Why clothes changed shape

Pick up a shirt and turn it inside out. Count the seams. There are probably fifteen or sixteen of them, most joining pieces that aren't rectangles, several of them curved, and at least one — the armhole — sewing a curve to a different curve without puckering.

Now try to imagine cutting and stitching that by hand, from cloth you also made by hand, for one person.

Clothing history usually gets told as a parade of styles, with dates and designers and a vague sense that people simply wanted different things at different times. That's the least useful version. Almost every large change in the shape of clothes was forced by something duller: the width of the cloth coming off the loom, the properties of the fibre, the cost of a seam, or the invention of a fastening. Taste moved into the space the technology opened, and not usually before.

The loom decided the pattern

On a hand loom, the weaver throws the shuttle across the warp and catches it on the other side. That means the cloth can't be wider than a person's arms can comfortably reach — realistically somewhere between about 70 and 90 centimetres, and often much less.

Cloth was also enormously expensive. Spinning alone took vastly more labour than weaving, and weaving took more than sewing. So the sensible thing to do with a narrow, precious strip of fabric is to use all of it, in straight lines, wasting nothing.

Which is exactly what most of the world's traditional dress does. The sari is a single uncut length. The Greek chiton and the Roman toga were draped and pinned rather than tailored. The kimono is cut from one bolt of cloth, the tanmono, roughly 36 centimetres wide and about twelve metres long, into straight rectangles and near-rectangles with essentially no waste — and it can be unpicked flat for washing and re-sewn. Tunics, kaftans, ponchos and smocks all work the same way: rectangles joined at the edges, with the fitting done by belts, folds and pins rather than by scissors.

These aren't primitive garments. They're the correct engineering answer to expensive narrow cloth.

John Kay's flying shuttle, patented in 1733, let one weaver knock the shuttle across a much wider warp with a cord instead of passing it by hand. Power looms followed from the 1780s. Cloth got wider and, over the next century, absurdly cheaper. Once a metre of fabric costs an hour of wages rather than a week's, you can afford to cut curved pieces and throw the offcuts away — and curved pieces are the only way to make cloth follow a three-dimensional shape closely.

Wool does something no other fibre does

The European tailored suit exists because of a property of sheep. It isn't an aesthetic story.

A wool fibre is covered in overlapping scales and has a natural crimp. Add moisture and heat and press it, and the fibres shift and lock — you can shrink one area of a flat piece of cloth and stretch another, permanently, and the fabric will hold a curve it was never cut into. Tailors call it shaping with the iron, and it's the whole basis of the trade. A jacket chest is moulded, not just cut. The roll of a lapel is built by hand-stitching a canvas interlining with hundreds of small stitches that hold a curve into the layers.

Nothing else really does this. Linen creases sharply and refuses to be moulded, so linen garments are loose and flat and always look slightly rumpled, because they are. Cotton drapes softly and takes dye brilliantly, which is why printed cotton conquered the world in the eighteenth century, but it won't be sculpted. Silk is strong and fine and drapes beautifully and does what gravity tells it.

So the shapes each culture could make were constrained by which animal or plant was to hand. Heavy fitted outerwear belongs to wool regions. Draped, layered, air-moving construction belongs to cotton and linen regions, where the job is shade and evaporation rather than insulation.

Curved cutting, and the invention of fashion

Something changes in Europe in the thirteenth and fourteenth centuries. Garments stop being draped and start being fitted — cut with curved seams, closed with rows of buttons or spiral lacing, shaped to the body underneath.

Buttons had existed for ages as ornaments. The buttonhole is the actual invention, and it arrives in Europe around the thirteenth century. A worked buttonhole plus a button gives you a fastening that's adjustable, repeatable and strong, which means a garment can be pulled tight and still be got into.

What follows is the first recognisable fashion cycle: shapes changing fast enough that clothes went out of date, sumptuary laws being passed to stop the wrong people wearing the right things, and moralists complaining. Once garments are cut rather than draped, the number of possible shapes explodes, and there's suddenly something to have opinions about.

The sewing machine changed the arithmetic of a seam

Barthélemy Thimonnier had machines sewing French army uniforms around 1830 and a mob of tailors destroyed them. Elias Howe patented a lockstitch machine in the United States in 1846. Isaac Singer's 1851 version, with a straight needle and a foot treadle that left both hands free for the cloth, was the one that worked in practice — and the patent fights got so tangled that in 1856 the main holders pooled their patents rather than keep suing each other.

The lockstitch mechanism is genuinely clever. Two threads, one from a needle above and one from a bobbin below, are locked together in the middle of the fabric by a shuttle or hook catching a loop of the upper thread and carrying it around the lower supply. Hand sewing can't produce that structure at all.

A hand-sewn seam runs at something like twenty to thirty stitches a minute. Early machines managed several hundred; modern industrial machines do thousands. The immediate consequence is that seams became almost free.

Watch what that does to the shape of clothes. More seams means more shaping, so garments could fit closely without heavy structure. Topstitching, flat-felled seams and finished edges got cheap. Ruffles, pleating and applied trim — previously the mark of someone who could pay for hundreds of hours of hand labour — appeared on cheap dresses within a decade, and promptly stopped signalling anything. Elaborate decoration went downmarket almost overnight, and by the end of the century plainness was doing the status work instead.

Ready-to-wear followed. Military contracts came first, because armies need thousands of identical garments quickly — and the measuring done for uniform contracts in the 1860s produced the first large body of statistical data on human dimensions, which is where commercial sizing came from.

Sizing has never really worked

Bespoke tailoring measures one person. Ready-to-wear has to guess, in advance, in bulk, and the guess has to be wrong for almost everyone by some amount.

The first serious attempt at a scientific answer was an American study around 1939–40 that measured roughly 15,000 women, and it fed into a voluntary commercial standard published in 1958. That standard was withdrawn in 1983, and sizing has drifted freely ever since — which is why a label number tells you very little across two brands, and why the same number has quietly come to mean more centimetres than it used to.

This is a measurement problem, not a people problem. Human proportions vary along many independent dimensions at once, and any sizing system collapses all of them onto a single number or two. No arrangement of numbers can survive that, and the shops solved it instead with fit models, stretch fabric and generous returns policies.

Fastenings, and the ones that changed silhouettes

A garment can only be as close-fitting as its opening allows. If you can't get into it, the cut doesn't matter.

• The zip. Whitcomb Judson showed a version in 1893 and it kept jamming. Gideon Sundback's 1913 redesign, with interlocking scoops on a tape, is the one that works. The name came from B.F. Goodrich in 1923, who used them on rubber galoshes and liked the noise. Men's trousers took it up in the 1930s. It gives you a strong opening that closes flat and disappears, which no button placket can do.
• Elastic. Vulcanised rubber from 1839 made elastic threads possible, and they went straight into waistbands and cuffs and stayed there.
• Velcro. George de Mestral got the idea from burdock burrs stuck to his dog and patented it in the 1950s. Loud, but it works one-handed.
• Elastane. Invented by Joseph Shivers at DuPont in 1958 and sold as Lycra from 1962, and this is the big one.

Elastane deserves the emphasis. Before it, close fit had to be cut and constructed, which meant darts, panels, boning, linings and a fitting. A stretch fibre blended into ordinary cloth at a few per cent lets a simple garment conform to a shape it wasn't cut for, and lets the same garment fit a range of people. Every close-fitting cheap garment of the last fifty years depends on it — jeans that move, jersey dresses, anything sold in four sizes to cover everybody, and the whole category of sports clothing.

It's also why a modern pair of jeans and a 1960s pair look similar and feel nothing alike.

Synthetics and the pleat that doesn't come out

Rayon arrived at the end of the nineteenth century — regenerated cellulose, sold as artificial silk, cheap and drapey. Nylon, developed by Wallace Carothers at DuPont and launched in 1939, was the first fully synthetic fibre: strong, fine, and famous for stockings before the war took the entire supply for parachutes. Polyester followed from British work in 1941 and went into mass production in the 1950s.

The property that matters most is thermoplasticity. Synthetics soften with heat and set as they cool, which means a pleat, a crease or a shape can be heat-set into the fibre itself. A pleat in wool is a temporary arrangement that a rainstorm undoes. A pleat in polyester is a permanent change to the material, and it survives the wash. Permanent-press clothing is that one property, sold.

The other effect was knitwear. Jersey and other knits are made of interlocking loops rather than crossed threads, so they stretch in every direction by geometry alone — the loops just change shape. Woven cloth barely stretches at all except diagonally. Cheap synthetic and cotton-blend knits gave the twentieth century the T-shirt, the sweatshirt, the polo shirt and the jersey dress: garments with almost no shaping in the cut, which fit anyway because the fabric moves.

Before that, the only way to get give out of a woven fabric was to exploit the diagonal. Madeleine Vionnet built a career on it in the 1920s and 30s, cutting on the bias so the cloth's natural diagonal stretch let a dress follow a form with hardly any seams. It uses a great deal more fabric and is difficult to sew straight, which is precisely why it read as expensive.

The shape was usually built underneath

For several centuries the outline of European women's dress was produced by a substructure, and the visible garment was more or less a cover for it. Change the engineering underneath and the whole silhouette changes in a season.

The cage crinoline, patented in 1856, is the clearest case. It replaced half a dozen heavy stiffened petticoats with a light frame of steel hoops on tapes, and it was cooler, lighter and easier to walk in than what it replaced. It also made the skirt enormous, because now the width cost nothing. The bustle did the same trick a couple of decades later by moving the structure to the back. When structural underwear went out in the 1920s, the outer garments went straight and loose within a few years, and when structured foundations came back after 1947 the outer shapes tightened again immediately.

The rule holds for men too, quietly. A tailored jacket's chest is a piece of shaped canvas with wool laid over it.

Status, and the deliberate uselessness of good clothes

Running underneath all of it is a trade-off between what a garment can do and what it says.

An enormous amount of high-status dress has been designed to demonstrate that the wearer doesn't do physical work. Trailing sleeves you couldn't lift anything in. Starched white collars and cuffs that show a single day's dirt. Powdered wigs. Long fingernails. Shoes for surfaces that are always dry. The signal is expense, and impracticality is a reliable way to prove expense because it can't be faked cheaply.

Sumptuary laws were the attempt to enforce the code by statute, and they were passed all over the place — restricting furs, particular dyes, silk, gold thread and specific garments to particular ranks. They almost never worked. Edo-period Japan restricted merchants from displaying silk, and merchants responded with plain, correct outer kimono lined with spectacular silk on the inside, visible only at the cuff and to anyone who mattered. Rules about surfaces produce hidden luxury.

Then men's dress did something abrupt. Around 1800, European male clothing dropped colour, embroidery, lace and ornament almost entirely and settled into dark wool with sober cut — a shift the psychologist J.C. Flügel later labelled the Great Male Renunciation. What replaced court dress was English country and riding clothing, cut close, made in fine plain wool, with the whole display transferred from decoration to fit and cloth quality. Beau Brummell became famous for clothes that were conspicuously undecorated and impossibly well made. The modern suit is riding kit that never went home.

Work put things back in

The other direction produced most of what people actually wear now.

Pockets are a good example of the trade running against practicality. Women's pockets used to be separate bags tied around the waist under the skirt, reached through a slit, and roomy. The slim high-waisted styles around 1800 left nowhere to hide them, and they were abandoned in favour of a small handbag that had to be carried. A storage problem was solved by removing the storage, and it never fully came back.

Jeans went the other way. Levi Strauss supplied the cloth and Jacob Davis had the idea; their 1873 patent is specifically for copper rivets at the points where pocket corners tear under strain. That's not styling, it's a repair of a known failure mode, and the rivets are still there on trousers that will never carry a tool.

War did the rest. Trench coats, flight jackets, cargo pockets, chinos, duffel coats, aviator sunglasses and the T-shirt itself all came out of military specification, where somebody had to justify every seam, and then walked into civilian wardrobes when the surplus hit the shops.

What actually sets the shape now

Paper patterns, graded to multiple sizes and sold by the envelope from the 1860s, let home sewers copy fashionable cutting. Mail order catalogues spread the same shapes across whole countries. Container shipping and offshore manufacturing then took the cost of a garment down to a level where the constraint stopped being cloth or labour at all.

What's left constraining shape is mostly the machinery. A garment that can be assembled by a standard sequence of machine operations is cheap; one that needs hand shaping isn't, which is why a properly canvassed jacket costs what it does and a fused, glued one doesn't. Knits are cheaper to fit than wovens. Stretch is cheaper than cutting.

Look for the seams that don't do anything and you'll usually find a piece of history. The buttons on a jacket cuff that don't undo. The vestigial belt on a coat that was designed for a horse. And the narrow panels joined down the length of a sari border, still carrying the width of a loom nobody uses any more.