Things Do Not Intend
A letter was written to persuade you. A chronicle was composed to be read. An object was made to be used, and that difference is the whole argument for material culture as evidence.
A cooking pot is not trying to convince anyone of anything. Its wall thickness, its soot pattern, its repairs, and the wear on its rim are the residue of thousands of ordinary acts nobody thought worth recording. This is unwitting testimony in its purest form: the object cannot lie about what was done with it, because it was never making a claim in the first place.
And things reach where writing does not. The people who left no documents — the illiterate, the poor, the enslaved, the colonized — left pots, houses, tools, rubbish, and bones. For most of humanity across most of the past, material culture is not a supplement to the written record. It is the record.
What an Object Testifies To
| Property | What it reports |
|---|---|
| Material | Where it came from; what trade reached this place |
| Technique | The skill available, and how labour was organized |
| Wear | What was actually done with it, and how often |
| Repair | What it was worth to keep — the clearest proxy for scarcity |
| Deposition | How it was discarded, hidden, or buried, and by whom |
| Distribution | Who had one and who did not |
Repair deserves the emphasis. A society that mends is a society where the object costs more than the labour to fix it; a society that discards has reversed that ratio. Nobody in either society ever wrote the ratio down — and you can read it off a rubbish heap in an afternoon.
Buildings and Landscapes Are Objects Too
A house plan is a diagram of social relations rendered in stone. Where are the doors? Who can see whom? Is there a room a servant cannot enter, or a passage that lets them move without being seen? Georgian houses added corridors — and corridors are a machine for privacy, replacing rooms that opened into one another with rooms you can be shut out of. Nobody published a manifesto about wanting privacy. The floor plan announces it anyway.
Landscapes work the same way, at another scale: field boundaries, terraces, hedges, and ridge-and-furrow record centuries of decisions about labour, tenure, and inheritance. Enclosure is legible from the air as straight lines cutting across curved ones — a legal revolution you can photograph.
The trap: objects survive selectively, and the bias is physical. Stone, ceramic, and metal survive; wood, textile, leather, and food do not. So the record over-represents the durable, and therefore the ceremonial, the elite, and the buried — and under-represents almost everything people actually touched all day. A society that built in timber and dressed in cloth can look impoverished next to one that built in stone, when the only real difference is what the ground kept.
The historian's move is to read the silence as physics, not as absence. "No textiles" means "the soil was acidic", not "they had no clothes". And the moment a waterlogged site turns up — a bog, a well, a shipwreck — the missing 90% reappears at once, and the picture built from the durable 10% usually turns out to have been wrong in exactly the direction the survival bias predicted.