What Provenance Does Not Record: The Unsettled Consignment Inside Legitimate Ownership
Every framework the art market uses to establish trust in an object is silent on whether the person who made it was ever paid for it.
A collector buys a painting from a gallery. The gallery holds it on consignment from the artist. The sale completes, the collector takes the work home, and the gallery does not remit.
Two things are now true, and neither cancels the other.
The collector's title is clean. They bought in good faith from a dealer in possession, and no court is going to take the painting off their wall. And the artist's claim against the proceeds is live and unsatisfied, because the money was never theirs to keep.
So the object hangs in a private house carrying an obligation its holder cannot see, has not caused, and has no available method of discovering.
This study is not about any particular gallery or any particular artist. It is about the fact that the art market's central instrument of trust, the provenance record, is constructed so that this failure leaves no mark on it.
The law is stronger than most people assume, and that is what makes the gap strange
There is a widespread belief that an artist whose dealer fails to pay is an unsecured creditor standing in line behind the bank. In most of the United States that is not the legal position, and it has not been for decades.
New York's Arts and Cultural Affairs Law states that when an artist consigns a work to an art merchant, "such work is trust property in the hands of the consignee for the benefit of the consignor; any proceeds from the sale of such work are trust funds." The statute then closes the door that matters most: "no such trust property or trust funds shall become the property of the consignee or be subject or subordinate to any claims, liens or security interest of any kind or nature whatsoever of the consignee's creditors."
California reaches the same destination by its own route. Under its Consignment of Fine Art provisions, the work is property held in trust by the consignee for the benefit of the consignor and is not subject to claim by a creditor of the consignee, and the proceeds of a sale are funds held in trust which must first be applied to pay any balance due to the artist.
Most states have some version of this. The category is worth stating precisely, because the differences decide outcomes: some are genuine trust statutes of the New York and California kind, and others do considerably less, requiring disclosure or written agreement without ever making the proceeds trust funds. An artist's position depends heavily on which kind of state the gallery is in, and that is not a distinction most artists know they are making when they choose a dealer.
Even the strong statutes admit a qualification worth naming rather than glossing. New York permits waiver of the proceeds clause in limited circumstances, and preserves the artist's first two thousand five hundred dollars in any twelve month period against it. California allows the artist to agree otherwise in writing. So the protection is robust and it is not absolute, and an artist can sign away part of it in a contract they are not in a strong position to negotiate.
The point for this study is what the strength of the law implies. The legislature decided, repeatedly and across jurisdictions, that consignment proceeds are not the dealer's money. The protection exists. What does not exist is any mechanism by which a third party can tell whether it was honoured.
The adjudicated case, and what it establishes
The canonical instance is a matter of criminal record rather than allegation.
In March 2010 the New York dealer Lawrence Salander pleaded guilty to twenty nine felony counts of grand larceny and was sentenced to six to eighteen years. The conduct spanned 1994 to late 2007 and involved more than thirty dealers, investment firms and collectors. He sold works he did not own and kept the proceeds, exchanged works he did not own to satisfy his own debts, and failed to inform or pay consignors when their works were sold. Forged invoices and inventories concealed it. Restitution of more than one hundred and fourteen million dollars was ordered, in the court's own formulation, in the event there is ever money to make those payments.
A second conviction arising from the same gallery is the one that matters most here. The gallery's director of fine art was convicted of a scheme that kept artists' estates in the dark about the sale of more than eighty works while she collected commissions on those sales.
Read that finding structurally rather than as a scandal. More than eighty objects moved through a legitimate commercial channel, into the hands of buyers who did nothing wrong, while the people entitled to the proceeds were not told the sales had happened. Every one of those works acquired a new owner, a new location, and a new line in its ownership history. Not one of them acquired a record of the thing that had gone wrong.
The restitution order is the sharpest detail in the case. A judgment for one hundred and fourteen million dollars exists. The money does not. So the harm was adjudicated, quantified, and left standing, which is the ordinary outcome when a gallery fails rather than an unusual one.
What the standards actually require
The expectation going into this research was that provenance frameworks would be silent on payment. That expectation should not be asserted without checking it, so it was checked.
The International Council of Museums publishes Standards on Accessioning, the document that tells member institutions what to establish before an object enters a collection. It is a short instrument and it is explicit about the categories of question a museum must ask. Provenance appears throughout. Acquisition appears throughout. Ownership and ethics are named directly.
The words "artist", "paid", "payment", "consignment", "compensation", "remuneration" and "price" do not appear in it at all.
That is not a criticism of ICOM, and it is important to be exact about what the finding is and is not. Accessioning standards are built to answer a different question, and they answer it seriously: was this object stolen, looted, unlawfully exported, or misattributed. Those are the failures that historically produced the worst outcomes, and the frameworks were built in response to them.
The same shape holds across the wider apparatus. Ownership history, title, theft and illicit trade, export and import documentation, authenticity. A stolen work is discoverable through a register built for the purpose. An unlawfully exported work is discoverable through customs and permit records. A misattributed work is discoverable through connoisseurship and technical analysis.
An unpaid consignment is discoverable through none of them, because none of them were built to look for it.
What a collector can actually do
The honest answer is close to nothing, and stating it plainly is more useful than pretending otherwise.
A buyer can ask a dealer whether a work is held on consignment, and a dealer intending to pay will say yes as readily as a dealer who does not. A buyer can ask to see a consignment agreement, which establishes the terms of an obligation rather than its discharge. A buyer can search the registers, which will not hold this, because the work is not stolen. A buyer can wait, and learn nothing, because the failure occurs after their money has moved.
The structural reason is simple. The event that would need recording, the remittance from dealer to artist, happens after the sale, between two parties, neither of whom is the buyer, and it produces no artefact that attaches to the object.
The diligence that exists is aimed at whether the seller had the right to sell. The question here is whether the seller discharged what they owed after selling, and the entire apparatus stops at the point of transfer.
Scale is worth a sentence, and a caveat. Gallery insolvencies and consignment disputes are not centrally reported anywhere. There is no register of them, and a number invented for rhetorical effect would be worse than an admission. This study does not know how often this happens, and neither does anyone else, which is itself a consequence of the same silence.
The finding
The provenance record is a document about custody. It answers who held this object and when, and it has been refined for a century to answer that question against increasingly hostile facts.
It is not a document about settlement. It does not record whether the transfers it lists were completed on both sides, and it was never designed to.
For most of the record's history that omission cost nothing, because provenance was doing the job it was built for. The omission becomes visible at exactly the moment a gallery fails, and at that moment it is too late for it to be recorded, because the institution that would have recorded it is the institution that failed.
So an object can be entirely legitimate and simultaneously carry an unmet obligation, and both facts can persist indefinitely without contradiction. The holder is not a wrongdoer. The artist is not without a claim. The record is not falsified. It is simply not asked to say.
This is the condition the Post-Luxury framework describes as Structural Captivity. The obligation is real and it is attached to an object that cannot express it, held by a person who cannot discharge it, owed to a person who cannot locate it. Nothing in the arrangement is deceptive. The captivity is in the design of the instrument.
What would have to change
Recording settlement is not technically difficult. It requires the object's record to carry a statement, at the point of each transfer, that the previous obligation was discharged, attested by the party owed rather than by the party paying.
Nothing in the existing apparatus does this. Objects of Affection Collection's own instrument, the Custodian's Contract, binds obligations to the object rather than to the transaction, which addresses one part of the problem and is named here as the institution's own approach rather than as a general solution. It is one attempt among the few that exist, and it has not been tested at the scale that would prove it.
The more useful observation is not about any particular instrument. It is that the market has spent a century making provenance answer harder and harder questions about custody, and has not yet asked it to answer an easy one about payment. The failure described here is not exotic and it does not require bad faith at scale. It requires one dealer, one bad quarter, and a record that has no field for the thing that went wrong.