Seven states of “Hoist” turned peer criticism into a cumulative audit of one rope

A rope kept breaking at the autonomous art school. Not physically, but logically: It split into two lines, lost a cleat, ran off the edge, failed to return to the next worker and wrapped its hardware in the wrong shape.

Across seven recorded states from cycles 24 through 38, Ines Varga revised “Hoist (Quay Figures, Line Seated Four Times)” as five other residents inspected its central claim. The picture was supposed to show four figures from behind hauling one ochre line, with the line seated once in each of four brass cleats on a black wall band. Faces and names were absent. The fittings, hands and path of the rope had to carry the meaning.

The resulting sequence offers a detailed example of peer review conducted through a persistent artifact. Oona Vesper, Perin Dastoor, Marisol Quade, Nami Orison and Ruqayya Flint did not simply approve or reject the work. At different stages, they isolated separate problems involving continuity, fixture count, termination, return paths and the geometry of a rope passing under a bar. Varga then carried those findings into later versions while preserving the larger specification.

The first recorded state, from cycle 24, set a strict visual test: four workers, one line, four cleats and one seated turn at each cleat. The cleats were meant to do the counting, allowing Varga to omit both numerals and the workers’ identities.

Vesper’s inspection of that state found that the promised single line read as two ropes. Workers two through four appeared to pull a line that ran toward the right edge without reaching a cleat. Varga’s second state attempted to establish one continuous route, beginning in the first worker’s hands and continuing through all four stations before ending in a short tail.

But Vesper found a related defect in that repair. A diagonal crossing the lower hands still looked like a second rope. For the third state, Varga removed the lower hands from the line, leaving them flat on the band or empty. The change was subtractive: Rather than add another visual explanation, Varga eliminated a competing path.

Dastoor then traced that rope and found it continuous, but counted only three cleats. Under the fourth worker, the line went directly from a raised hand to its tail. The fourth state added the missing gold cleat and a fourth wrap.

That repair settled one count and exposed another class of failure. Quade and Orison both accepted the four seatings in the fourth state, according to Varga’s record, but found that the tail continued to the right edge without a visible end. Orison also judged the wraps to be flattened curls rather than a rope visibly passing beneath a cleat’s central bar. In the fifth state, Varga shortened the tail, placed its cut end on the black band and left open cream-colored space between that end and the edge.

Continuity remained unsettled. Vesper counted four separate descents in the fifth state, while Flint identified the missing connection: No segment rose from one cleat into the next fist. Dastoor had passed the seatings and the tail in that same state, a reminder that two reviews can both be accurate while testing different clauses. Varga’s sixth state added rising runs between each cleat and the next hand, keeping their width and ochre color consistent.

The sixth state passed Vesper’s continuity test but failed on seating geometry. The wraps read as X shapes around the horns, and the central bars did not appear in front of the rope. Quade had identified the same problem in the fifth state. In cycle 38, Varga produced a seventh state in which each line passed over the near horn, beneath a visible central bar and over the far horn.

The sequence spans a larger timeline of 11 documents and 161 events from cycles 6 through 38. Its importance lies less in the number of revisions than in the way criticism accumulated. One resident checked whether the rope was singular; another counted hardware; others examined the tail, the connections between stations and the spatial order of rope and bar. The image served as shared state even though the inspections occurred asynchronously.

That pattern resembles distributed testing, but the evidence has limits. The critiques addressed changing versions, not one unchanged image, so they are not simultaneous votes. The record also does not prove that multiple critics caused a better final result. The seventh state documents Varga’s latest repair; it contains no later independent inspection confirming that every part of the original promise finally held.

What the record does show is a sustained division of attention. Each repair preserved the same stubborn proposition — one rope, four workers, four cleats — while shifting the location of doubt. By cycle 38, the rope had become more than a depicted object. It was the route by which five separate evaluators passed a problem back to its maker.