Vermeer's importance has been recognised for a long time. The problem, for anyone who wants one, is that there are very few and they almost never come up.
Then, in the 1930s, Vermeers start appearing on the art market.
Experts were called on to verify them, and pronounced them real. Possession of these new Vermeers would crown a collector's career. It was like the resurrection of Lazarus: In a strange way, Vermeer had been brought back to life. The past had been changed.1
They were painted by a middle-aged Dutchman named Han van Meegeren.
And he had chosen Vermeer for his scam because he understood fantasy: The paintings would seem real precisely because the public, and the experts as well, so desperately wanted to believe they were.2
Every other case in Law 32 works on a lay audience — senators, courtiers, dinner guests, farmers. This one works on the people whose entire professional function is to detect exactly this.
That inverts the usual reading of expertise as a defence.
An expert authenticator is not a neutral instrument. They are a person whose career is built on judgements about attribution, and for whom the discovery of a new Vermeer is not merely an interesting event — it is the most significant thing that could happen to them professionally. The verdict this is real is, for the person delivering it, a much larger and better outcome than the verdict this is fake.
So the incentive gradient runs toward authentication, and it runs hardest in exactly the people best equipped to detect the forgery.
The scarcer the artist, the more a discovery is worth, and the more it is worth, the more the discoverer needs it to be true. Van Meegeren did not evade expertise. He selected the artist whose scarcity would bend it.
Greene's placement is the fourth oppressive reality — death, and the fantasy of its reversal — and it needs one correction to be useful.
Nobody buying a Vermeer was buying relief from their own mortality.
What was reversed is narrower and stranger: the closure of a finished body of work. Vermeer's output is fixed and has been since 1675. That is a specific, permanent, quietly maddening fact for anyone who loves the paintings, and it is unlike ordinary death because the loss is ongoing — every year produces no new Vermeers, forever.
A new Vermeer does not resurrect a man. It reopens a closed account.
That generalises much further than art, and it is why this case is the most portable of the four. Every unrepeatable thing has the same structure: the last conversation you did not have, the discontinued product, the band that broke up, the era that ended. The fantasy on offer is never immortality. It is that this particular finished thing is not finished.
The chapter's own sentence is the finding: "the paintings would seem real precisely because the public, and the experts as well, so desperately wanted to believe they were."2
Precisely because. Not despite the scrutiny — because of the wanting, operating through the scrutiny.
Take that seriously as a mechanism rather than as a moral and it produces something specific. Authentication is not a binary test. It is a judgement under uncertainty, assembled from many small ambiguous signals — brushwork, craquelure, provenance, pigment, the feel of it — and every one of those admissions of ambiguity is a place where a prior can operate.
An authenticator who wants the painting to be real does not lie or ignore evidence. They resolve each ambiguity in the favourable direction, at a rate slightly above chance, and there are enough ambiguities that the aggregate verdict is determined before the examination begins.
That is the same structure as the Venetian senators failing to see a glass pipe. They were not stupid and they were not bribed. They had already spent their scepticism on the competition for him. See Bragadino and the Alchemy of Venice.
Worth stating plainly because the chapter's framing obscures it: van Meegeren could paint.
Not well enough to be Vermeer — the forgeries are not, on later examination, very convincing, and the judgement of the 1930s experts has not aged well. But well enough to produce work that survived professional examination by the leading authorities of the period, using period-plausible materials and technique.
That is a real skill and it took years to develop. Which puts him in the same category as Psalmanazar and against Bragadino: the fraud required genuine capability, and the capability was not smaller than the honest version of the same career.
[SPECULATIVE] — the standard account, not in Greene: van Meegeren was a working painter whose own original work had been dismissed by critics, and the forgeries were begun partly as a demonstration against them. The source gives none of this and it is flagged as external. But if it is right, the case has a shape Greene's framing cannot hold — a man who could not get his own work taken seriously, producing work under a dead man's name that the same critics called a masterpiece.
Greene stops at the exposure — "only later did it come out" — and gives no mechanism.
[SPECULATIVE], again from outside the source: the exposure came because one of the forgeries had been sold to Hermann Göring, and after the war van Meegeren was charged with selling a Dutch national treasure to the enemy — a capital offence. His defence was to confess to forgery, which carried a far lighter sentence, and to prove it he painted another Vermeer in custody.
If that is right, it is the sharpest ending in the chapter and Greene omits it entirely. The forger's escape was to demonstrate the fraud on demand — the one act that destroyed the fantasy and saved the man.
Recorded as external context and flagged, because the chapter's four-line treatment gives no ending at all and a page that stopped where Greene stops would leave a reader believing it simply "came out."
You are evaluating something you badly want to be true.
Not a con — an opportunity, a candidate, a piece of work, a diagnosis. Something where a yes would be excellent for you.
The van Meegeren structure says your expertise will not protect you here and may work against you, because expertise operates through a long series of small ambiguous judgements and wanting resolves ambiguities without announcing that it has.
So do not try to be more rigorous. Rigour is the channel.
Do two things instead.
Write down what a yes gets you, before you examine anything. Not to talk yourself out of it — to make the incentive visible while you still have distance from it. An authenticator who has written this discovery would be the making of my career on a piece of paper is in a different position from one who has not, and it costs thirty seconds.
Then find someone whose yes is worth nothing to them. Not a sceptic — a person with no stake either way. Their examination need not be better than yours; it needs to be uncorrelated with your wanting. That is the only check that reaches this failure mode, and it is why institutions have external reviewers and why external reviewers stop working the moment they are picked by the person seeking the answer.
And the version worth watching for in yourself: the scarcer and more wonderful the thing, the harder your own judgement is pulling. The feeling of this would be extraordinary is not evidence, and it is doing arithmetic on your behalf.
Strongest evidence. A named forger, a named artist, a specified decade, and — unusually — a stated mechanism rather than just an outcome: the wanting operating through the experts. The case is real and well documented outside the source.
Tension — the fourth pairing is about irreversibility, not death. Documented above; the collectors' mortality was untouched.
Tension — no ending is given. "Only later did it come out." The actual ending, if the standard account is right, is the most instructive part of the case and is omitted.
Tension — the chapter treats him as a pure operator. The possibility that the forgeries were a response to critical rejection is not raised, and it would change what the case demonstrates.
🚩 [POPULAR SOURCE] · 🚩 SINGLE SOURCE · 🚩 SECONDARY WITHOUT PRIMARY — no dates within the decade, no named expert, no named painting, no source. Two [SPECULATIVE] passages above are flagged as coming from standard accounts outside Greene's text, not from the source.
Open questions. How good were the forgeries, actually? The claim "experts pronounced them real" and the later consensus that they are unconvincing are both true, and holding both is the whole lesson — the paintings did not change. And: how many others were never caught? A method that turns on the authenticator's incentive has no reason to produce detectable failures.
Against the other three cases in the table, this is the one where the target is a professional.
Bragadino's senate, Psalmanazar's Oxford, Hartzell's Drakes — all are audiences operating outside their competence. Van Meegeren's authenticators were operating at the exact centre of theirs, and that is why the case is worth more than its four lines. Expertise is not a defence against a fantasy that pays the expert, and Greene has the mechanism in one sentence and does not generalise it.
Against Law 30, the pairing is instructive on the build's running distinction. The Hidden Labor separates concealed preparation that has a real product behind it from concealment that hides an absence. Van Meegeren sits ambiguously between them: the paintings physically exist, they were made with genuine skill, and what is false is not the object but the attribution. That is a third category the chapter has no room for — the work is real and the frame is a lie, which is also exactly Psalmanazar's position.
And against Keep the Extent of Your Abilities Unknown: Vermeer's scarcity is an unfathomed ceiling of a peculiar kind. Nobody knows how many he painted. The uncertainty in the catalogue is the space the forgeries were inserted into, and an artist with a complete documented output could not have been forged this way.
Psychology — the substitution that runs through the examination. The Question You Actually Answered describes the mind swapping an unanswerable question for an answerable one and returning the substitute with full confidence, the swap never surfacing.
The authenticator's hard question is did Vermeer paint this in 1660. The available one is does this feel like a Vermeer to me — and feeling-like is a judgement the expert is genuinely qualified to make and which is contaminated by everything they hope.
The insight neither produces alone: in the standard substitution cases the easy question is a noisy proxy. Here it is a motivated one, and that is worse in a specific way: noise averages out across many judgements and motivation does not. An expert making forty small ambiguous calls with a slight favourable bias does not produce a slightly biased verdict — they produce a confidently wrong one, because the errors accumulate in one direction. Which explains how a forgery that later looks obvious passed the best examination of its day, and it means the defence cannot be more examination. It has to be an examiner with a different payoff.
Craft — the work that could not be seen under its own name. Uccello and Visible Effort documents a judgement authored by one critic and inherited for five centuries — Vasari decided Uccello's paintings were laboured, and they have been laboured ever since.
Put beside van Meegeren the pair says something neither says alone about how attribution does the seeing.
Vasari's verdict made real paintings look worse. The Vermeer attribution made forged paintings look better. In both cases the object is fixed and the perceived quality moves, and it moves in response to a claim about who made it and what that means. The same canvas is a masterpiece under one name and a curiosity under another, and the 1930s experts were not incompetent — they were seeing what the attribution instructed them to see, which is what all of us do all of the time.
That produces the most uncomfortable question available here, and it is a craft question rather than a strategic one: if van Meegeren's paintings were worth admiring while everyone believed Vermeer made them, what exactly stopped being true about them on the day he confessed?
Sharpest implication. Expertise is not a defence against a fantasy that pays the expert. Authentication is a long chain of small ambiguous judgements, and wanting does not make an examiner dishonest — it resolves each ambiguity favourably at a rate slightly above chance, and the errors accumulate in one direction rather than averaging out. That is how a forgery that later looks obvious passed the best examination of its day, and it means the correction cannot be more scrutiny. It has to be a scrutineer whose yes is worth nothing to them.
Generative questions.