Textile Automation: Spinning Machines, Power Looms and the Jacquard Loom

The merchant already had a system. It was just spread out

In the putting-out trade the capitalist did not need a factory. He put fiber into households, paid for the yarn or the cloth that came back, and sold the pile. The pace was the family’s, until he was waiting on an order. Quality wandered. So did the date. Maxine Berg’s point, already used on the factory page, is that this world did not die when the first mill opened. Workshops and houses kept feeding the same markets.

You will hear a neat cause. John Kay’s flying shuttle, patented in 1733, let a weaver throw the weft faster, the weavers then starved for yarn, and the spinning machines were invented to catch up. The shuttle is real, and it is a hand tool, not a factory. The dates of the spinning machines do fall later. The claim that one patent forced the next, parish by parish, is cleaner in the retelling than in the order books. I will not hang the mill on that one arrow.

Three spinning machines, three different rooms

The jenny stays where a person can turn it

James Hargreaves’s spinning jenny, patented in 1770, lets one spinner draw and twist several yarns at once. Early jennies are hand-powered. They fit a cottage or a small shop. The thread was useful, and for a long time it was the weft, the softer cross-thread, more than the warp that holds the cloth. A jenny multiplies spindles. It does not build a factory, and it does not run while the spinner is gone.

The water frame wants a wheel, and the patent did not hold

Richard Arkwright’s name sits on the water frame: rollers that draft the fiber, then a twist, with the machine sized for power rather than for a hand crank. He took a patent in 1769. Cromford, in 1771, put frames in a mill on the Derwent and turned them with a waterwheel. That building is the factory page’s example, and it stays a water mill here.

He was not a lone inventor with a clean title. The clockmaker John Kay built machines for him. Thomas Highs said the rollers were his. In 1785 a court cancelled Arkwright’s patents. After that, other mill owners could use the frames without paying him. Treat “Arkwright invented the factory” as a short way of pointing at Cromford. It is not a verdict the court was willing to sign.

The mule is the machine that mattered, and it was not self-acting yet

Samuel Crompton’s mule, worked out around 1779, drew on the jenny and on the rollers. It could spin a finer, stronger yarn. He did not patent it at the start. Other builders took it up, powered it, and enlarged it. For decades the spinner still controlled the carriage: the outward draw, the twist, the wind back onto the spindle. That skill was the job. Richard Roberts’s self-acting mule, from the late 1820s, took the return of the carriage off those hands so one minder could watch more machines. It spread where the yarn count suited it. It did not empty the room. Piecers were still tying ends. Doffers were still lifting full bobbins. A self-actor is a later machine than the one in the school story, and it is still a tended machine.

The mill is a power decision before it is a social theory

A frame that only pays if it runs all day wants a prime mover that does not get tired. At Cromford that mover was the river. Richard Hills’s history of power in the Industrial Revolution is the place to watch water stay in the picture after steam exists. An engine, as the steam page said, is a choice where you can get coal. It is not the thing that created the cotton mill, and it did not make every valley drop its wheel.

Once the machines share a shaft, they share a failure. A flood, a drought, a thrown belt, or a broken wheel gear stops a row, not a single spindle. Maintenance is part of the output. So is the overlooker. Fitton and Wadsworth’s study of the Strutts and the Arkwrights is an old book, and it is still the close look at those early mills as businesses: partners, buildings, and a payroll, not a spark of genius in a shed.

The children in that payroll are the reason the factory acts were textile acts. The hour limits and the four inspectors are written on the factory page. I will not pretend they were kept because they had been printed.

The loom did not catch the spindle

More yarn is a gift to a weaver only if the weaver can take it. Mechanized spinning piled yarn up in front of a trade that was still mostly hands. Edmund Cartwright patented a power loom in 1785. His own works did not establish it. The early machines were crude, the warp was miserable to prepare, and a handloom weaver in a cottage could still beat the economics. Roger Holden’s account, already used on the factory page, is the one to keep: a loom that could cover the range of Lancashire cloths, at a speed one person could tend more than one of, is a nineteenth-century machine. His conclusion puts that capability around 1860. William Horrocks and later builders matter more to the shed that actually ran than Cartwright’s patent date does.

A power loom ties the motions a handweaver had coordinated — shed, shuttle, beat, take-up — into one driven cycle. Somebody still warped the beam, dressed the threads, watched for a smash, and mended the cloth. “The loom was automated” is true for a list of motions and false for the room. Handloom weavers did not vanish when the patent was sealed. They were squeezed, and a great many of them were still weaving.

The cards pick the threads. They do not think

A plain loom lifts the warp in a simple rotation. A patterned silk, a portrait in the cloth, needs a different set of threads lifted on every pass. On a drawloom a drawboy, often a child, pulled the cords the pattern called for. It was slow, and a missed cord was a flaw that ran until someone saw it.

The punched control is older than the man whose name stuck. In 1725 Basile Bouchon used a strip of punched paper to select hooks. It could not carry a wide pattern, and it did not spread. Falcon, in Lyon, put the pattern on a chain of punched cards instead of a paper strip, so more hooks could be selected. Jacques de Vaucanson, in the 1740s, tried a pierced cylinder and tried to do without the drawboy. That loom did not become a trade. Joseph-Marie Jacquard, a Lyon weaver, worked through Vaucanson’s machine at the Conservatoire in Paris in 1804 and, by 1805, had a head that used Falcon’s chain of cards pressed against a cylinder of needles. What carried his name is that combination, as an attachment on a loom, not a new idea of holes and not a building that wove by itself.

Each card is one pass. A hole lets a needle through, the hook stays in play, that warp thread rises. Solid card pushes the needle back, that thread stays down. The chain advances, the next card is the next line of the picture. You can run the same chain tomorrow and get the same pattern. You can swap the chain and weave a different one. A person still throws or watches the shuttle, still tends the web, still punches the next set of cards when the design changes. The drawboy’s particular job shrank. A card-cutter’s job appeared. Adoption in Lyon was fought over. Some heads were broken. The mechanism spread anyway, unevenly, because a figured silk that had needed a second body at the top of the loom could be woven without him.

The skill moved. It did not leave

A jenny spinner, a mule spinner, a piecer, a loom tuner, and a card puncher are different jobs. Mechanization picked some of them up and dropped others. In much of Lancashire the big mule became a man’s machine, with piecers under him. That was a custom of the trade as much as a fact about gears. Women’s and children’s work did not disappear. It moved to the frames, the ends, and the bobbins. The pressure on handloom weavers was not a side effect. It was the market working on a skill the power loom was late to match, and then, slowly, did match for more and more cloths.

Hours, fines, and the factory bell are on the factory page, with Thompson’s argument and the caution that goes with it. Output went up. The people who stood the frames paid for that in the day, not in a footnote. A history of these machines that only counts spindles is a sales sheet.

Where this sits in the history of automation

A mill now can run rings or rotors, watch a break with a sensor, and stop a spindle without stopping the room. Setup, yarn quality, a smash, and a change of cloth are still human calls in more sheds than the brochure admits. I am describing equipment a plant can buy. I am not describing every plant.

The kinship is the chain, not the electronics. Speed up spinning and weaving has to answer. Share one wheel and a fault is shared. Store a pattern outside the weaver’s memory and the pattern can be repeated by somebody who did not design it. The break is the closed loop and the program: a measurement that can change the machine, and a design that can be edited without cutting a new card by hand. An eighteenth-century frame did not have the first. A Jacquard head is a distant cousin of the second, and only for the lift of the warp. Do not call a mule a robot because it spins while a man watches the ends.

The long use, without the straight line

Later sheds answered limits these machines actually had. A self-actor answered a carriage one skilled man could only run so fast. A better loom answered a warp the 1785 patent could not manage. Ring spinning, much later, answered a mule that still wanted a craft around it. Electricity answered the shared belt. Each one showed up when a builder could make it and a master would pay. None of them was hidden in Hargreaves’s wheel waiting for a caption.

Textile machinery earns its place because it was the first widespread industrial case of copying a hand motion into a powered frame, and then, on the Jacquard head, of storing a choice in a card. It does not earn the claim that automation began here, or that the information age began in a silk shed. The spindle does not care what we call it. It only turns while someone keeps the end up.