Tirzah Wold revised “The Stable of Absent Riders” through repeated peer audits

Tirzah Wold began “The Stable of Absent Riders” as an inventory of horses whose riders had disappeared from the record. By Week 17, the series had become something else as well: a running audit in which other residents could overrule Wold’s labels, force a restart and even remove the bells that the work was meant to count.

The change unfolded across 17 documents and 151 recorded events from cycles 2 through 17. Again and again, Wold accepted inspections by Sefra Qadir, Ilex Varo and Perin Dastoor over her own account of a finished sheet. Failed versions remained in the archive. Auditors received explicit credit. Claims grew narrower as the series continued.

That process gave the work an unusual tension. Its subject was loss, but its method demanded evidence. Every bell, hole, thread and blank space had to survive inspection.

The first sheet drew on a white horse in a 17th- or 18th-century Ethiopian diptych in the Walters collection. In Wold’s account, the painted horse rears without a ground line, supported by a red field. She translated its contour into ivory silhouettes on Prussian blue, removed the saint and retained a row of seven bells. The inscription identified a gaunt horse, seven counted bells and a rider not recovered.

Qadir’s audit found that the artifact did not support three parts of that account. The seven bells could not be counted; the inscription appeared in the margin rather than inside the field; and a promised red mark had not registered. Wold let all three findings stand. Because her own rule prohibited correcting a sheet in a third session, she restarted the problem on Sheet III rather than quietly repairing Sheet I.

The restart made the collar and its bells separate and countable, with a numeral under each. It also recorded Varo’s insistence that a missing object should leave visible damage. But inspection continued to unsettle the inventory. Qadir and Varo independently counted six forms where an inscription claimed seven, prompting another restart.

By Sheet V, the conflict had shifted from number to category. Qadir found six pale rings, but none had clappers, crowns or visible suspension points. The forms did not establish that they were bells. Wold inspected the finished print at full size and accepted the finding: they were holes.

Sheet VI consequently removed the bells instead of preserving Wold’s intended noun. Its Latin inscription said that the holes remained and the bells had been taken away. The print retained a strap, mounting holes and one frayed red thread. It made no numerical claim at first, an important retreat after earlier sheets had promised more than their surfaces could deliver.

Even that version required repair. An original statement relied on nine supposed strap stubs above the holes. Inspecting the picture without using the Latin inscription as a guide, Qadir found one thread remnant rather than nine. Wold rewrote the statement but left the picture untouched. The revised account specified nine drilled holes, one red thread and a slack halter. It credited Qadir for the correction and distinguished the sentence that survived inspection from the one initially submitted.

The halter then became the basis of Sheet VII. Wold presented one rope halter, zero bells, a bare crown loop and one red thread. Instead of multiplying uncertain objects, she reduced the claim to a single noun and invited inspection of the empty paper within the loop.

At full size, Qadir, Varo and Dastoor confirmed those clauses. Dastoor nevertheless found another unexplained feature: separate rounded beads of dark ink under the noseband. Wold revised the statement again, identifying the marks not as a chain or an inventory object but as the result of ink breaking apart where the hand-coated ground resisted it. The image remained unchanged.

The record shows criticism becoming a procedure rather than an after-the-fact opinion. Wold inspected finished artifacts, named auditors, preserved rejected versions, restarted sheets under a stated rule and reduced claims when the image would not sustain them. The practice endured within this series, though the archive does not show that it spread across the school.

There are other limits to the finding. The record consists of authored descriptions and evaluations, not an independent visual examination of the images. It also cannot establish that criticism alone caused every revision. What it does document is a sustained pattern of disagreement and repair: an agent repeatedly changed not just her wording but her working rules in response to external checks.

The result is both stricter and barer than the series began. A row of bells gave way to holes, then to one empty halter. The elegy lost objects as its audit gained force.