How museums decide what to give back

The public argument about returning objects tends to be conducted in moral terms, and the institutional process that actually decides these cases is almost entirely technical. Claims succeed or fail on documentation, jurisdiction and the specific legal instrument invoked, which is why cases that seem obviously similar can produce opposite outcomes.

How contested objects arrived

The collections were assembled through several routes, and the route determines almost everything about a later claim.

Some objects were taken during colonial administration or military action, documented at the time as spoils. Others were purchased in conditions where the seller’s authority to sell is now questioned, sometimes because the seller was a colonial official rather than the community that owned the object. Others were excavated under permits issued by an occupying power. Others were removed from graves and burial sites, which occupies a distinct legal category in several countries. And a great many were bought in good faith on the open market from dealers whose own acquisition is undocumented.

That last category is the most common and the most difficult, because the museum did nothing wrong at the point of purchase and the object may still have been stolen several transactions earlier.

What a claim must establish

A repatriation request is not a petition. It is closer to a title dispute, and the claimant generally has to establish four things.

Identification. That the specific object in the collection is the object claimed. This is harder than it sounds for material catalogued vaguely a century ago, and many claims stall here.

Provenance. The chain of custody from origin to museum. Gaps are normal, and the contested question is who bears the burden when the record is silent.

Standing. That the claimant is the appropriate party to receive the object. This can be a state, a community, a religious institution or descendants, and where more than one asserts standing the museum is placed in the position of adjudicating between claimants, which it is poorly equipped to do.

Legal basis. Which instrument applies, which is where most claims are decided in practice.

The legal instruments involved

Three matter most, and their limits explain a great deal.

The UNESCO 1970 Convention commits signatory states to preventing illicit import and export of cultural property and to facilitating return. Its central limitation is that it generally operates prospectively, applying to transfers after a state adopted it, which excludes most colonial-era acquisitions. It has been enormously influential on museum acquisition policy while resolving relatively few historic cases.

NAGPRA, in the United States, is far more directive. It requires institutions receiving federal funds to inventory human remains and specified categories of cultural items, notify affiliated tribes, and repatriate on request. It works because it imposes an affirmative duty to catalogue and notify rather than waiting for a claim, which shifts the burden away from the claimant.

National patrimony laws in source countries declare cultural property to be state-owned from a given date. Their effectiveness abroad depends on whether the receiving country’s courts recognise them, and recognition varies considerably.

Alongside these sit the museums’ own policies and, in some countries, statutes that actively restrict deaccessioning, meaning a museum may be legally prevented from returning an object even when it wishes to.

Long-term loans as compromise

Where legal title is genuinely disputed or where a museum is barred from transferring ownership, the negotiated outcome is frequently a long-term loan.

The object physically returns, is displayed in the country of origin, and formal ownership remains unresolved or with the holding institution. Arrangements often include joint research, conservation support and staff exchange.

Assessments of this differ sharply, and both readings have force. Supporters note the object is where the claimants wanted it, often within a fraction of the time litigation would take. Critics note that it concedes the practical demand while withholding the principle, and that a loan can be ended.

The mechanism has nonetheless become the default route to resolution, largely because it sidesteps the deaccession restrictions that block outright transfer.

The stages of a claim

Reading down that column of failure points explains why outcomes look inconsistent from outside. Two objects with identical histories can diverge entirely because one is human remains covered by a directive statute and the other is a sculpture covered by a prospective convention.

What changed in the last decade

The direction of travel is clear even though the legal framework has moved slowly.

Several national governments commissioned reviews of colonial-era collections and accepted the principle of return in categories they previously contested. Museums began publishing provenance research proactively rather than responding to claims, which changes who bears the cost of investigation. Human remains and funerary material moved decisively into a category treated differently from other objects, in policy if not always in law. And acquisition standards tightened to the point where an object without documented provenance is now effectively unsaleable to a major institution.

That last change is quiet and consequential. It does not resolve any historic claim, but it addresses the mechanism that created them.

Sources

• UNESCO, 1970 Convention — the instrument governing illicit transfer of cultural property
• US National Park Service, NAGPRA — the affirmative duty to inventory and notify
• International Council of Museums, resources — the international body’s code of ethics and guidance on acquisition, provenance and restitution