London, the early 1700s. A mysterious young man named George Psalmanazar arrives from Formosa — the island now called Taiwan, off the coast of China, and to most Englishmen a place of pure rumour.
What happens next is the part that stops being funny:
Oxford University engaged Psalmanazar to teach the island's language; a few years later he translated the Bible into Formosan, then wrote a book—an immediate best-seller—on Formosa's history and geography. English royalty wined and dined the young man, and everywhere he went he entertained his hosts with wondrous stories of his homeland, and its bizarre customs.1
He was a Frenchman.
After Psalmanazar died, however, his will revealed that he was in fact merely a Frenchman with a rich imagination. Everything he had said about Formosa—its alphabet, its language, its literature, its entire culture—he had invented.2
He confessed in his will. Not before.
Most impostures are thin. This one is not, and the thickness is what made it work.
He did not claim to be Formosan and leave it there. He produced:
That is not a lie. It is a body of work, and it took the labour of a scholar to produce.
Which produces the case's central strangeness: the effort required to fake a language convincingly enough to teach it at Oxford is comparable to the effort of learning a real one. He was not avoiding work. He was doing an enormous amount of work in a direction that could not be checked.
The verification structure is the whole operation and it is unusually clean.
No one in England had been to Formosa. No Formosan speakers, no texts, no travellers, no consular records. The island existed and was known to exist, which is essential — an invented country would have failed instantly — and nothing about it was available.
So the only test anyone could apply was internal consistency, and internal consistency is exactly what a diligent forger can supply without limit. Ask him a question about Formosan grammar and he answers; ask again next week and the answer matches, because he built a system rather than a story.
The scrutiny available was of the kind he could pass, and none of the kind he could not. Greene names the enabling condition precisely: he "had built on the English public's ignorance of the place."2
That is a general property worth extracting: an unfalsifiable claim plus a rigorous internal structure reads exactly like scholarship, because scholarship is largely what internal rigour looks like from outside.
The most consequential move in the story is not one Psalmanazar made.
Oxford engaged him to teach. Once that happened, the question of Formosan authenticity stopped being is this man telling the truth and became is Oxford's appointment sound — and a great many people now had a position in the answer.
Every subsequent verification runs through an institution that has already endorsed him. Students learning the language are producing evidence of the language's existence. The Bible translation is commissioned work, which is to say the arrangement's output is also its credential.
The institution did not merely fail to catch him. It became the thing that made catching him expensive, and it did so through ordinary competent behaviour — appointing an expert, using him, publishing the results.
This is the same structure as Bragadino's Venetian senate voting the funds before the demonstration. The scepticism budget is spent by the institution's own investment, and after that the deceptive act does not need to be good, because nobody is looking at it any more. See Bragadino and the Alchemy of Venice.
Worth being exact about who wanted what, because Greene's placement depends on it.
Psalmanazar's fantasy — being an exotic foreigner — is his own private business. The fantasy he was selling belonged to his audience, and Greene names it:
British culture's rigid control of people's dangerous dreams gave him the perfect opportunity to exploit their fantasy.2
Under the second oppressive reality: fixed social codes, and the fantasy of a world with different ones.
What his hosts bought at those dinners was not information about Formosa. It was an evening in a place where the rules were different — the "bizarre customs" are the product, and the more bizarre the better, because their function is to establish that the codes at home are contingent rather than natural.
Which explains a detail that would otherwise be a mistake: he made Formosa strange rather than plausible. A cautious forger invents a boring island. He invented an interesting one, and the implausibility was the merchandise.
He kept it up for his entire life and told the truth only where it could cost him nothing.
That is a choice with a shape. A deathbed confession is not repentance in any operative sense — it is the disclosure timed to arrive after all the benefits and none of the consequences.
But it also tells you something about the run itself. He was never exposed. There is no unmasking scene, no rival scholar, no returning traveller. A structure sound enough to survive decades of contact with a university was sound enough to survive indefinitely, and the only thing that ended it was the author.
Which is a more disquieting outcome than a fraud getting caught. The system had no mechanism that would have found him.
This one's honest use is defensive, and the defence generalises well past impostors.
Somebody is presenting expertise in an area where you cannot check and nobody around you can either. A market you have never operated in, a technical domain outside your training, a jurisdiction, a language, a period.
Do not test their consistency. Consistency is the thing a well-prepared fabricator has and a genuinely knowledgeable person often lacks — real expertise is full of it depends, unresolved disagreements and things the person has not thought about. Fluent, complete, internally seamless answers to every question are a warning rather than a reassurance.
Instead, do the two things Oxford did not.
Find one external anchor. One person, document or record outside the claimant's own account. Not a reference they supply — one you find. Psalmanazar survived because no such anchor existed for Formosa in 1704; almost nothing today is genuinely in that position, and the reason people skip this step is effort rather than impossibility.
Notice when your own institution's endorsement becomes the evidence. The moment your reason for believing includes we already hired them or we've been using this for two years, the verification has become circular and every further year makes the check more expensive rather than less necessary.
And on your own account: if you are ever tempted by the Psalmanazar route, note that he did the work. The labour of faking it competently was not smaller than the labour of doing it, and it bought him a life of never being able to stop.
Strongest evidence. A named individual with a documented publishing record, a named institution, a specific and extraordinary output (a Bible translation into an invented language), and a confession he authored himself. This is among the best-attested cases in the whole chapter.
Tension — the labour is enormous and the chapter treats him as a fantasist. "Merely a Frenchman with a rich imagination" undersells a decades-long scholarly production by some distance.
Tension — he was never caught. Greene presents the will as a resolution; on the facts it is an author ending his own operation, and the absence of any detection mechanism is the more significant finding.
🚩 [POPULAR SOURCE] · 🚩 SECONDARY WITHOUT PRIMARY — no dates beyond "the early 1700s", no title for the best-selling book (it was An Historical and Geographical Description of Formosa, 1704 — [SPECULATIVE], supplied from outside the source and flagged as such), and no citation of the will. 🚩 The claim that he translated the Bible into Formosan is extraordinary and is given without support.
Open questions. Was the language genuinely consistent, or did nobody look hard enough to find out? Those are very different achievements and the chapter cannot distinguish them. And did anyone doubt him at the time — a case with no recorded scepticism over decades is either a very good forgery or an under-researched account.
Against the other three cases in Law 32's table, this one is the outlier on effort.
Bragadino needed a glass tube. Van Meegeren needed to paint well, which is a real skill but a known one. Hartzell needed a mailing list. Psalmanazar needed to build a language. He is the only figure in the chapter whose fraud required more work than the honest version of the same career.
Which is the finding this page contributes to the corpus-level pattern: the chapter treats all four as instances of "playing to fantasy," and they differ enormously in what the operator has to be capable of. Greene's framing is uniform; the underlying labour ranges from a sleeve trick to a lifetime of scholarship.
Against Law 30's material the connection is direct and Greene does not make it. The Hidden Labor argues that concealed preparation is the asset. Psalmanazar's concealed preparation was immense and its product was nothing — which is precisely the line that page draws between hidden labour and fraud, and this case sits on the wrong side of it while requiring more of the labour than Houdini did.
And against Keep the Extent of Your Abilities Unknown: Gracián says an unfathomed capacity generates veneration. Formosan was unfathomable in the strict sense — there was no bottom to sound — and the veneration lasted a lifetime.
Epistemics — the claim that cannot be checked and the structure that substitutes for checking. Signal Mimicry: The Arms Race Between Scam and Legitimate Branding tracks how the markers of legitimacy get counterfeited faster than institutions can develop new ones.
Psalmanazar counterfeits the marker that is hardest to fake and therefore trusted most: systematic depth. Not credentials, not manner, not references — an entire consistent body of knowledge.
The insight neither produces alone: the arms-race frame assumes the defence is a better signal. This case shows the strongest available signal being the exploitable one, because systematic internal consistency is what genuine expertise produces and it is also what sufficient labour produces. There is no signal that distinguishes them, and adding rigour to the test makes the forger's position stronger rather than weaker, since rigour is what he has and uncertainty is what the honest expert has. The only defence is categorically different: an external anchor, which is not a signal at all but a fact outside the claimant's control.
Craft — the invented world that has to hold together. Worldbuilding as Ascending Questions treats the construction of a fictional culture as a discipline of consequence-following: each invented fact generates further questions, and the world's credibility comes from having answered enough of them that the unanswered ones feel merely unvisited.
Psalmanazar is doing exactly that, professionally, for an audience that does not know it is fiction.
What the pairing produces is the observation that the two practices are technically identical and differ only in the reader's contract. A novelist and Psalmanazar solve the same problems with the same methods; a Formosan alphabet is a worldbuilding artefact whether or not anyone is being deceived. The fraud is not in the making — it is entirely in the frame, in the single claim I am from there, which required none of the skill and carried all of the deception. That is a strikingly small hinge for a lifetime of work to turn on, and it suggests the interesting question about this case is not how he built the world but why he needed it to be true, which the chapter, reading him as an operator, never asks.
Sharpest implication. The only test available to his audience was internal consistency, and internal consistency is exactly what sufficient labour supplies — so rigour made the forgery stronger, not weaker. Fluent, seamless, complete answers to every question are a warning; genuine expertise is full of it depends and unresolved disagreement. The only defence is not a better test but an external anchor, a fact outside the claimant's control — and Oxford's appointment removed even the appetite for one, because after that the institution's own credibility was the thing being verified.
Generative questions.