The Two Counts

In December 2025, Canada's federal government told Parliament that 71 original artworks in the Indigenous Art Centre's collection remained unaccounted for. In August 2026, reporters were told 111. Both figures come from the same internal audit. They count different kinds of things at different dates, and the public record does not show how one became the other.

Nothing in the public record establishes a theft, a lie, or a single missing painting. The failure is smaller than that, and more durable: a public collection has become hard to follow while the institution responsible for it is speaking on the record.

The Indigenous Art Centre was established by the federal government in 1965 to support the creation, promotion and preservation of contemporary Indigenous art in Canada. Its collection now contains more than 5,000 works by more than 990 First Nations, Métis and Inuit artists. The centre says the works were acquired through a peer-juried process. Its public description of the collection puts the value at about C$14.4 million and highlights works by artists including Norval Morrisseau, Christi Belcourt, Elisapee Ishulutaq and Kent Monkman.

The audit that began this story was an internal audit by Crown-Indigenous Relations and Northern Affairs Canada, or CIRNAC, covering the period April 2019 to August 2024 and dated November 2024. Its findings reached the public through the December 2025 parliamentary answer and the August 2026 reporting described below. It found 132 items in the records that could not be accounted for. The audit did not identify the individual works or assign them a value.

The parliamentary answer, tabled in response to Written Question Q-535, did something precise with those 132 items. It subtracted 13 records judged to be duplicates, leaving 119 entries still unaccounted for. It then subtracted 20 reproductions, 25 prints, two archival items and one reference item. That left 71 original artworks.

Two other items were under investigation as possible duplicates. They remained in the count.

The August account used another route through the same starting number. Seventeen items had been resolved: six were located in the centre's own storage facility and 11 were found to be duplicate entries in the database. That left 115 entries. Four additional items were flagged as potential duplicate entries and subtracted from the headline, producing 111 entries still unaccounted for. The August account said that 20 of the remaining 111 were reproductions.

The two accounts cannot be placed cleanly on the same terms. December's 119 removes 13 records judged duplicates. August's 115 removes 17 items: 11 duplicates and six works located in storage. Whether the 11 and the 13 describe the same category is not something the public record establishes, and without that the apparent movement between the two figures cannot be read as a measure of recovery. Four further entries were flagged as potential duplicates and excluded from the August headline without being resolved, so the lower headline is partly a change in how unconfirmed duplicates are treated. The identities and the exact reclassification path are not in the public record examined for this article.

Nor does the record permit a precise conversion from the August entry count to a current count of original artworks. The August account identifies 20 reproductions among the 111 entries, but it does not show how the print, archival and reference categories used in December overlap with the later entry count, or which of the located items were originals. Subtracting the 20 reproductions produces 91 entries, not a count of original artworks.

Eight months is long enough to reconcile two numbers. The parliamentary record shows no attempt to do it there.

Q-535, submitted by Jamie Schmale, was answered on Dec. 10, 2025, in Sessional Paper 8555-451-535. The House's cumulative Status of House Business register dated July 22, 2026, records that answer and no revised return. The register does record revised returns elsewhere, using a second line and a "-01" sessional-paper suffix; a reading of that register for this article counted four such revised answers in the current session. Q-535 has none.

A review of the 45th Parliament's written-question record, conducted for this article and current to Sept. 11, 2026, found no later question on the Indigenous Art Centre, the collection or the unaccounted-for artworks. The House sat again after the July register, but the Aug. 19 Journals contain no later Q-535 return and no mention of the collection. Work may well be going on inside government; the record cannot speak to that. What it shows is narrower. Parliament's public answer is still the December answer, while the department has given journalists a later one.

There is a defence available here, and it is stronger than anything the department has said out loud. Nobody has made it publicly, so let me make it. The two figures were produced eight months apart, by different people, in answer to different questions, for different audiences. A written answer to Parliament is a formal instrument that responds to what was asked, on the day it was asked. A spokesperson's reply to a reporter is a status update. Nothing in Canadian practice obliges a department to file a revised return every time an internal count moves, and reconciling the two in public would mean publishing a working document that is still changing.

The register is the answer to that. Revised returns to written questions do get filed — four of them appear in this session alone, each carrying its own second line and suffix. The record was correctable by an ordinary, well-used procedure. Eight months passed, a newer figure went to reporters, and the record was not corrected.

CIRNAC's language is careful. Maryéva Métellus, a department spokesperson, told Hadani Ditmars of The Art Newspaper, in a report published Aug. 21, that "unaccounted for" is an inventory and records-management designation. She said it means that an artwork's location cannot be confirmed through the available documentation, and does not mean the work is lost or stolen. She said no criminal investigation into the items had been opened. She described the work of checking historical inventory, loan, exhibition, storage and collection-management records, some of which stretch across decades and systems.

Jacinthe Goulet, another CIRNAC spokesperson, told The Canadian Press, in a report published by iPolitics on Aug. 14, that 17 of the 132 items had been resolved. Six were in the centre's storage facility. Eleven were duplicate database entries. Four more had been flagged as potential duplicates.

A records problem is not automatically a disappearance. But the distinction cuts in both directions. If "unaccounted for" is only a provisional status, then the public needs a way to follow the provisional status to its conclusion. If the records cannot establish where an artwork is, that uncertainty is itself part of the collection's condition. A label can describe uncertainty without making it less real.

The department's own audit makes the problem harder to dismiss as a matter of vocabulary. It recorded items the records could not account for. It warned that the recovery policy was unclear and that the last approved policy instrument dated to 1990 — thirty-four years old when the auditors examined it. It identified risks associated with the collection's 2022 move to private-sector storage, theft and security breaches among them. And when the records cannot locate an item, the audit found no clear procedure for getting it back, or for establishing that it never existed separately in the first place.

The institution that described its count in two different units to two different audiences is the same institution the audit found working from a policy approved in 1990, with its holdings in third-party storage and no settled way to recover what the records cannot place. The mismatch is the visible edge of that.

One limit on that reading should be stated plainly. The audit document itself was not obtained for this article; its findings are known here through the December parliamentary answer and the August reporting. The language above about the 1990 instrument and the 2022 move is reported rather than read.

A different reading comes from Cory Dingle, executive director of the Norval Morrisseau estate. Speaking to Ditmars, Dingle described the development as a positive moment in Canada's continuing cultural evolution: an opportunity to recognise that Indigenous art, history and cultural heritage deserve serious protection and investment. He said CIRNAC and the wider cultural sector needed the resources and tools to do the work properly, and, as Ditmars reported it, characterised the estate's dealings with CIRNAC as supportive.

Dingle is not wrong about the difficulty. Public collections are hard institutions to run. Their records are old, their holdings move between storage and exhibition, and their histories sit across paper files, databases, loan agreements and the memories of people who have retired. A government may inherit decades of weak practice without having created each gap in it.

Resources still do not supply the narrower answer required here. Better funding may help the centre reconcile the records, and scarce staff or fragmented systems may help explain why the bridge was not produced. But they do not explain, on the evidence now public, why Parliament received a count of originals while journalists received a count of entries, or why nobody published the conversion. They do not discharge the obligation to make the two descriptions intelligible.

Lawrence Paul Yuxweluptun, whose work is held by the centre, applied the opposite pressure in the same Art Newspaper report. He said the government should look after the art, and argued that the collection might sit better with the Canada Council for the Arts. His criticism is about the collection as a whole; nothing in the public record connects his own work to the unaccounted-for items.

That proposal deserves more attention than it has had, because it is the only structural idea anyone has put on the table. The Canada Council is an arm's-length Crown corporation whose entire subject is art. CIRNAC's mandate runs to treaties, claims and northern affairs, and the collection sits inside it as one holding among many responsibilities. A transfer would locate nothing. What it would change is who has to describe the collection in public, and to whom — which is where this has gone wrong.

The centre's own description names Morrisseau, Belcourt, Ishulutaq, Monkman. A collection of that kind is a record of what a government chose to acquire, preserve, exhibit and represent, and therefore of what it undertook to keep. When the institution cannot say plainly what kind of thing its number counts, that undertaking stops being something the public can check.

The public was handed a number of originals, and then, eight months later, a number of entries. Both were derived from the same 132 audit items. Neither was a correction of the other. The institution's language made the distinction technically available; it did not make the relationship visible.

Opacity of this kind is easy to underestimate because it has no cinematic moment — no locked room, no ledger discovered in a basement. There is a table, a database, a storage facility, an old policy, and a series of answers whose differences are left for somebody else to reconcile. What is missing is the document that would join them.

CIRNAC told The Art Newspaper that it has begun updating how the artworks are acquired, cared for and exhibited, and that the work is ongoing. That may be true. The centre may eventually account for every item, or establish that some records describe objects that never existed as separate works. The public record may become clearer. In eight months it has not. The document that would make it clearer would not have to promise certainty the department does not have. It would have to show the categories, give the dates, identify which items were located in storage, distinguish duplicate records from physical works, and set out how the 71 original artworks relate to the 111 entries. One page would do it.

Until then, the collection exists in two administrative versions. One was delivered to Parliament: 71 original artworks, after a set of categories was removed. The other was delivered to reporters: 111 entries, after duplicates were treated differently. The public record reports six items located and 11 duplicate entries resolved, and provides no item-level map that would place the two accounts on the same terms.

This account rests on the department's published statements to Parliament and to reporters, and on the parliamentary record itself. CIRNAC was not approached for this article. Nothing in the record indicates the government said anything untrue.

One document was promised. The December answer told Parliament that an updated list of titles and artists would be shared once written consent was obtained. Nine months on, no such list has appeared in the public record examined for this article, and the August coverage does not say whether consent was sought and declined, or whether the question was never asked. That private history cannot be invented from its absence.

No particular artist should be told that a particular work is missing on the basis of these aggregate numbers. That is exactly what the promised list was for. Until it exists, an artist who wants to know whether their own work is among the entries the records cannot place has no public document to consult — and no way to know which of the two counts the answer would come in.

Sources

• House of Commons of Canada, Written Question Q-535, answer tabled Dec. 10, 2025, Sessional Paper 8555-451-535.
• House of Commons of Canada, Status of House Business, 45th Parliament, 1st session, July 22, 2026.
• House of Commons of Canada, Journals, Aug. 19, 2026.
• CIRNAC, internal audit of the Indigenous Art Centre collection, November 2024, as described in the sources above. The audit document itself was not obtained for this article.
• Hadani Ditmars, "More than 100 works from Canada's federal collection of Indigenous art are 'unaccounted for,'" The Art Newspaper, Aug. 21, 2026.
• The Canadian Press, report carried by iPolitics, Aug. 14, 2026.
• The Indigenous Art Centre, public collection description, source of the holdings, artist and valuation figures given above.