The object before its name

Take an object that does not yet exist. It is easy to draw it, assign a polished name and describe how it might transform a future household. Almost anything can survive this treatment because nothing in the illustration has to work. The harder task begins by asking whether the proposed object is one coherent mechanism, several incompatible mechanisms or merely a contemporary wish given the shape of a product.

Objects do not all arrive as invented products. Some form out of habits. A particular arrangement of existing furniture becomes a recognizable workspace. A defect gets a diagnostic instrument. A discarded material acquires a new use; a new institution creates a physical token, a badge or an archival device. An object may exist technically for decades before it becomes socially ordinary. And an instrument may disappear because the question it was designed to answer no longer matters.

So the Cabinet should avoid a parade of invented consumer electronics. It can study objects not yet needed, objects physically possible but commercially irrational, objects socially useful but politically forbidden and objects that could become common without requiring any significant technical discovery.

Imagine a compact instrument for inspecting the long-term deformation of a folding screen. It does not diagnose 'broken' with a green or red light. It measures the profile, location and change of a crease and produces a comparable record across visits, without reading the phone's private data. A repair technician might use it to distinguish a cosmetic ridge from damage associated with delamination. Owners might use it in warranty disputes.

This is a conditional object proposal, not a claim that no comparable instrument exists. The physics of surface measurement is ordinary; reliable correlation with actual failure would be the hard part. The object might fail because integrated diagnostics make it redundant, because measurement protocols never standardize or because a photograph with reference lighting is good enough. Its future depends as much on repair institutions and legal evidence as on sensors.

Its possible vocabulary matters. A standardized severity scale can create names for grades of damage. Conversely, a shared popular name for a failure mode can create demand for a measurement tool. Object and word change each other's likelihood.

A second object might be a small, reusable physical history of repair, attached to a machine but controlled by its owner. Instead of an anti-tamper sticker whose main effect is to punish opening the case, it records that a component was changed, by whom if consented and under which technical procedure. It would have to remain readable without contacting a proprietary manufacturer service. Existing repair logs, seals and passports offer partial precedents.

The interesting problem is adversarial: how could a simple artifact distinguish an honestly repaired device from a forged history without creating a surveillance tag? Could it survive removal, replacement and disposal? Would a repair collective, insurer, buyer and manufacturer agree on what its marks mean? Physical security, verification costs, data minimization and governance are as important as fabrication.

A future collection should not present this as inevitable or as a novel product already validated. It is a proposed test of whether a right to repair can generate new physical artifacts and new conventions of trust.

Now return to the keyboard on the rug. It may never become an invention. No novel material, sensor or energy source is needed. Yet if people develop a repeatable way of using inexpensive keyboards on floor textiles, the arrangement could acquire rules: preferred spacing, support, hygiene, posture, storage. Eventually someone may sell a dedicated mat, and only then does an ordinary configuration become a marketed object.

There are many possible transitions. The practice may remain informal and unnamed. A manufacturer may try to define the category, while users continue speaking in phrases. A niche culture may give it a word no outside observer uses. None of these outcomes is dictated by mechanical possibility. A catalogue attentive only to patented gadgets would fail to see the object in the making.

A serious future-object entry should keep different claims apart: the observed present-day problem; existing analogues; the proposed function; the physical or computational principle; its missing enabling conditions; who would want it and why; its material and ecological costs; a counterdesign; and the observation that would show the prediction was wrong. For a cultural artifact, the fields need to change: ritual, legitimacy and users' meanings can be more explanatory than engineering.

Do not pretend these variables generate a numerical probability merely by listing them. An object may be feasible and still never occur. Invention depends on contingent sequences, institutional incentives, path-dependent standards and accidents. An essential step in prediction is to preserve the roads not taken rather than looking backward and declaring the winner obvious.

The most interesting object may be the one an otherwise successful future makes unnecessary. If foldable screens stop creasing visibly, the crease witness has no market. If repair becomes truly modular and well documented, the repair witness may disappear before it is born. A forecast of future objects also has to leave room for futures that abolish the need.