capacity is not the ability to seize power but the ability to _inhabit the interval_. To dwell in the space between command and execution long enough to recognise that the outcome remains open. To see that what appears as necessity is actually a series of choices that have been made to appear as necessity. To refuse the compression of time that collapses the gap into automatic response. This is not a call to heroism. The structures that compress the interval are powerful precisely because they make heroism almost impossible. They fragment responsibility, distribute accountability, and train response until the interval disappears. But the interval is never entirely gone. It remains, however compressed, however hidden, however disguised as mere reaction. And in that compressed space, the possibility of refusal persists. * The Portuguese soldier who received a carnation in his rifle barrel did not, perhaps, experience himself as making a world-historical decision. He simply did not fire. * The Nepali policeman who stepped aside did not announce a political philosophy. He simply did not advance. * The ICE agent who shot Alexzandria Prettey did not, in all likelihood, experience herself as closing a moral possibility. She simply fired. The difference was not in the structure of power, which in all three cases was formidable. It was in the relationship to the interval. In the first two, the interval was inhabited. In the third, it was bypassed. * * * **VIII.** --------- This reframes how we understand political action. The goal is not always to _seize_ power—which, if power is an event rather than a thing, may be a category error. The goal is to _inhibit its performance_. To ensure that, at enough points along the distributed chain of execution, the interval does not close. To make hesitation contagious, delay visible, refusal thinkable. The carnation in the gun barrel was not a counter-structure. It was an interruption of the interval. It was an act that said: _the performance ends here._ This is also why the question of accountability matters so urgently. When the interval is compressed to the point of invisibility, when action follows command without the space for refusal, responsibility becomes impossible to assign. The system produces outcomes—deportations, arrests, killings—without any single actor appearing to have decided. The loaded magazine fires, but no one pulled the trigger. This is not a bug in the system. It is the system functioning exactly as designed: distributing action so that no one bears the weight of the outcome, while ensuring that the outcome reliably occurs. * * * **IX.** ------- To recognise the temporality of power is not to deny the role of institutions, beliefs or structures. It is to understand their dependence on something more immediate. Power does not reside in systems. It exists only in their enactment. And enactment is never guaranteed. At each moment, an action must be completed. At each moment, the interval must close. At each moment, the possibility of reversal must be overcome. Most of the time, this happens seamlessly. The gap is invisible. Obedience appears automatic. Power seems stable. But this appearance is deceptive. Because the stability of power is not a property of the system itself. It is an emergent effect of countless successful actions. And as the events in Lisbon and Kathmandu suggest, those actions can fail. Not necessarily through confrontation, but through something quieter: a hesitation, a delay, a refusal to complete the act. As the events in Minneapolis suggest, they can also succeed—in ways that reveal what power is capable of when the interval is compressed beyond recognition. * * * **X.** ------ Power is performed. And therefore it is unstable. And therefore it is never fully owned. Power persists only because, at the final moment, someone continues to act. The clerk clicks the button. The engineer approves the deployment. The agent fires the weapon. Each time, the interval closes. Each time, the potential becomes actual. Each time, what could have been otherwise solidifies into what is. And yet—at every threshold, in every unfinished moment, before every action resolves into completion—the possibility remains. The same possibility that opened in Lisbon on that April morning. The possibility that the soldier will not fire. That the official will not sign. That the operator will not press the button. That the performance, at the final moment, will be withdrawn. There is no guarantee that it will not. There never was. The only guarantee is that the interval exists. The space between command and action, between order and execution, between what is demanded and what is done. In that space, however compressed, however hidden, however disguised, the possibility of refusal remains. And in that space—not in institutions, not in structures, not in the accumulation of force—the question of politics finally resides. * * * To the extent possible under law, this work has been waived of copyright and dedicated to the public domain. For details, see the [**Creative Commons CC0 1.0 Universal Public Domain Dedication**](https://creativecommons.org/publicdomain/zero/1.0/). --- *Originally published on [Lynne Heartwing](https://paragraph.com/@lynne-heartwing/the-gap)*