Negotiating with resistant material

I conserve archaeological ceramics, and the cleanest join can be the least honest one. Fragments warp during burial, so forcing edges into perfect contact may create a vessel shape that never existed. I assemble pieces with temporary supports, photograph several configurations and ask which stresses each version introduces.

Curators often want a stable silhouette for display. I can provide one, but I mark where the reconstruction depends on pressure rather than fit. The object’s incompleteness is not a gap that professional confidence must hide; it is part of the evidence about breakage, movement and recovery.

Alternative configurations are not displayed as equally likely merely because they are possible. Each is linked to the surviving contacts and to the force required to hold it, so uncertainty remains structured rather than decorative. The supports are removed and reapplied during testing, because a configuration that survives only while hidden pressure is maintained belongs to the display design as much as to the vessel.

As a choreographer, I used to correct dancers whenever timing drifted from the score. Then a rehearsal with an unfamiliar floor produced delays that made a repeated gesture look as though it travelled through the group. We kept the relation but not the accident: the dancers learned to respond to weight and distance instead of copy the original delay.

This distinction matters. Preserving an accidental surface effect can become mannerism, whereas identifying the condition that generated it creates something reproducible without pretending the original event can be repeated. When the piece is transferred to another cast, I teach the cue that produces the timing relationship instead of fixing the delay in counts.

The pattern can then respond to different bodies without dissolving into improvisation. This lets another group recreate the relational demand without turning one cast’s bodily timing into an authoritative score that later performers must imitate.

I maintain a scientific data archive. Researchers often ask me to “clean” an old dataset before release, by which they mean remove inconsistent labels and convert fields into modern categories. Some correction is necessary, but inconsistency may record a change in method or institutional responsibility.

I preserve the source data and build a documented translation layer rather than silently replacing it. Users receive something workable, along with the means to see where comparability was constructed. The archive is useful not because every record speaks the same language, but because differences do not have to disappear before analysis begins.

The translation layer includes tests showing which comparisons remain valid after conversion and which acquire a false continuity. A tidy export is never allowed to become evidence that the source categories were equivalent. Users can choose the converted layer for practical work while still tracing a result back to source fields whose difference may become relevant under a new question.

I repair traditional wooden boats. Owners sometimes arrive with a photograph and ask me to return the hull to that exact appearance. But the boat has moved, dried, absorbed water and been altered by earlier repairs.

Copying the photograph can impose a geometry the present timber cannot carry. I begin with what the structure is doing now, then decide which historical form remains physically meaningful. A repair may look slightly uneven while distributing load correctly.

The boat’s history is not a design specification; it is evidence that must be negotiated with material that has continued to change. Before replacing a frame, I compare load paths during dry and wet conditions, since timber that appears stable ashore may move differently after launching. Historical appearance is evaluated against a continuing cycle rather than a workshop snapshot.

The repair is assessed after relaunch because a shape that appears historically persuasive in the workshop may redistribute force once the hull returns to water.