Tenure Without Persons

In 1990 Roger Kimball published Tenured Radicals: How Politics Has Corrupted Our Higher Education. The book argued that the American humanities had been captured by people whose political commitments came before their scholarship. It is a polemic. It was meant to wound, and it did.

The U.S. Department of Education catalogued it. The record is ED328110 in ERIC, the federal education research database. Publication type: Books; Opinion Papers. Two hundred and four pages. Harper and Row, ten East Fifty-third Street, New York, eighteen dollars and ninety-five cents. Full text not available. Descriptors assigned: College Faculty. Curriculum. Curriculum Development. Educational Trends. Feminism. Higher Education. Humanities Instruction. Ideology. Liberal Arts. Marxism. Moral Values. North American Culture. Political Influences. Political Issues. Politics of Education. Role of Education. Technical Writing. Trend Analysis. Western Civilization.

Before the irony, the mechanism. The phrase everyone remembers has been misread for thirty-six years, both by the people who loved it and by the people who hated it.

“Tenured radicals” does not describe an ideology. It describes a change of state, like water freezing. The adjective does almost all the work. What alarmed Kimball was never the content of the beliefs, which had circulated harmlessly in the same buildings for two decades. It was that the beliefs had gained a permanent claim on the institution’s resources, through a procedure that nobody who disagreed could reverse.

Tenure is a ratchet, a device that turns one way and cannot be turned back. The American version was first argued for in the 1915 Declaration of Principles of the newly founded American Association of University Professors (AAUP) and formalized in the 1940 Statement of Principles, issued by the AAUP with the Association of American Colleges. The 1915 Declaration described the university as an intellectual experiment station where new ideas could germinate and their fruit be allowed to ripen until, perchance, it became part of the accepted intellectual food of the nation. The threat it was built against was the trustee: the wealthy donor or political appointee who would fire a professor for reaching an unwelcome conclusion. It worked. It is one of the more successful institutional inventions of the twentieth century, and it worked exactly as specified.

That is the problem, and the whole problem. A device that protects a scholar from dismissal for their views protects every scholar from dismissal for their views. It cannot be built any other way, because a version that discriminated on content would be the very thing tenure exists to prevent. So the ratchet turns for whoever is holding it when it turns. It has no politics. It has a direction.

The best demonstration of this was written by someone who very much did not want it to be true and published it anyway. That is to his credit, and it is why the document is worth reading now.

On November 21, 1991, Michael P. McDonald, president and general counsel of the Center for Individual Rights (CIR), spoke at the Heritage Foundation. The talk was published the following year as Heritage Lecture 371, “Defending Academic Freedom.” CIR had just set up an Academic Freedom Defense Fund, offering free legal help to students and professors on the wrong end of what was then universally called political correctness. McDonald summed up their approach in five words borrowed from the public interest litigator John Banzhaf: “sue the bastards.”

They had been winning.

• A Georgetown law student wrote in the student paper that the school applied a double racial standard in admissions. He was charged with breach of confidentiality, settled on favorable terms, and graduated.
• A fraternity at George Mason was punished for an offensive charity contest. It won on First Amendment grounds before Judge Claude Hilton.
• A professor at Delaware had been blocked by her university from accessing an outside foundation’s research money. She got the ban voided in arbitration.
• At the City College of New York, the philosopher Michael Levin had written off campus about race and intelligence. The college responded by convening a committee on whether his views amounted to conduct unbecoming a scholar, and by running shadow sections of his classes so students could avoid him. Levin won before Judge Kenneth Conboy, who found the college had punished him solely for his expressed ideas and ordered it to stop harassing him.

Then McDonald told the room what happened next, in a section he headed with a warning that victories may come back to haunt you and be used against you.

At the same college, in the same period, Leonard Jeffries chaired the department of African-American studies. He was under mounting pressure over a speech in Albany that summer and over material he had long distributed in class, and his reappointment as chairman was thought to be in danger. City College officials, McDonald reports, literally waved a copy of the Levin decision in the faces of those demanding action, and said their hands were tied. Jeffries was reconfirmed, for a shortened term.

Consider the shape of that. A conservative public interest firm builds a tool to protect a professor from an administration that wanted him gone for his opinions. Within weeks the same administration picks up the tool and uses it to protect the professor those litigants regarded as the worst actor on that campus. It does so while claiming, with what McDonald called appalling but predictable disingenuousness, that it had no choice. Both professors sued the same college president. The two case names differ only in the plaintiff.

McDonald did not pretend this away, and his conclusion is the most durable thing in the lecture. Judicial protection of the principle of academic freedom, he wrote, does very little to improve the quality of what is being done in the name of academic freedom. That is a man watching his own tool work perfectly and produce a result he hated, and understanding why. The tool did not care what it protected. It had to be that way. That is what made it a tool rather than a faction.

Now back to the catalogue record. It is the same finding arriving by a quieter road, and it is funnier.

Kimball’s charge was that the education establishment had been captured. The establishment’s response, at the level of the federal database, was to file him. It did not refute him, suppress him or answer him at all. It gave him a number, a publication type and nineteen controlled-vocabulary terms (one of them Marxism), and noted the retail price.

It is tempting to read this as suppression by burial. I think that reading is wrong, and the truth is less flattering to everyone. Nobody buried it. The record is exactly as accessible as every other record, which in a database of well over a million items is practically the same as being hidden, and that takes no intent at all. This is not censorship, and it is not tolerance either. It is processing. An apparatus does not need to refute a criticism. It only needs to index it.

Notice what the descriptors accomplish, without any malice, applied by a cataloguer following a manual. Ideology is a term in the thesaurus. So are Marxism, Feminism, Moral Values and Western Civilization. Every charged word in the book has a neutral slot waiting for it, and once the words are in their slots the book is no longer an accusation. It is a document about certain topics, sitting beside other documents about the same topics. The polemic went in and a record came out. That transformation is the institution’s entire immune response, and it works better than any rebuttal, because a rebuttal would have conceded that there was something to rebut.

The publication type is the detail I cannot stop thinking about: Books; Opinion Papers. It is not wrong. It is exactly correct, which is what makes it devastating. A man writes two hundred and four pages to say that scholarship has been displaced by opinion, and is filed, accurately, under opinion.

All of this would be a period piece about universities, except that the interesting institution is now the other one, and it solved the same problem in the opposite direction.

Corporate America has no tenure. Nearly everyone in an American company is employed at will, meaning they can be dismissed on a Tuesday for a reason nobody has to state. By the logic so far, that should make a firm the least capturable institution in the country: no ratchet, nothing irrevocable, no protected cadre, and every belief held only as long as its holder holds a badge.

The opposite happened. Firms found a way to grant tenure without granting it to anybody. A belief written into a mandatory system gets exactly what tenure confers: a permanent claim on the institution’s resources and attention, immune to the departure of the person who introduced it, defended by procedure rather than by argument. Consider the required training module, the form field that cannot be left blank, the metric on the dashboard, the clause in the vendor contract, the policy with an effective date and no expiry. None of these can be fired. None will be reviewed, because a process only gets reviewed when it hurts, and, as I have argued elsewhere, automating a process removes exactly the pain that would have led someone to remove it.

So ask directly: who holds tenure inside a large firm? Not the chief executive, whose median time in the job is a handful of years and who can be fired for two bad quarters. Not the staff, who are at will by definition. Not even the founders, who get diluted and then managed out. The systems hold it. The tenured radical of corporate America is a required field in a form, and it has better job security than anyone reading this.

This is a better arrangement for the institution and a considerably worse one for every person in it, and the asymmetry is exact. In the university, the radical had to win tenure. That meant surviving a review, being named, and being identifiable ever afterward as the person whose commitments those were. Whatever else that system did, it kept belief attached to a human being who could be argued with, mocked, or outlived. The corporate version detaches the two. The belief becomes permanent and the believer stays disposable. Nobody has to defend the module, because its author has left and the module remains.

That also explains why the corporate fight feels sour and unsatisfying in a way the campus fight, whatever its excesses, did not. You could name Jeffries. You could name Levin. Both stood up and said what they thought under their own names, and were sued and sued back. There is nobody to name in a compliance workflow. There is only an “owner” field, and it usually points at a distribution list.

A short note, because it connects to an argument I have made at length elsewhere (“The Org Chart Travels”).

Tenure itself traveled. So did the American speech code, the American administrative office with its American job description, and the American mandatory module. None of them crossed a border as a proposition for the people who would live under them to evaluate. They crossed as templates, through procurement, because a single global standard is cheaper to administer than forty local ones. What arrives in a country is never the argument. It is the apparatus, and the apparatus is portable in a way an argument never has been.

This is the same property that made the Levin decision usable by the people it was won against. A tool that works regardless of who holds it will eventually be held by someone you did not expect, and the more neutral and well built it is, the more certain that becomes. Neutrality is not a safeguard against capture. It is the precondition for transfer.

So: will there be radicals elsewhere who win tenure? On the Moon, on Mars, on whatever comes after?

Yes, and quickly. They will not look like radicals, and there will be no tenure for them to win.

The setting. Start with what an early settlement actually is. It is a workplace, operated by a corporation, staffed entirely by at-will employees. It has no legislature, no courts of its own, no First Amendment, no faculty senate and no AAUP. Whatever protections exist arrive as company policy, which the company may amend. The dissenting engineer on Mars, the one who thinks the seal specification is wrong or the medical protocol unethical or the rotation schedule a slow injury, has no Levin decision to wave at anybody. This will be the least protected working population in the history of the species. It will also be made up of extraordinarily capable people who each cost a hundred million dollars to deliver. That combination has never existed before, and nobody has thought through its consequences.

What becomes permanent. Something there will become irrevocable, because something always does, and the raw material is in plain sight. On a world where the atmosphere is inside the building, every rule can be justified as life support. “This is safety-critical” is the most powerful conversation-ending claim ever devised. It is more powerful than “compliance requires it” or “the board has decided,” because the person overruling it is proposing to gamble with a hull. Some of those rules will genuinely hold up the structure. Others will be a preference that arrived on the first ship in a document written by someone in Hawthorne, California who is now retired. After ten years nobody will be able to tell which is which, and asking will carry a cost that asking about a training module never did.

That is tenure. Not for a person, but for a paragraph. The first permanent institution off Earth will not be a parliament, a university or a church. It will be a checklist. It will outlast everyone who wrote it, and it will be defended by people who have no idea why each item is there and every reason not to be the one who removed it.

The paragraph does not originate on Mars. It is written on Earth, to satisfy an authority on Earth, in a payload assembled by an organization with a headquarters and a general counsel. It is mandatory on arrival for reasons nobody at the destination was consulted about.

The case for the checklist. I should concede how good the checklist is, because the argument is worthless if it reads as a complaint about safety procedure. The device has an excellent founding story. On October 30, 1935, at Wright Field, Boeing’s Model 299 took off, stalled, crashed and burned, and two men died. The aircraft was not defective. It was the most capable bomber anyone had built, and it had one more system than a man could hold in his head. Someone had left the gust locks engaged. The Army could have decided the design was too much aeroplane. Instead a group of test pilots wrote down what to touch and in what order, on an index card. The service-test fleet that followed, flown under the new card, logged nearly two million miles without a serious accident, and the design became the B-17. Everything defensible about procedure is in that card.

The strongest case for making a paragraph permanent is the same move at its maximum. The submarine Thresher was lost with all hands in April 1963, with one hundred and twenty-nine dead. Within weeks the Navy set up SUBSAFE, a certification program covering everything exposed to sea pressure, with documentation requirements that strike outsiders as pathological. No SUBSAFE-certified submarine has been lost since. The one boat the fleet lost afterward had not been through the program. That is sixty years of evidence. It is no rhetorical concession to say that some procedures should outlive everyone who wrote them. It is the reason the ratchet exists, and the reason nobody sensible proposes to dismantle it.

Why items stay regardless of merit. But notice what actually holds those items in place. It is not their merit. It is an arithmetic available to every individual in the building. Removing a procedure has a limited upside, spread thinly across everybody, and an unlimited downside owned entirely by the person who removed it. Nobody is ever promoted for deleting an item, and somebody is certainly ruined by deleting the wrong one. For every individual the expected value of deleting is negative, and for the organization it is positive. That structure needs no defenders, no ideology and no conspiracy. The rule survives because the cost of killing it is concentrated and the benefit is diffuse. A ratchet is just an asymmetry that nobody owns the far side of. This is why merit stops predicting survival. The good item and the stupid item are held in place by the same mechanism, and once their authors are gone the two are indistinguishable from inside.

Why it is worse off Earth. Off Earth, that asymmetry stops being procedural and becomes physical, and this part has no terrestrial equivalent. The checklist will not be a card or a wiki page. It will be software, and much of it will be interlocked: the hatch does not cycle until the sequence is acknowledged, the printer will not release the part until the traveler document is signed, the suit will not report ready until the item is closed. Removing a paragraph then means a code change, a review board, a regression test, and a signature from someone who will be asked, if anything goes wrong, why they touched it. On Earth an outdated rule wastes an afternoon. In a pressure vessel it is a door that will not open. The difference between a rule and a wall has collapsed into a configuration file that a contractor in Hawthorne wrote before the crew was selected.

The rehearsals. The rehearsals for all this already exist, and they are not reassuring.

• The International Space Station is governed by an intergovernmental agreement whose twenty-second article settles criminal jurisdiction by nationality. Your passport, not your location, decides whose law applies to you, aboard a structure no single nation owns. That is jurisdiction as cargo in its purest form. The crew code of conduct requires obedience to the commander.
• Closer to the ground, Antarctic winter-over is a company town run by contractors under a federal agency. It has no local court, no press, and no exit for months, and the workforce is expensive to deliver and easy to replace at the end of the season. A report commissioned by the National Science Foundation and published in 2022 found harassment and assault in the program. More to the point here, it found that the people affected did not believe the reporting channels were for them.

That is the best-supervised remote workplace humans currently operate, with a functioning legal system one flight away and an agency with a reputation to protect. Mars gets none of those three.

The missing amendment clause. So what the first settlement will lack is not protection, which everybody discusses, but an amendment clause, which nobody does. Every durable institution on Earth has one, and it is always the least glamorous document in the founding: the procedure for undoing things, the majority required, the person whose job it is to ask whether an item is still needed. Amendment clauses are written by people who expect to be wrong. First crews are selected, screened and celebrated for the opposite quality. The eighteen months in which the procedures are written will be the months when being wrong is least survivable and least admitted. The paragraph that becomes permanent will not be the one anybody argued for. It will be the one that arrived with no owner, and so with nobody empowered to retire it.

The one cheap fix. That suggests the one intervention available, and it is cheap enough to be embarrassing. Make every item carry its provenance: the incident, the date, and the name of the person or body that put it there, in a field attached to the procedure, visible to the person carrying it out at three in the morning. An item that can name its fatality holds up the structure and should be defended by anyone who touches it. An item that traces back to a preference held by someone in Hawthorne who has since retired is a preference, and it should expire on a cycle unless somebody currently alive is willing to re-approve it under their own name. This is the only form in which the usual test survives on a world with no control group. You cannot switch a rule off for two quarters and count what remains when the sample is the whole population present. But you can, at zero cost, require the rule to say why it exists.

Who the radicals will be. This returns the essay to where it started, a catalogue record. The polemic will be forgotten and the accession number will not, because the number is the part designed to persist. On Mars the accession number will be a document control identifier on a procedure. It will still be resolving long after everyone named in the metadata is dead, and the only question that matters to the person reading it will be whether the record says what the item is for. Nobody thinks of the numbering scheme as the institution. It always is.

The radicals, then, will be whoever happened to be present when the settlement decided what it would not revisit. They will not have argued for permanence and may not have wanted it. They will simply have been in the room, holding the pen, during the eighteen months when the procedures were written and everybody was too busy staying alive to argue over the wording. That is how it has always worked. Nobody wins tenure by arguing for tenure. They win it by being early.

The instinct at this point is to ask how to keep the wrong ideas out. That instinct should be abandoned, because the record is clear that it cannot be done.

Kimball’s side could not do it. They were sophisticated, well advised, and correct about the specific abuses they litigated. Their own tool turned in their hands within a single semester, in the same building, against the same defendant. If a firm of lawyers who spent their careers thinking about exactly this problem could not build a ratchet that only turned their way, then a policy committee’s confidence that it can screen for the right permanent beliefs is not a plan. It is a mood.

The question that can be answered is simpler: which of your rules is revocable, by whom, on what evidence, and how would you find out? Any institution should be able to produce, on request, a list of the things it cannot undo, with the name of whoever could undo each one if it could be undone. Almost none can. The list exists in every case and is longer than management believes. It gets discovered the way City College discovered it: late, in public, when somebody waves a copy of it in your face.

So write the list. Do it on a campus, in a firm, and above all in the first eighteen months of a settlement, when everything still looks provisional and nothing is. Mark each item with what would have to be true for it to be reversed. If the honest answer for an item is “nothing,” then that item is now part of your constitution, and you should say so out loud. That is what a constitution is: not the document that says what you believe, but the document that lists what you have agreed not to reopen.

And keep the record, ED328110, somewhere you can see it. A man wrote a book accusing an institution of being unable to tolerate criticism. The institution catalogued the book, assigned it nineteen descriptors, listed its price, declined to hold the full text, and moved on to the next item. Nobody suppressed anything, and nobody had to. The apparatus was never going to argue with him. It did not need to win, because it was still going to be there in 2026, indexing, when everybody involved in the argument was gone.