The Regime Is a Channel

The Letter

In the late 1970s, a young Ugandan academic stood at a security table at Entebbe airport while soldiers went through her luggage. Anyone leaving Uganda was searched, and the search was not for weapons. It was for information: names of the abducted, accounts of what was happening in the country, anything that might reach a foreign newspaper or a church committee or a conference hall in Geneva. In her papers was a letter, an ordinary letter, a page of pleasantries addressed to someone abroad. The soldiers read it and handed it back.

Between the lines of that letter, written with a fountain pen dipped in lemon juice instead of ink, was the real message: who had been taken, and where, and when. The juice dries invisible. Held over heat, it turns brown and the words come back. The recipient knew to warm the page. The state, which controlled the airport, the army, the courts, and the most powerful radio transmitter on the continent, did not.

Nakanyike Musisi told that story this morning at the Stratford Festival, at the second session of CBC Ideas' five-part series on how authoritarian regimes rise and fall. She was not describing something she had read. She was describing her own pen. She lived the whole of Idi Amin's rule as a student, took her degree from his hand because there was no one else's hand to take it from, and smuggled the news of the disappeared past his soldiers in citrus and cover text. Then she said the sentence that this essay exists to unpack: dictatorship also teaches us to be innovative.

In the first essay in this series, Mussolini's Italy taught us that resistance has a timer, that the cheap, legal, bloodless forms of self-defense expire as a regime consolidates, and that the moral weight therefore falls on acting early. Uganda poses the harder question. Italy's timer ran for years while elites talked themselves into accommodation. Uganda's ran out in a single day in January 1971, when a military commander seized a state that colonialism had equipped with an army, a police force, and prisons, but never with institutions capable of stopping him. Most people who live under authoritarianism do not get a Matteotti moment. They wake up one morning and the window is already closed. The Amin panel was, in effect, a seminar on what self-defense looks like after the timer expires, and its answer kept converging on a single, unfashionable word: channels.

One Frequency

The first essay described how a consolidating regime "puts a radio in every kitchen so that no one can fail to be reminded who is the only option." In Italy that was a metaphor for the stage in which a regime absorbs the press and occupies public space. In Uganda it was simply the machinery of government.

By an accident of timing, Amin inherited the most powerful radio broadcasting service in Africa. His predecessor, Milton Obote, had built out the network with medium-wave transmitters that could reach the distant provinces short wave never touched. The historian Derek Peterson, who has spent years in Uganda's local government archives and published a popular history of the regime last year, described at Stratford what that network became after the coup. Radio Uganda was not state propaganda in the sense we usually mean, a channel of flattery running alongside ordinary life. It was the operating system of the state. Provincial officials were obliged to listen because the president gave orders over the air, sometimes about their own districts, sometimes about the minutest matters of their own jobs. Listening was, in Peterson's phrase, a discipline of citizenship. The dictator did not send memos. He broadcast, and the country arranged itself around the broadcast.

Peterson offered the analogy himself, so I am not reaching for it: Radio Uganda played a role in the 1970s not unlike the role a certain social media platform played in American government in the late 2010s, a channel through which a newly empowered executive could issue instructions directly, publicly, and at whim, while everyone whose career depended on his mood monitored the feed. The audience in Lazaridis Hall laughed, the way people laugh when a comparison lands closer than expected.

It is worth sitting with why the comparison lands. The lesson is not that radios are dangerous or that platforms are fascist. It is that every authoritarian project is, among other things, a bid for channel monopoly. Amin did not need every Ugandan to believe him. He needed every Ugandan to receive him, on a frequency where no reply was possible, until the broadcast became the shared reality that everyone had to orient toward whether they credited it or not. A one-way channel with mandatory reception is not a medium of persuasion. It is an instrument of coordination, and it coordinates the regime's men just as much as its subjects. The soldiers at Entebbe going through Musisi's luggage were downstream of the same transmitter she was.

That is the frame in which the rest of the panel's stories snap into focus, because every act of survival and resistance the three panelists described was, at bottom, a channel operation.

The Kingdom That Did Not Listen

Ask what internal resistance to Amin looked like and the honest answer is: almost nothing survived. The exile groups operated from London and Dar es Salaam. Inside the country, the panel could point to exactly one movement that kept itself alive across the whole period, and it is not the one a screenwriter would invent. In the Rwenzori mountains on the Congolese border, a separatist movement called Rwenzururu had declared its own kingdom in 1962, before Amin, before even the crises that produced him. It had its own king, its own parliament, its own legal structure, and an army equipped with sticks and stones. It held out until 1982, outlasting Amin's regime entirely.

Sticks and stones did not do that. Peterson's account of how they actually survived is the most quietly radical thing said on that stage. The Rwenzururu maintained what he called habits of mind. They refused to listen credulously to Radio Uganda, the vehicle of the dictatorship's claims about reality. And they built what he called their own media economy, a way of communicating among themselves that centered local history, local identity, and local culture, and that treated the capital's inflated announcements about racial and political liberation as noise from someone else's transmitter. They did not jam the regime's channel, which was impossible. They starved it of the only thing a channel needs, which is reception, and they kept a second channel alive on which their own account of the world could circulate.

Notice what this is and what it is not. It is not 3.5 percent of the population in the streets; the Chenoweth threshold from the civil resistance plan assumes an unconsolidated regime with pillars that can still defect, and Amin's Uganda was past that. It is something prior to mobilization and humbler than revolt: the preservation, under a channel monopoly, of an independent loop between people who trust each other. Every future act of resistance depends on that loop existing. A strike cannot be called, a defection cannot be coordinated, a disappearance cannot even be counted except over a channel the regime does not own. The Rwenzururu matter not because a mountain kingdom is a model anyone can copy, but because they demonstrate the minimum viable unit of resistance under consolidated authoritarianism, and it is not a weapon. It is a working channel plus the habit of disbelief.

Musisi's lemon juice is the same discovery made portable. Her letters worked because the information traveled inside a channel the state could see but not read, camouflaged as the state's own approved traffic. The Entebbe search was a checkpoint on the visible network; the real network ran between the lines, secured by a shared secret about heat. And the panel supplied a third example without flagging it as one. Nam Kiwanuka, the Ugandan-Canadian journalist whose family fled on foot to Kenya, is in Canada because a woman in London, Ontario heard her father's story through a church's missionary grapevine, before social media, before the family had so much as a photograph to send. A refugee sponsorship that saved five children traveled person to person along a channel of congregations, entirely outside any state's broadcast. Three stories, one shape: the regime holds the transmitter, and life routes around it on lower power.

The Flame at Scale

It would be comfortable to leave this as history, and wrong, because the panelists themselves refused to. Kiwanuka described her relationship with social media as love and hate in the same breath: the same platforms that let Ugandans organize and document are the platforms on which criticizing the president's appearance can end in a courtroom or a disappearance, because posting is publishing your name, your face, your location, and your social graph to the same address. Uganda's current government, forty years into Yoweri Museveni's rule, has shut down the internet outright at election time. The channel that mobilizes is the channel that targets. This is not a paradox. It is what happens when the second channel, the one resistance depends on, is a rented room in a building the powerful can enter.

And here the modern situation is genuinely darker than 1977 in one specific way. Amin's censors were men. Every letter opened at Entebbe cost minutes of a soldier's attention; every intercepted message had to be read by someone; the lemon juice worked because holding each individual page over a flame does not scale. Surveillance had a labor cost, and the labor cost was the citizen's margin. Machine learning deletes that margin. A classifier reads every message, transcribes every call, matches every face, and flags every anomaly at once, for approximately nothing, forever. An AI-equipped state holds every letter to the flame simultaneously. The asymmetry that let a graduate student beat a military dictatorship with citrus has been engineered away, and the same models that read everything can now also write everything, flooding the shared channel with synthetic consensus so that the disbelief the Rwenzururu practiced becomes harder to aim.

Against that, the citizen's side of the ledger has exactly one entry of comparable weight, and it is mathematics. End-to-end encryption is lemon juice that scales: a channel that can be observed but not read, secured by a shared secret, running camouflaged inside the visible network. It is the single technology that restores the 1970s asymmetry in the citizen's favor, which is precisely why the legislative pressure against it never stops. The proposals cycle through Brussels and London and Ottawa under different names, chat control, client-side scanning, lawful access, and they all reduce to the same design: every letter shall be held to the flame before the envelope is sealed. Each proposal arrives dressed for child safety or counterterrorism, and each would build, in permanent infrastructure, the checkpoint at Entebbe. The first essay in this series argued that authoritarian infrastructure should be opposed while its abusers remain hypothetical. There is no better test case than scanning mandates, because a scanning mandate is Radio Uganda run in reverse: mandatory transmission from every kitchen.

Money, it turns out, obeys the same grammar, and the Amin lecture proved it in passing. The regime's signature domestic program was the "economic war": the expulsion of fifty to eighty thousand Ugandan Asians with whatever they could carry, the seizure of their businesses by decree, the confiscation of religious minorities' property, a tribunal that could sentence a shopkeeper to death for the price of cigarettes. Every asset with a title, a registry, or a license changed hands the moment the state decided its holder was an enemy, which is to say that registered wealth is also a channel the regime owns. What crossed the border with the expelled was what needed no registry. That story deserves its own essay and will get one, but note for now that it is the same story: the ledger is a channel, and the question of who can read and censor it is the question of who owns you.

Keep Your Own Channel

The thread running through this series is a timer. The Mussolini panel showed that the cheap forms of self-defense expire as a regime consolidates, and that the moral act is to spend them early. The Amin panel showed the other side of the clock face: what remains when the timer has already run out is exactly what you built before it did. The Rwenzururu did not improvise their media economy in 1971; they had been practicing it since 1962. Musisi's generation did not invent covert channels after the coup; they repurposed schoolgirl chemistry the regime's men had never needed to learn. The channel you control on the day the window closes is the one you kept in working order while everything still looked normal.

So the timer argument acquires a corollary, and it is the most practical sentence I can write about any of this. Acting early does not only mean protesting early, voting early, refusing early. It means building early. Use the encrypted messenger while it is legal, and normalize it among people who think they have nothing to hide, because the network is only as useful as the number of ordinary people on it. Keep a channel to the people you trust that does not pass through a platform's terms of service: the feed you subscribe to, the server in the closet, the mailing list, the congregation, the standing Sunday dinner. Hold some part of what you know, publish, and own in forms that need no one's permission to exist. None of this is exotic, and that is the point. The Rwenzururu's tools were local history and a refusal to believe the radio. Musisi's tool was a lemon.

Amin held the most powerful transmitter in Africa, an army, and a treasury fed by everything he could confiscate, and the durable resistance to him was a mountain parliament that stopped listening and a letter that read as blank. The regime is a channel. So is the alternative to it, and only one of them can be built after the coup. The rule has a timer, and one of the things the timer is counting down is the time you have left to build channels nobody else owns.