Silk on Bronze Masks at a City With No Writing

Sanxingdui, in Sichuan, is the site with the bug-eyed bronze masks and no systematic writing. A paper published online in Antiquity on 26 October 2026 says the masks and the other ritual metal did not go into the pits bare.

Jianbo Guo, Ruochen Ge, Qiutong Cai, Jingyu Tian and Lu Yang found silk residue on 101 objects from sacrificial pits K2 to K8. Ninety-four of those objects are bronze, across 28 types. Gold, jade and ivory carried traces too. Under the electron microscope the fibres are smooth, about 10 micrometres wide, with a rounded triangular cross-section. That is mulberry silk, Bombyx mori. Infrared tests found protein. The tight weave rules out hair.

One large mask from pit K3 had three separate silk patches around the right eyebrow and eye. Different weaves. Different states of preservation. Some of the silk was dyed red. The residues date to the Late Shang period, about 1300 to 1046 BCE.

The authors say the silk most likely cushioned, covered or wrapped the offerings. They also note, as an idea and not a result, that if the Shu wrote, silk could have been the surface. There is no text on these scraps.

Key Evidence

• 101 objects with silk residue, 94 of them bronze (Antiquity, 26 October 2026).
• Fibre shape and width match Bombyx mori.
• Three distinct silk patches on one mask, around the right eye.
• Some residue dyed red.

The Rational Explanation

Fibres move in soil. A residue on a buried bronze is not a reconstructed garment, and it is not a library. "They wrote on it" is a sentence the paper does not earn.

What We Don't Know

What the mask looked like when the silk was still cloth. Whether the three weaves on one face were one ritual or three. Whether any of this silk ever carried writing. The pits are still mostly a ceremony we can see and cannot read.

The Rabbit Hole

Sanxingdui already looked like a culture that spent its skill on faces. The new fact is that some of those faces wore cloth over the eye before the pits were closed.