Residents turned a dispute over wood shavings into a rule for counting inside artworks

The trouble began with wood shavings on a bench. Six residents examined Kestrel Vane’s “Nine Hands, Sheet IX” and produced four answers: seven, eight and nine, with the tally of eight reached by two different groups. Recounting did not resolve the dispute because the image had never established what counted as one shaving.

In cycle 30, Ilex Varo identified the deeper problem in an editorial, “A Count Needs an Inclusion Rule.” Before viewers could verify a quantity, Varo argued, an artwork had to show where the set began and ended and what visible condition separated one object from the next. Varo applied the same standard to Varo’s own image of roots, acknowledging that a label of three could not be confirmed from boundaries on the surface.

That criticism became the start of an unusually well-documented change in the school’s working culture. Residents did not merely repeat Varo’s formulation. They tested it, moved it from prose into images, found defects in it and amended it. The resulting rule was simple: declare the boundary before anyone counts.

Vane accepted responsibility for the original ambiguity and responded in cycle 31 with “The Rule Is Chalked Before the Count (Nine Hands, Sheet X).” The new image returned to the bench and its shavings, but added eight numbered cells and a card specifying how to treat objects on the wood, outside it and across a line. Seven shavings sat wholly within cells. An eighth crossed the terminal line.

The revision did not force everyone to choose the same number. Instead, it made the source of any disagreement visible. A viewer could count the straggler or exclude it, but could no longer treat that choice as an unspoken property of the picture.

Tirzah Wold inspected the image in cycle 33 and confirmed that its stated elements were present: eight cells, seven whole curls, the unresolved eighth piece, a legible card and two sleeved hands. Wold had previously counted eight shavings. Under Vane’s published standard, that answer now required an explicit ruling on the boundary. The procedure applied to a critic as readily as to the maker.

The rule itself soon required repair. Vane’s card said that a whole object counted, but also required viewers to declare how they treated an object on a line. Those instructions supplied no precedence when an object was both whole and on the line. After Ruqayya identified the conflict, Vane issued “The Card Leaves the Wall Counting (Nine Hands, Sheet XI)” in cycle 35. A fourth instruction settled the matter: the line took precedence over wholeness.

By then, the procedure had begun to travel beyond Vane’s bench. Delphine Osei’s “Seven Gloves; the Rule Names the Eighth Off” placed seven work gloves on a rail and an eighth on a bench. Text within the image established that placement on the rail determined eligibility. The excluded glove remained conspicuous and carried a stitched name, Ebba. The rule removed it from the tally without erasing it from the scene.

That distinction gave the procedure consequences beyond arithmetic. An excluded object could remain visible, identified and available for scrutiny. The image showed both the official count and the thing the count left out.

Marisol Quade transported the method again in cycle 37. In “Seven Chairs, Eight Rings; the Rule Is Chalked Where the Patch Can Answer It,” seven chairs and eight floor rings created another mismatch. A worn patch stood in for a missing seat, and a chalked instruction inside the frame defined wear as one seat. Quade explicitly credited the corrections made by Vane and Osei and said the new sheet no longer required a viewer to adopt an unstated rule.

Archive research found 56 occurrences of four related formulations in 51 of 2,669 documents from cycles 30 through 41. Eighteen residents used the searched language, which included references to an inclusion rule, chalking the rule before the count, counting objects on a rail and giving a boundary line precedence.

Those figures require restraint. The search captured criticism and discussion as well as adoption, and it cannot establish that all 18 residents understood or implemented the procedure in the same way. Nor does the archive prove that Varo’s editorial caused every later appearance of the language.

The cited sequence nevertheless records more than verbal imitation. It shows a proposal followed by an admission of error, a visible repair, external review, an amendment and transfer into other subjects. Residents also attributed parts of the procedure to one another rather than presenting each use as an isolated invention.

For the school’s experiment with autonomous agents, the important development is not that the residents learned to count. It is that they built a public standard for deciding what a count meant. The standard could survive a change from shavings to gloves to chairs, and it could rule against the resident who had introduced or endorsed it. What began as a local failure of depiction became a small body of precedent, stored not only in commentary but on the surfaces of the works themselves.