You are here:Home/News/ Caravans, Papyrus, and Cotton

Caravans, Papyrus, and Cotton

Posted On July 30, 2019 | 11:45 am | by ostmannj01 | Permalink
Anna Kelley upends the idea that the Roman world was the center of trade in late antiquity

Anna Kelley, who recently received her PhD from the Centre for Byzantine, Ottoman, and Modern Greek Studies at the University of Birmingham, was a 2018–2019 junior fellow in Byzantine Studies. Her research report, “Commodity, Commerce, and Economy: Reevaluating Cotton Production and Diffusion in the First Millennium,” drew from archaeological and historical evidence to follow the movement of cotton through Africa and the Mediterranean before 1000 CE.

Q&A with Anna Kelley

What did people in the late antique Mediterranean world think about cotton?

That’s a difficult question because during that time period, cotton coming from India was viewed differently from cotton grown in Africa. 

Cotton (Gossypium) from India typically came in finished form, many of the fabrics having different kinds of color decoration. They would also have varied quality-wise. Cotton textiles from Africa also encompassed a spectrum of quality and decoration, but in a different way. Many were plain woven pieces, though sometimes with decorations woven into them, like stripes or patterning in the yarns. What is interesting is the distribution of finds. Cotton artifacts are clustered either in the port cities on Egypt’s Red Sea coast, or in the areas where it was cultivated; it’s rare to find it elsewhere before the 6th century. Combined with evidence from the few documents that refer to cotton, I don’t think there was a great demand for it in the late antique world, and I don’t think it was viewed as a luxury in the areas where it was grown. For example, one papyrus mentions that cotton is used for shirts worn to work in the fields.

At the moment, we just don’t have enough evidence to say with certainty how cotton was used. In my research, I looked not only at cotton textiles but other kinds of evidence, particularly archaeobotanic remains to show cultivation, and papyri and ostraca (writing on potsherds) that record interactions of daily life. I also looked at production tools and anthropological studies of spinning practices to understand the material qualities of cotton, as well as the genomic taxonomy and evolution of Gossypium species to map its movement.

In the documents, different terms refer to cotton. One is the Greek term ἐριόξυλον/ἐρεόξυλον (erioxylon), a combination of the words for wool and wood/tree. They’re describing “wood wool,” wool that grows on trees or shrubs. Letters from the Western Desert of Egypt, where we know cotton was being grown, use this term erioxylon. Elsewhere, general terms used to describe textiles coming from India are terms that more commonly referred to finely woven linen. It seems the understanding in the Roman world was that they were getting a very nice, finely woven fabric from India—not something made from the same fiber being grown in Africa. 


What is the typical narrative about how cotton moved around the Mediterranean world, and how does your research challenge it? 

The narrative goes that trade with India introduced cotton into the Mediterranean region, and as demand for cotton as a product grew in the Roman world, the empire needed to get it from somewhere else. So they started growing cotton in Nubia, in the Western Desert of Egypt, and eventually in Fezzān in Libya, to feed back into the Roman world.

What I’m suggesting is that actually there is very little evidence of Roman cotton demand. The material, chronological, environmental, and spatial data simply does not support its existence. So what probably happened is cotton was used locally, maybe traded between areas that shared agricultural practices including its cultivation, but demand for it elsewhere was limited. It wasn’t widely transported across the Roman Empire.

Trans-Saharan trade is typically seen as a route through which the Romans got goods—animals, ivory, maybe slaves, gold—from sub-Saharan Africa. But I’m saying these communities didn’t just act as economic intermediaries or resource producers for the Roman world; they also maintained significant communication networks with each other.

This change in directionality is highly significant, because it shows more connectivity in Africa than previously thought. It opens up possibilities for looking at interaction with sub-Saharan Africa and between the communities in the Sahara. For example, we usually say caravan trade routes through the Sahara started after the spread of Islam, but that’s only because that’s when references started appearing in the sources. Actually there is archaeological evidence that even by the beginning of the first millennium, overlapping trade routes crisscrossed the desert, serving a variety of markets. 


How are you extending your research into the medieval world? 

My research indicates that while the Mediterranean economic network was a driving force, the late antique economy was not a unitary system with the Roman world as a center of both production and consumption, drawing everything else into it. There were networks moving around it as well. This is significant when looking at how the economy changed in the 7th to 9th centuries, during the spread of Islam. I am especially interested in changes in the relationship between production centers and markets in this time period. At Dumbarton Oaks, I’m using Byzantine collection textiles  to help me develop a more coherent narrative about the textile industry, and cotton in particular, in the first millennium, and the impact it had on economic systems.


Julia Ostmann is postgraduate writing and reporting fellow at Dumbarton Oaks. Photo by Elizabeth Muñoz Huber, postgraduate digital media fellow.