Dhaka Muslin: The Finest Cloth on Earth, and How It Was Killed

For most of recorded history, the finest cloth on earth was made by hand in a stretch of river country in eastern Bengal. Dhaka muslin was cotton spun so fine that a sari of it could be drawn through a finger ring, and a length of it laid on grass and left overnight was said to be indistinguishable from the dew. The Romans bought it. The Mughal court reserved the best grades for itself. European aristocrats paid fortunes for it. And then, within roughly a century, it was gone — not outcompeted by something better, but administered out of existence. That last part is the part I keep coming back to.

What “woven air” actually meant

The superlatives around this fabric sound like folklore until you look at the numbers. Ordinary good-quality muslin runs somewhere around 40 to 80 threads per inch. The finest Dhaka cloth has been measured at warp densities in the range of 800 to 1,200 threads per inch — an order of magnitude finer than what a skilled weaver anywhere else was producing. The yarn itself, by Mughal-era accounts, ran to counts of 300 and above, which is finer than most modern industrial spinning can reliably manage even now.

The grades had names that tell you how people experienced the stuff. Abrawan — running water. Shabnam — evening dew. Malmal khas, the special muslin, woven for the emperor and not sold. There is a well-travelled story about a Mughal princess being scolded for appearing immodestly dressed when she was in fact wearing several layers of it. I would treat that particular anecdote as court legend rather than documented fact, but the physical property it points at is real and measurable: the cloth was close to transparent, and it weighed almost nothing.

None of this was a trick. It was the output of an enormously deep production system, and that system is the actual subject here.

Why it could only be made there

Three things had to line up, and all three were local.

The first was a specific cotton. Not cotton in general — a particular variety, phuti karpas, generally identified as Gossypium arboreum var. neglecta, which grew along the banks of the Meghna river a short distance from Dhaka. Its fibre was short and, by the standards of any spinning machine, useless: too weak, too prone to fraying, exactly what an industrial spindle cannot handle. In human hands, in the right conditions, that same fibre took a twist no long-staple cotton could match.

The second was the river valley itself. Humidity was not incidental to the process, it was part of the equipment. The finest spinning was done in the damp hours around dawn and dusk, and accounts describe women working on or beside the water, because dry air snapped the thread. When a production process depends on the weather at a particular bend in a particular river, it is not portable. You cannot license it. You cannot relocate it to Lancashire.

The third, and the one that actually mattered most, was accumulated human skill. Getting from a cotton boll to a finished piece involved on the order of a dozen or more distinct operations, each with its own specialists — cleaning the fibre, sometimes with the small jaw of a boalee fish; spinning; warping; weaving; washing and finishing. Dhaka in its peak years supported a very large weaving population; figures in the tens of thousands get quoted, and while I would not lean hard on any single number, the order of magnitude is not in doubt. A spinner capable of 300-count yarn was the product of a childhood spent learning it from someone who had spent their childhood learning it. That kind of skill lives in people, not in documents, and it dies with them.

A technique that exists only in the hands of living practitioners is one bad generation away from being a mystery.

The market that made it, and the market that unmade it

Before the eighteenth century, Bengal was one of the manufacturing centres of the world. India as a whole is commonly estimated to have produced something like a quarter of global industrial output around 1750 — the estimates vary by method and are contested at the edges, but the picture they describe is consistent. Bengal supplied a huge share of what European trading companies carried home. Textiles were the anchor commodity of the entire Europe–Asia trade.

Which is precisely why it became a political problem in London. Indian cloth was better and cheaper, and English wool and silk interests said so loudly. Britain’s answer was not a better loom. It was legislation. Import duties on Indian goods climbed through the late 1600s. The Calico Acts of 1700 and 1721 banned first the import of printed and dyed Indian fabrics and then, more sweepingly, the domestic use of most calicoes in England. This was straightforward protection of a weaker industry against a stronger one, decades before the spinning jenny existed.

Then the asymmetry became structural. After the Company’s acquisition of the diwani of Bengal in 1765, the same power that governed the producers also set the terms of trade. Indian cloth entering Britain faced duties commonly cited in the range of 27 to 59 per cent in the early 1800s, rising to something like 70 to 85 per cent by the 1810s. British machine-made cloth entering India paid a nominal few per cent. Exact figures differ between sources and between years, and I would hedge any single number, but the shape is unmistakable and no serious historian disputes it: the tariff wall ran in one direction only.

Add Manchester’s mills coming online, capable of producing acceptable cloth at prices no hand-spinner on earth could match, and you have a market being flooded from one side while being locked out from the other. India’s share of world manufacturing output fell from roughly a quarter in 1750 to a low single-digit percentage by the late nineteenth century. That is not a slow drift. That is a demolition.

What happened to the weavers

Alongside the trade rules sat the procurement system. Under Company and agent control, weavers were increasingly bound by advance contracts, tied to particular buyers, and paid at prices set by the buyer rather than the market. Contemporary critics of the Company’s conduct in Bengal described coercion, confinement, and fines. William Bolts, a disaffected former Company man writing in 1772, is the most cited of these, and his account is genuinely damning about the conditions imposed on Bengali artisans.

Bolts is also the origin of the story most people have heard: that weavers’ thumbs were cut off to stop them working. I want to be careful here, because that claim is contested. Historians have pointed out that the passage is ambiguous, that it may describe weavers mutilating themselves to escape forced work rather than British officials mutilating them, and that there is no evidence it was ever policy. I do not think the story should be asserted as fact, and I am not going to assert it. I also do not think the argument needs it. The documented mechanisms — tariff asymmetry, price-setting, contractual bondage, the collapse of the courts and nobility who had been the buyers of the highest grades — are more than sufficient to explain what happened, and they are worse in a way, because they were legal, deliberate, and written down.

The demand side went too. When Mughal authority fragmented and local courts lost their revenues, the patrons who commissioned malmal khas stopped commissioning anything. A craft at that level cannot survive on the mass market; it needs someone willing to pay for months of one person’s labour.

The double death

So the weavers left. They went into agriculture, or into worse-paid coarse weaving, or into nothing. By around 1800 British observers were already writing about the finest Dhaka muslin in the past tense. Cultivation of phuti karpas stopped, because nobody was buying the yarn made from it, and the plant — commercially useless for anything else, unusable by machines — went extinct in cultivation. Preserved specimens ended up in a herbarium at Kew, which is a fairly exact image of what colonialism did: the plant filed in London, the industry dead in Bengal.

Both halves of the knowledge died. The genetic material and the human technique. Losing one would have been recoverable. Losing both meant that when people later wanted the cloth back, there was no thread to pull.

Superior technique does not win on its own. Whoever writes the rules of trade decides which technique gets to exist.

The rule underneath the story

Here is why I think this matters beyond textile history. There is a comfortable story in which handmade things lose to machines because machines are better, and this is simply what progress looks like — regrettable, impersonal, nobody’s fault. Dhaka muslin does not fit that story. The machines never made anything close to it. The mills won a market they were not technically capable of winning, because the market’s rules were rewritten by the people who owned the mills and governed the producers at the same time.

That is the pattern I keep finding when I look at how technology gets captured: the decisive fight is almost never over which technology is better. It is over who sets the terms — the tariffs, the standards, the contracts, the platform rules. The East India Company is the clearest historical case precisely because it held both roles at once, as merchant and as sovereign, and could therefore make its commercial preferences into law. The same logic that turned common land into private property during the enclosure movement was at work here: a resource held and worked by many, converted by legal instrument into an asset controlled by few. And the line runs forward as much as back — the arrangement that took production out of Bengali households and put it inside Lancashire mills is recognisably the same arrangement I traced in from cotton mill to smartphone, where the value made by millions of dispersed people gets funnelled into a mechanism somebody else owns.

Deindustrialisation, in this case, was a policy outcome. Calling it progress is a way of not naming the decision-makers.

Trying to get it back

Since around 2014 there has been a serious effort in Bangladesh to reconstruct the fabric — botanists searching for surviving relatives of phuti karpas, DNA work on herbarium specimens, replanting near the old growing areas, and training spinners and weavers almost from first principles. Related jamdani weaving, which survived as the coarser descendant of the tradition, was inscribed by UNESCO as intangible cultural heritage in 2013, and Dhakai muslin received Geographical Indication status in 2020. Reports of the revival work describe cloth reaching counts in the low hundreds — impressive, and still short of the historical peak.

I find the revival genuinely moving and also slightly beside the point. You can, with a decade of national effort, partially reconstruct a lost fabric. What you cannot reconstruct is the ecosystem that made it ordinary: the growers, the spinners, the finishers, the buyers, the whole dense economy in which a child could learn 300-count spinning because everyone around her already knew it. That took centuries to build and about a century to destroy, and the destruction was cheaper than the building by an enormous margin.

That asymmetry is the lesson I would keep. Capability is slow and cumulative; policy is fast and decisive. A country can hold the best technique in the world and still lose the industry, if it does not hold the pen.