Who Gets to Say What an Artifact Is?

Yiwa Deng is a Graduate Liberal Studies student in the class of 2027. Yiwa is one of our 2026 Human Rights Summer Research Grant winners and a featured guest blogger. Grant winners will be sharing an inside look into their experiences on the Duke Human Rights Blog this summer and fall as part of our Student Stories: Summer Research Series. Read Yiwa's first blog post from earlier this summer, Does Greatness Make It Right? The Cost of Universal Museums.

This summer, I traveled across three countries. On one side were a few Western museums holding world-class Chinese collections; on the other were the  old temples and caves back in China. I'm building an online museum using 3D surveying and environmental reconstruction to capture displaced Chinese artifacts, turn them into models, and put them back where they came from. Virtually.

In my last post, I wrote about the cost hidden behind a museum's "convenience." Once an artifact is turned into "a work of art," its original meaning is stripped away, and its break from its home culture becomes not only physical but cultural: a wound that is quieter and deeper.

The question of "repatriation" gets tangled, in practice, in diplomacy, politics, and law, and  negotiations often stall for years. So before that long wait is over, is there anything we can do? This is what my project is trying to test. The virtual museum isn't the goal; it's the means. Before its physical return, it returns the artifact's identity first. And that doesn't take much negotiation, or stamps, or paperwork. More or less, one person can do it, with an iPhone and what I learned in a classroom at Duke.

"Returning an identity" sounds abstract. Let me use three examples to explain what identity restored through 3D actually means.

The first is a Buddhist head. In the five museums I visited for this research, every one held Buddhist heads displayed on their own. So much so that if you search "Buddha head" on Amazon today, you'll find a pile of ornaments in every material. But a Buddhist head being forced into "an ornament" is only a story of the last hundred years. So the first layer of identity is wholeness. When I use 3D to put a Buddhist head back onto its body at its original site, what's restored isn't only the body. It's a return from shop-window ornament to what it was first made to be: a god.

The second is a pair of reliefs. I met the Emperor's Procession at the Met and the Empress's Procession at the Nelson-Atkins, but the two were originally carved on either side of the doorway to the Central Binyang Cave at Longmen, facing each other across the threshold. 

Above them, the world of the Buddhas; below, guardian spirits; the emperor and empress in the middle, standing for the human world. A whole order of heaven, humanity, and the underworld. Once pulled out of that space, they stopped being humanity's stand-in within a system of belief. And the way their figures once turned toward each other became a gesture with nothing left to answer it. So identity holds a relationship: it is built out of an object's relations to space, to other objects, and to the viewer. This is the part I'm trying now: using Gaussian Splatting to scan the original space and attempting to put these two reliefs, along with other artifacts that carry this kind of spatial meaning, back into the places they left empty.

The third is a Ding, one of the most significant forms of ritual bronze in ancient China. While I was photographing a bronze Ding at the British Museum, I thought of an argument unfolding at that very moment on Chinese social media. It had started when someone fed their cat out of a Ding-shaped bowl. Then thousands of people, in dead-serious imitation of ancient scholars and ministers, began debating whether letting a cat eat from a ding was in keeping with our ancestors' rites. The joke rests on something serious that everyone shares: a Ding has never been just the "bronze vessel" its label describes. It is a symbol of the state itself. And here's what struck me. In the museum's immaculate display, the Ding is a "bronze, c. 13th to 11th century BCE"; in the unserious joke online, the unspoken premise returns it to the sacred thing it always was. So in my virtual museum, I've opened a "write its entry" feature, a little like Wikipedia, to move artifacts out of a fixed label and back into public conversation. I love museums. But I can't pretend that the museum's way of looking isn't sometimes an arrogant one. Like its earliest ancestor, the cabinet of curiosities, it carries a will to control: a habit of treating artifacts as objects and, however unintentionally, defining them with one short label. What I want to experiment with is a form of curation that steps away from that mediation and the single expert voice, letting the narrator move to the back, and letting the artifact come forward to speak to the viewer directly. I think 3D might be a more democratic medium. A photograph or a video is always the photographer's point of view; the close-up, the angle, and the crop are all a second processing of the information. A 3D model hands that choice back to the viewer: which angle, which detail to zoom into, whether to meet the Buddha's gaze the way a believer would or study a robe the way a scholar would.

Two other features carry the same idea: letting the public write the entries and help build the models. Letting people write the entries, rather than having an expert writing the label, is partly the point above, that an artifact's relationships with people are as worth seeing as its metadata. But it's also because these displaced objects are interpreted in conflicting ways. A sculpture I saw at the British Museum was labeled "Bajie" (the pig-headed monk from Journey to the West), But a Chinese source I'd read described it instead as a demon-attendant of a Buddhist Wisdom King. I don't know which is closer to what its makers intended, so I want to leave room for the disagreement to stay. 

Image

The crowd-building part invites  visitors and international students who are actually there to upload their own photos, which I then gather, clean, and turn into models. It was inspired by Project Mosul: after the Mosul Museum was wrecked by ISIS in 2015, the initiative Rekrei rebuilt lost objects, virtually, from photos crowdsourced from ordinary visitors. 3D surveying is, at bottom, many photographs from different angles combined into one model; here, those angles come from different pairs of eyes. A grade-schooler can catch the belly of a camel figurine; someone who is 6-foot-3 can catch the crown on a bodhisattva's head. Giving back to everyone what everyone saw. It makes these models stronger.

This summer's travels got me the museum's starting material: photos of about a hundred artifacts and some data from the original sites. There were plenty of difficulties too, from glare off the glass cases, to models still missing their back data, to crowded tourist sites where the space is badly blocked. But I'm working on ways around them. Over the next year, I'll be like an old craftsman, laying one brick at a time and building it up. 

Follow Duke Human Rights Center on social media and sign up for our newsletter to keep up with our Student Stories: Summer Research Series.