The Org Chart Travels

A regional office outside Kuala Lumpur runs a ninety minute module on unconscious bias. The slides were built in California. They have been localized, which means the stock photography was changed and the text was translated, and they retain a case study involving a hiring panel, a candidate with a name the local staff cannot place, and a legal framework that does not exist in Malaysia. Everyone completes it. Everyone is bored. Nobody, in the entire chain between the vendor and the room, was ever asked whether the thing being described is a problem in that country.

This is what the export of an ideology actually looks like, and noticing it correctly matters, because almost every argument about whether the phenomenon is global proceeds as if ideas travel the way ideas are supposed to travel: by persuasion, in the local language, contested by people who live there. That is not the mechanism. The mechanism is procurement.

Start by separating three items that are habitually confused: a word, a set of underlying disputes, and a machine for distributing policy.

The word is American and carries American history. It comes from Black vernacular speech, an admonition to stay alert to racial injustice, traceable in recordings to the 1930s, revived through a song in 2008 and through the protests that followed Ferguson in 2014. Around 2019 it was captured by its opponents and inverted into a pejorative, which is the ordinary fate of any term that becomes useful to the other side.

The consequence is that outside America the word functions almost exclusively as a weapon. Nobody in Lyon or Seoul describes themselves as woke. They are described that way, usually by a politician, usually in a sentence about foreign contamination. Any question of the form does wokeness exist in country X therefore has a trivial answer about vocabulary and a serious answer about substance, and the two get switched in most conversations.

The serious answer begins with what actually crosses a border reliably, and it is not an argument. It is an organizational chart.

A multinational adopts a policy at headquarters. The policy propagates to subsidiaries because consistency is easier to administer than variation and because counsel prefers a single global standard to forty local ones. A vendor is engaged, and the vendor sells the same product in every market because that is the only way the economics work. A university builds an administrative office modeled on an American one, because the American one is the prestige template and its job descriptions are available to copy. A platform writes moderation rules in one jurisdiction and applies them in a hundred and ninety.

Notice what has happened by the time any of this reaches a person. The idea did not arrive as a claim to be evaluated. It arrived as a mandatory field in a system, a required module before the quarter closes, a policy in the handbook with an effective date. It has already won, procedurally, without ever having been argued in the country where it landed.

That mechanism explains most of what people find infuriating about the phenomenon abroad, and it does so without anyone having to agree about the merits. Even someone entirely sympathetic to the underlying claim is entitled to notice that it was not put to them. Compliance is not persuasion. It is the substitution of an org chart for a debate, and it produces the particular resentment reserved for things that are simultaneously moralizing and non-negotiable.

None of which is new, and the priority is embarrassing enough to be worth stating plainly. In 1970 a French Marxist published an essay in La Pensée called Ideology and Ideological State Apparatuses, and the section above is largely his.

Louis Althusser's distinction was between the repressive apparatus and the ideological ones. The first is the state's capacity for force: government, courts, police, army, unified, coercive, holding ideology in reserve behind the violence. The second set runs the other way round, ideology first and the coercion concealed and symbolic, and he listed them: religious, educational, family, legal, political, communications, cultural. His central and genuinely strange claim was that ideology does not live in anybody's head. It has a material existence. It exists in practices, in rituals, in the apparatus itself. Not what the staff in the regional office believe about the module. The module.

The mechanism he proposed for how it takes hold is interpellation, and the image is a policeman in the street shouting hey, you there. The person turns around, and in the turning becomes a subject: somebody addressed, somebody the hail was meant for. The turn is the entire operation. Belief is not required and was never requested. Only recognition is required.

Which accounts for the detail in Kuala Lumpur that otherwise looks like the apparatus failing. Everyone is bored. Everyone completes it. Boredom is not resistance, and completion is not assent, and the system asked for neither. The employee clicking acknowledge while privately mocking the slides has turned around. What the module wanted was the timestamp, in a database, retrievable by counsel in a jurisdiction eleven time zones away, and it got exactly that from the skeptic and the believer alike, which is why nobody administering it has ever cared which is which.

Now the amendment, because the essay has aged in an interesting direction. Althusser held that the school had supplanted the church as the dominant ideological apparatus. On current evidence the employer has supplanted the school: an adult spends more waking hours inside it, it grades continuously rather than annually, and its sanctions are material in a way a bad mark never was. But the deeper amendment is to the middle word of his term. Every apparatus on his list is national. Church, school, family, courts, parties, press: each of them bounded by a state, staffed by its nationals, arguing in its language. He called them state apparatuses because in 1970 an apparatus and a state were the same size, and that had been true for long enough to look like a law of nature.

It is not true now. The dominant apparatus has no state. It has a headquarters, a general counsel, a vendor contract and operations in a hundred and ninety jurisdictions, and it can install a practice in Malaysia without the Malaysian state holding any view on the matter, or being asked, or in most cases finding out. Strike the middle word. What is left is an ideological apparatus of a kind that did not previously exist, the first to reproduce a set of relations across social formations rather than within one.

That difference is the whole source of the resentment described above, and it is worth being precise about why. Being formed by the apparatus of your own society is invisible from the inside. It feels like weather, or like common sense, or like nothing at all, which is the condition Althusser was trying to make his readers notice. Being formed by somebody else's apparatus is not invisible in the least. It arrives with a foreign case study attached, and it feels like an instruction, because that is precisely what it is. The imported version fails at the one thing the domestic version does well, which is to disappear.

Two cautions, and the second is the serious one. He was a Marxist working out why the revolution had not arrived, and nothing here is that argument. More importantly, the framework as he left it has a defect that its admirers rarely concede: if everything is ideology and the hail always lands, then no observation counts against the theory and the analysis ends in a shrug. Ideology has no history, he wrote, and is eternal. That is the sort of claim which cannot be wrong, and cannot being wrong is a serious flaw rather than a strength. The mechanism is worth keeping. The fatalism is not, and there is a cheap way to test the difference, which is the subject of a later section.

Meanwhile there are real disputes in these countries, and they run on domestic material that the imported framework describes badly.

France is the sharpest case, because it holds a rival universalism of its own. Republican citizenship and laicite are not a hedge against identity politics, they are an explicit alternative to it, and the state does not collect ethnic statistics, which makes American racial categories close to unusable as an administrative matter. The National Assembly voted to strike the word race from the first article of the constitution in 2018, though the revision was never completed. Ministers have attacked what they call wokisme as an American import threatening the republican model, and inclusive writing has been restricted in official documents and in schools. Whatever one thinks of that, it is not a country failing to have the conversation. It is a country that had a different conversation first and reached a different answer.

Germany runs into its own memory culture. Holocaust remembrance is institutionalized to a degree with no American parallel, and imported anti-colonial frameworks have collided with it directly and publicly, in the controversy around a Cameroonian theorist in 2020 and in the antisemitism scandal that engulfed a major art exhibition in 2022. The result is not a German version of the American argument. It is a genuinely different argument in which two moral vocabularies with strong local claims are in conflict.

Canada and Australia organize around settler and Indigenous relations rather than the American racial binary, and the Australian referendum on an Indigenous Voice to Parliament failed in 2023 by roughly sixty to forty, carrying no state. India has run identity politics of enormous consequence since well before America exported anything, structured around caste and religion, with reservation policy older than most of the vocabulary now being imported onto it. South Korea's central conflict is gender, conducted in entirely domestic slang, and it has swung national elections. Latin America generated mass feminist movements on its own, then generated its own noisy reaction against them. Chinese internet users coined their own epithet for Western progressives, white left, with no assistance from anybody.

The pattern is consistent. Where the conflict is real it is local, and where the vocabulary is American the vocabulary is usually the least interesting part of it.

One structural fact accounts for more cross-country variation than any amount of cultural analysis, and it is almost always omitted.

The United States has a constitutional prohibition on regulating speech. Germany, France and the United Kingdom do not; they have functioning hate speech statutes, press regulation and public order law. So the same impulse to police expression that runs through a statute in Europe has nowhere official to go in America, and it flows instead into employment, platform policy, professional association rules and social sanction.

American cancel culture is therefore not evidence of a uniquely censorious population. It is what censoriousness looks like in a system that has closed the legislative route. The European version is calmer in the same way that a regulated utility is calmer than an unregulated market, and just as final. A German prosecuted for a post and an American fired for one have both been sanctioned for speech. Only one of them can appeal to a court, which is a point in Europe's favor, and only one of them is subject to a rule enacted by a legislature that can be voted out, which is also a point in Europe's favor. The American arrangement gets the worst of both: no due process and no democratic authorship.

The tidy story in which an American ideology spreads outward while local traditions resist is wrong in one important respect. The counter-movement is at least as transnational and at least as professionalized, which follows directly from the preceding section: an apparatus is indifferent to its payload, and anyone who enjoyed the analysis while it was aimed at a compliance department should expect it to work identically pointed the other way.

An Argentine president denounces wokismo from a stage in Davos in the vocabulary of an American think tank. A Hungarian government builds an entire electoral identity around gender ideology as a foreign imposition. Russia moves from a law against propaganda to designating an international movement as extremist, and markets the whole package abroad as evidence of Western decadence. A Ugandan legislature passes severe legislation and frames rights advocacy as colonialism in modern dress, which is rhetorically effective precisely because the earlier export really was clumsy and really did arrive as a condition attached to aid.

That last point deserves more discomfort than it usually receives from either side. The manner of the export handed genuinely repressive governments a defensible-sounding argument, and the people who pay for that are local, vulnerable and now less protected than before. Delivering universal claims as compliance does not merely fail to persuade. It discredits the claims themselves, and it does so most severely in the places with the least margin for error.

Which leaves the question worth actually answering: how much of this is universal?

There is a thin version that generalizes without difficulty, and it is thinner than its advocates like and more substantial than its opponents admit. Do not harass people at work. Hire and promote on the merits, and be able to show your reasoning. Apply rules to everyone. Give someone a hearing before you ruin them. None of that requires an American vocabulary, most of it predates the vocabulary by centuries, and no country needs an imported module to understand it. It is ordinary liberal proceduralism and it survives translation because it is a claim about how institutions should operate rather than about how history should be narrated.

Then there is the thick version, which is culturally specific and travels badly: a particular racial taxonomy, a particular national guilt narrative, a particular theory of how identity determines standing in an argument, and a particular set of speech conventions produced by American institutions in a specific decade. Transplanted, these generate category errors rather than progress. Applying American racial categories in a country that legally refuses to collect them is not a bold reform; it is an administrative impossibility. Importing a settler-colonial framework into a society whose central historical wound is a genocide it perpetrated in Europe produces a collision, not a synthesis.

The honest position is that the thin version should be universal because it already nearly is, and the thick version should stay home, not because it is wicked but because it is parochial, and parochial things become ridiculous in transit.

The instrument, for anyone running a multinational, is three questions about any people policy arriving from headquarters.

Was this argued in the country where it will apply, in the local language, by people who live with the consequences, or was it translated? Who is the customer for it, and if the honest answer is the parent company's general counsel, say so in the memo rather than presenting it as a local initiative. And the decisive one: if headquarters stopped requiring this tomorrow, would the local organization keep any part of it?

That last question separates a genuine commitment from an act of compliance more cleanly than any survey, and the answer is usually known to everyone in the building before the question is asked. A practice that would not survive the removal of the mandate was never adopted. It was administered.

It is also the test the theory of apparatuses never supplied. An account in which everyone is always already a subject has no control group, and so it can name the apparatus and cannot measure it. A mandate is better than a theory in exactly one respect: it can be switched off. Switch it off in one region for two quarters, tell nobody it is an experiment, and count what remains. Whatever survives the withdrawal of the hail was a belief. Everything else was a timestamp.

The pattern will not stop at national borders, because the mechanism has nothing to do with nations. It is a property of organizations that operate in more than one place and prefer one policy to several.

The first permanent settlement off Earth will be operated by an entity with a headquarters, a general counsel and a global standard. Its personnel policies will be written to satisfy an authority on a different planet, they will arrive in the payload alongside the water reclamation hardware, and they will be adjusted at a cadence set by legal exposure in a jurisdiction that is eight light minutes away and has never seen the place. Nobody on Mars will have been asked. The module will still be mandatory, and it will still contain a case study about a hiring panel.

The legal substrate for this already exists and is more accommodating than people assume. The Outer Space Treaty forbids national appropriation of a celestial body, so nobody will own the ground. But its eighth article provides that the state on whose registry an object is carried retains jurisdiction and control over that object and over any personnel thereof, and a habitat is an object. The ground is unclaimable and the building is not, which means the settlement will be governed the way a ship is governed, by the flag it is registered under rather than by the water it is sitting in. Jurisdiction arrives as cargo. That is not a loophole; it is the only workable answer anybody has found to the question of which law applies where no law was in force, and it has the incidental property of guaranteeing that every apparatus described in this essay lands intact, with a warrant, at the destination.

A private version has already been drafted, which is the part that should hold the attention. The Starlink terms of service, a consumer broadband contract, have for years carried a clause providing that for services on Mars, or in transit to it, the parties recognise Mars as a free planet, that no Earth-based government has authority or sovereignty over Martian activities, and that disputes will be settled through self-governing principles established in good faith at the time of settlement. Read that as what it is. It is not a constitution. It is a choice-of-law clause, sitting in the same instrument that governs a billing dispute about a dish on a roof in Nebraska, and it does not create a government so much as decline to name one. Self-governing principles to be established in good faith is a promissory note, unsigned, drawn on a future in which the promisor will also be the employer, the landlord, the utility, the medical provider and the atmosphere. The document that disclaims sovereignty over the settlement is a document the settlers will have accepted by clicking through it.

Now the light lag, because it is the detail that makes the mechanism worse rather than better. Mars is between three and twenty-two minutes from Earth at the speed of light depending on where the two planets happen to be standing, so a question and its answer cannot occupy the same hour. This looks at first like a constraint on central control, and for anything requiring judgment it is: nobody in Hawthorne can help with a seal failure, and everybody knows it. But recall what the apparatus actually wanted. Not a conversation, not agreement, not even attention. A timestamp, in a database, retrievable by counsel. Interpellation has no return leg. The policeman shouts, the subject turns, and the turning is the whole of the operation, which means it survives a delay of twenty-two minutes exactly as well as a delay of none, because nobody was waiting for the reply. Every part of the relationship that requires a round trip will devolve to the settlement within a decade. The one part that does not require a round trip is the part that will not.

So the division of authority is predictable, and it is close to the opposite of what a settler would choose. Discretion will be delegated wherever delay is lethal: the abort call, the surgery, the decision to vent a module. It will not be delegated over policy, because policy is the cheapest thing in the payload and the most expensive thing to be caught without. Training content weighs nothing, ships at light speed, costs a vendor licence, and its absence is the first fact a plaintiff's counsel will establish. Under those economics no general counsel will ever propose that the module stay home. The settlement will therefore be trusted with the hull and not with the handbook, and the people living inside the first structure in history where the atmosphere is a management decision will find that the thing headquarters retained most tightly was the paperwork.

Which is where the test proposed above stops working, and it is worth conceding that plainly rather than being caught with it. Switch the mandate off in one region for two quarters and count what survives requires a second region. Mars will be one settlement, of a few hundred people, under one operator, for a long time. There is no control group, no comparison population and no possibility of a quiet natural experiment, because everyone would notice, and because the operator would be running the experiment on the only sample that exists. The instrument that separates a belief from a timestamp is unavailable at precisely the site where the distinction matters most, and it will stay unavailable until there is a second settlement with a different flag on the registry, which is an argument for competition in a place where competition sounds like a luxury.

The thin version and the thick version separate more sharply out there, not less. The thin version becomes more necessary than it has ever been on Earth: apply the rules to everyone, give a person a hearing before you ruin them, be able to show your reasoning. In a workplace with no courts, no legislature, no press and no exit, ordinary proceduralism is not a nicety imported from a law school, it is the entire stock of protection available to a human being, and it should be written down early by people who know they are writing a constitution and not a policy. The thick version becomes farce. A taxonomy built from the specific history of one continent of one planet, delivered to a room where nobody has seen either, in a case study about a hiring panel that met in a city eight months' travel away, is parochial in the exact sense this essay has been using: not wicked, just from somewhere, and visibly so once the somewhere is far enough away to see whole. The demographic field will still be on the form. It will still carry a jurisdiction in its label. It will just have run out of jurisdictions to name.

That is not a prediction about ideology. It is a prediction about org charts, which are the only thing in this entire argument with a demonstrated capacity to cross a vacuum.