Published online 17 February 2009 | Nature | doi:10.1038/457945a

Rethinking silk's origins
Did the Indian subcontinent start spinning without Chinese know-how?

Philip Ball

Ancient silk strands have been found in artefacts from Pakistan’s Indus valley.I. Good
New findings suggest that silk making was not an exclusively Chinese technological innovation, but instead arose independently on the Indian subcontinent.

Ornaments from the Indus valley in east Pakistan, where the Harappan culture flourished more than 4,000 years ago, seem to contain silk spun by silk moths native to the region. What's more, the silk seems to have been processed in a way previously thought to have been a closely guarded secret within China.

There is hard and fast evidence for silk production in China back to around 2570 BC; the newly discovered objects are believed to date from between 2450 BC and 2000 BC, making them similarly ancient. There have been no previous finds of manufactured silk at sites outside China before about 1500 BC.

"This is the first evidence for silk anywhere out of China at such an early date," says Irene Good of Harvard University in Cambridge, Massachusetts, one of the authors of the study. "It was a complete surprise."

The objects come from two sites in the Indus valley: the city of Harappa itself, the hub of the Indus civilization, and Chanhu-daro in Sindh province, about 500 kilometres to the south. They were collected from archaeological excavations in 1999 and 2000 conducted by the Harappa Archaeological Research Project (HARP), a US–Pakistan collaboration. Because of the sheer volume of artefacts amassed so far, they have only recently been studied in detail.

Good, working with HARP directors Richard Meadow of Harvard University and Jonathan Kenoyer of the University of Wisconsin in Madison, used an electron microscope to look at the fine structure of silk strands found in necklaces and bangles.

The precise shape of the individual silk threads — determined by the shape of the orifice through which they are extruded — is characteristic of the species of silk moth that produced the strands.

In a paper in the journal Archaeometry, the researchers show that the Harappa samples — two metal ornaments — contain silk from species of Antheraea moths indigenous to south Asia (I. L. Good, J. M. Kenoyer and R. H. Meadow Archaeometry doi:10.1111/j.1475-4754.2008.00454.x; 2009). The origin of the Chanhu-daro silk, threaded through soapstone beads, is less clear, but it may be from one of the same species. Chinese silk comes from the domesticated silk moth Bombyx mori.

The Harappan silks seem to have been made by a process called reeling, in which the strands are collected on a bobbin rather than being twisted in short segments into a thread. The researchers say that reeling was thought to have been part of a silk technology known only to China until the early centuries AD. Now it seems that knowledge was not so exclusive.

"Archaeology in early China is showing increasingly there were connections outside China," says Shelagh Vainker, a silk expert at the Ashmolean Museum in Oxford, UK. "It doesn't seem unreasonable." But she sees evidence for silk production in China "significantly earlier" than 2500–2000 BC, which would suggest China could still claim priority.

"I believe that the people of the Indus civilization either harvested silkworm cocoons or traded with people who did, and that they knew a considerable amount about silk," says Good. She does, however, acknowledge that some might find this challenging: "National pride is at stake with such a discovery as this."