Straw has to learn the curve

The braid goes in under the presser foot almost straight. On the other side a spiral begins to come out. At first the change is minimal: one band overlaps the one before it, a line of stitching holds it there, and the disc grows in diameter turn after turn. Then the surface has to stop lying flat. The braid has to lift, build the top of the dome, come down the side and, at the point the model fixes, change direction again to open into the brim.

The material never changes. The geometry does. On the inside of every curve the braid has to travel a shorter distance, on the outside it has to follow a wider circumference, and if it is guided at the wrong tension the surface puckers, throws folds or opens up. The machine goes on sewing: it does not know whether the spiral is becoming a hat or a perfectly regular mistake. It is the operator’s hand that decides how much braid to feed, how much to overlap, when to increase the diameter and when to start closing it.

Straw is born straight. The craft begins when it has to learn the curve.

A wheat grown so as not to become bread

The modern history of Florentine straw is generally dated to Signa in 1714, when Domenico Michelacci introduced a crop of the wheat variety known as grano marzuolo, grown not mainly to produce food but to produce straw for plaiting. The agronomic principle was simple and radical. The wheat was sown very thickly: the plants, forced to compete for light, put out long thin stems, and shallow furrows made it possible to pull them up before the ear had finished ripening and before the culm had grown too thick and stiff.

The plant was harvested at exactly the moment when it would still have been unfit to produce mature grain. That straw was not the waste left over after an ordinary harvest: it was the result of a crop designed around the characteristics of the stem — length, fineness, flexibility, colour and uniformity. To become material for a hat, the wheat had to stop becoming food ahead of time.

Historic Florentine straw did not come about because someone had found an elegant use for what was left over in the field: it came about because a whole field was organised to produce fibre, and the raw material already contained an industrial decision. Before the braid, before the block and before fashion, there was a plant grown so that its most useful part would not be the grain.

Quality begins in the space between two nodes

A wheat stem does not have the same diameter and the same elasticity along its whole length. It is interrupted by nodes, and the sections between one node and the next — the internodes — have different characteristics. Historic processing selected above all the finest and most regular parts: the ear, the nodes and the less suitable sections were removed, and the resulting strands were graded by size and quality, bringing together materials similar enough to be plaited together.

It is work less visible than the plaiting and far more important than it looks. A braid made of strands with different diameters does not react uniformly to tension: a thicker stem tends to dominate the structure, one too fine can snap or leave a weak zone. Colour has to be watched as well, because natural straw is not an industrial beige that is always the same, and it carries differences that come from the plant, the soil and the drying. If the variations are distributed with intent they can build depth and movement; if they gather in one place by chance, they read as a stain.

The uniformity of the hat does not begin at the seam. It begins in the selection of elements that, on their own, are not uniform.

After the harvest the straw was dried and exposed to light, and the historical sources also record whitening with sulphur fumes and, in later periods, other bleaching systems. This is not the romantic formula of Tuscan sun producing a golden fibre by itself: the drying has to reduce the moisture without making the stems brittle, and the exposure can change the colour without cancelling every difference. Straw that looks dry, moreover, goes on answering to its surroundings: too dry, it becomes stiffer and can break as it is folded; too damp, it can distort. To be plaited and blocked, the material has to be in an intermediate condition, flexible enough to accept the gesture and stable enough to keep it. The hatter is not working dead matter, but a fibre that responds to water, heat and pressure long after it has left the field.

The braid is a structure, not a decoration

A straw braid can look like a ribbon. It is not one. Every strand passes over and under the others in a precise sequence, the tension spreads the forces across the width, the edges have to stay regular and the pattern has to repeat without producing sudden narrowings.

Historically Florentine braids were made with different numbers of strands: braids of seven, eleven and thirteen ends are documented, among others. The number, though, is not on its own a ranking of quality. What counts is the fineness of the straw, the regularity of the elements, the type of plait, the width obtained and what the semi-finished material is destined for. A wider braid covers more surface at every turn and makes its own structure more visible; a finer one allows smaller curves and details, but takes many more passes to build the same volume.

Density changes the behaviour too. A braid that is too loose can distort during the sewing, one that is too tight can resist narrow curves. The plait therefore has to hold regularity and availability to movement at the same time: stable enough to take the machine, flexible enough to change direction without opening.

The braid does not cover a structure that already exists. It is the structure.

An industry built inside houses

For centuries an essential part of the supply chain did not happen inside large factories, but in homes, on doorsteps, in stables, in courtyards and in the gaps left by farm work and housework. The plaiting was done above all by women: the trecciaiole, the straw plaiters, received the prepared straw and produced long braids that would then pass to merchants, carriers, sewers and manufacturers. The work could be done while children were watched, while walking, or alongside other occupations.

This apparent compatibility with domestic life has often been told as a form of freedom. The house did not remove economic discipline: it moved it outside the factory. The braid had to meet a length, a width and a quality, it was paid by the piece — cottimo — and the price could fall when the market went into crisis, when supply rose or when the merchants transferred the risk to the workers. Plaiting fast for many hours was not a traditional pastime: it was a skill set inside an international supply chain. That it was practised at home and almost only by women did not make it less industrial. It made it easier to underestimate.

In May 1896 the trecciaiole of the Firenze area mounted a broad mobilisation. The crisis in the sector, international competition, overproduction and the devaluation of women’s work had cut earnings to levels barely possible to live on. The workers blocked roads, stopped transport and brought into public space a craft that the organisation of production had kept scattered through the houses. It was not an edifying legend about humble, courageous women: it was a conflict over the economic value of the work.

The trecciaiole were not asking for their skill to be admired as cultural heritage. They were asking for it to be paid. It is a distinction that stays essential, because celebrating a craft without discussing the conditions that let the people practising it live from it means keeping the product and losing the person. The braid does not only tell of knowledge handed down: it tells who had the power to set its price.

Sewing a spiral means changing measurement all the time

Traditional construction with braid can begin at the centre of the tip. The first turns produce a small, almost flat surface; as the diameter grows, the operator has to decide when the spiral should bend and build the volume of the dome. There is no universal line at which every hat changes direction: it depends on the model. A low wide dome needs a different expansion from a tall narrow crown, and a flat boater is not built the way a cloche is.

The machine sews the new braid to the braid already mounted; the operator controls the relation between the two. She can vary overlap, angle and tension by small amounts, guiding the material as the surface goes from horizontal to curved and then almost vertical. At the point where the crown has to become brim the direction changes again and the spiral has to open: if the transition is forced a sharp fold appears, if it is widened too early the crown loses height, if it is made too late the brim starts from the wrong circumference.

An early error, moreover, does not stay where it was born. A small difference in width changes the way the next turn rests on the one before; too much tension can start to close the diameter. Every new turn inherits what it finds, and a millimetre repeated over many circumferences can become a visible deformation. That is why the hand guides the braid at the point of the stitch while the eye watches the volume as it grows: whether the centre is still centred, whether the dome is rising symmetrically, whether the brim is opening without twisting.

The craft consists in seeing the error while it is still small enough to be corrected. Not when the spiral has already turned it into the final shape.

The thread is a variable as well. It has to stop the spiral opening during blocking, use and maintenance, without stiffening the surface: if the tension is too low the braids can separate and leave gaps, if it is too high the stitching compresses the fibre, ripples the edge and changes the curvature. The position of the stitch counts too: in some hats the seam belongs to the design, in others it has to all but disappear. Quality does not consist in the absolute absence of construction marks, but in every mark having a function compatible with the project. The machine measures the stitch. The craftsman judges the hat.

Not everything called straw begins as a braid

The word straw brings together different constructions. Some hats are made by sewing a long braid in a spiral; others start from a cone, a hood or a capeline already woven as a single body, and the semi-finished piece is damped, heated or steamed before being blocked over a hat block. Raffia, sisal, buntal, hemp, paper fibres, yarns and synthetic materials can also be plaited or turned into semi-finished bodies for hat-making.

They do not all react in the same way: elasticity, strength, capacity to absorb moisture, memory of shape and response to stiffeners all change. A fine wheat braid can bend elegantly and snap if it is forced when too dry; a palm fibre can take a different treatment; a paper braid can be surprisingly regular and answer to water according to characteristics of its own. A competent hatter does not apply the same cycle to whatever the trade calls straw: first he has to understand what he really has in front of him.

The word “natural”, in any case, says nothing about how strong the material is, nor whether it has been dyed, stiffened or treated for greater stability, and it does not on its own demonstrate a lower environmental impact: cultivation, bleaching, dyeing, transport, waste and durability are part of the material history as much as the plant origin is. In the same way a synthetic material is not necessarily without technical quality, and can offer uniformity or performance useful to a particular model.

The right question is not whether the hat is natural enough to deserve trust: it is whether the material has been clearly identified, worked consistently and chosen for a real function. The hatter does not sell a moral category. He builds an object that will have to take sun, heat, sweat, handling and travel.

The block makes final whatever it is handed

After the sewing, or after the semi-finished body has been prepared, the hat is blocked. Blocks can be wood, metal or other materials able to take pressure, moisture and heat, and they define the height, the angle and the volume of the crown, while other formers control the brim. The material is made temporarily more available with moisture, steaming or heat, then pulled, fitted and pressed onto the block; as it dries or cools it tends to keep the geometry it has reached.

It is a decisive stage. It is not a resurrection. A block can correct small differences, regularise a surface and define the model, but it cannot give continuity back to a broken braid, remove a seam that has puckered the whole crown, or make badly chosen material uniform.

If the body is too small, forcing it can open the plait or weaken the fibre; if it is too large, the compression can produce folds and gathers. If the stiffener is excessive the material can go rigid without really following the block; if there is too little the hat can lose its geometry quickly. The block makes final whatever it is handed. It does not automatically make it right.

Every block keeps the head it was designed for

The blocks kept in the hat factories do not serve only to produce different models: they contain measurements, periods and different ways of imagining the relation between a hat and a body. A tall dome changes the visible proportions of the face, a wide crown changes the point at which the hat rests, an angled brim can shade the eyes, hide part of the face or oblige the wearer to change posture.

Inside the same model different sizes are then needed, and the circumference of the head is the first figure but does not describe the whole shape: two people can have the same measurement and a different relation between length and width, and a perfectly circular hat can press on the forehead and leave space at the sides. Every block is a tool and is also a technical memory: it keeps the head someone had designed the hat for, even when that person no longer exists.

Quality is often judged by looking at the hat from the front, but the body judges what it cannot see as well. The inner edge has to rest without creating pressure points, and a wide brim must not turn every movement of air into an uncontrollable shift. Where the sweatband is sewn changes the fit too: if it pulls it reduces the size and can distort the crown, if it is left too loose it shifts and loses its function.

The hat must not be immobile: it has to move with the person without constantly losing its place. The craft does not end when the object looks right on the block. It ends when that shape can be inhabited.

The brim amplifies the error

The brim moves away from the crown, and the further it grows the more visible every small asymmetry becomes. A minimal difference at the starting point can turn into an obvious variation along the outer edge; a slightly stiffer area lifts while the opposite one drops; an error of tension can produce waves that no decoration will really manage to hide.

The brim has to have enough structure to keep its profile, but it does not have to be rigid: a soft model may call for controlled movement, a boater has to hold a far sharper line, a wide brim meant to shade from the sun has to carry its own weight without becoming a plank. Stiffeners, stitching and the thickness of the material all contribute to the result: more stiffener does not mean more quality and can produce a hard or brittle surface, less structure does not mean more authenticity and can turn a designed brim into a shape that changes with every use. The hatter has to decide which movement belongs to the hat and which, instead, shows it giving way.

The brim is the part furthest from the centre. It is also the part where the centre reveals whether it was built correctly.

Ribbons, bows, flowers, veiling and feathers belong to millinery and can move the same body from everyday use to a ceremony. They should not, though, serve to hide the construction: a ribbon can cover the seam in the zone where crown and brim meet, but it does not stabilise a badly made transition, and a flower can draw the eye away from an asymmetry without removing it. A band that is too tight can distort the crown, a decoration that is too heavy can drag the hat down. The milliner works on a structure already built and has to change its language without compromising its function: decorating does not mean adding, it means knowing how much the hat can still take.

The cappello di Firenze is not one single raw material

The historic fame comes from the wheat straw grown and worked in the Florentine territory — the cappello di paglia di Firenze, the straw hat of Firenze — but contemporary manufacture is wider. The firms of the district work braids, fabrics, yarns, felts, leathers and plant fibres that also arrive from other parts of the world, and they produce hats, accessories and semi-finished bodies for their own brands or for international maisons. Saying cappello di Firenze should not therefore suggest that every fibre was grown at Signa: it can indicate the place where design, choice of materials, sewing, blocking and finishing happened.

It is a distinction that does not weaken the territory. It makes it more credible. A manufacture can hold a very strong local identity while working materials that come from international supply chains, as happens in jewellery, in leather goods and in tailoring. The problem begins when the geographies are confused.

The name Panama has produced one of the most persistent misunderstandings in hat-making. The hat is traditionally woven in Ecuador with fibres taken from the toquilla palm: crown and brim grow progressively out of the weaving, not out of the sewing of a Florentine wheat braid, and the quality depends on the preparation of the fibre, the fineness of the weave and the time needed to build the volume. The international name does not describe the place of production, and still less does it license calling any pale wide-brimmed hat a Panama.

A Florentine hat factory can buy a genuine toquilla body, select it, block it, dye it or finish it to its own design. That work has value, but it does not turn the fibre into Tuscan straw and it does not make Florentine the knowledge with which the body was woven. Luxury has no need to simplify provenance in order to seem more precious: it should be confident enough to declare how many skills and how many territories it contains.

The sample is simpler than the production run

A hat made as a prototype can receive continuous attention: the material is chosen from many, every step is watched, a small irregularity can be corrected at once. When the model is approved the problem changes. The same identity has to be reproduced across a larger quantity of natural materials that will never be perfectly alike: the colour can vary from one batch to another, one braid can have slightly different elasticity, a woven body can meet the block with a different tension. Production must not pretend these differences do not exist: it has to establish how far they can be accepted without moving the product away from the sample, and to tell a natural variation from a production deviation.

Two hats do not have to be indistinguishable: they have to be recognisable as the same model, offer a compatible fit and keep the characteristics promised. Craftsmanship does not consist in producing something different every time. It consists in carrying a real difference through without losing the project.

Contemporary manufacturers use modelling software, electronic cutting machines, dies and presses for this, and the presence of these tools does not turn the hat into an impersonal product. Software makes it possible to check measurements and proportions before cutting, but it does not decide whether a brim is too heavy for that material. An electronic machine repeats a profile with great precision, but it does not always recognise a difference in elasticity that makes part of the material less suitable.

The specialised braid machine holds the stitch, the feed and the speed: it is still the operator who guides the volume growing around the needle. Technology reduces some of the swings; it does not remove the decisions. And it works in the opposite direction as well: among the sewing machines, the irons and the presses that have passed through many generations in the hat factories, an old machine does not automatically make the hat it produces better, because it can offer a tactile control still worth having or be worn out and unsuited to a recent material. A tool is not to be judged by its age, but by the quality of the decisions it lets you carry out.

The hat goes on reacting after it is sold

The shape does not become unchangeable when the hat leaves the workshop. Plant fibres go on absorbing and releasing moisture, light can change their colour and prolonged pressure can distort crown and brim. A hat crushed inside a suitcase does not return to its original shape simply because it was sold as a quality product: it depends on the material, the construction and the type of blocking. Some models are designed to be folded or rolled and others are not, and the commercial label “packable” should describe a real characteristic, not an invitation to mistreat any straw hat.

Water calls for care as well. Light rain may leave no permanent consequence on a material treated properly; soaking the hat through can dissolve the stiffeners, change the tensions and leave distortions as it dries. The direct heat of a radiator or a hairdryer can speed the drying unevenly and make some fibres more brittle. Maintenance therefore has to follow the maker’s instructions and the actual nature of the material, because the word straw is not enough to establish a universal method.

A good hatter does not hand over a shape alone. He hands over the information needed not to destroy it.

A seam can open, a ribbon wear out, a brim distort, and not every damage requires replacing the object. But an effective repair has to respect the way the hat was built: whoever works on a spiral-sewn braid has to find the point where the continuity broke and work without pulling the rest of the surface, and whoever reshapes it has to know the material, the stiffener and the temperature it can take. A patch applied over a break can hide the mark and stiffen the area, moving the stress a few millimetres further along.

Repairing does not mean bringing the hat back to an imaginary condition of absolute newness: it means giving it back enough structure and function to go on being used. Durability does not depend only on the strength it started with. It depends on the object having been built in a way that can be understood, and on someone still existing who is able to work on it.

Keeping the museum does not automatically keep the craft

The Museo Civico della Paglia di Signa, the town’s straw museum, keeps hats, braids, sample books, blocks, sewing machines and photographs, and makes visible a supply chain that the finished product tends to compress into a single elegant object. It is indispensable. It is not sufficient. A museum can show how the straw was plaited and tell who the trecciaiole were, but it cannot guarantee that a young worker today finds the time, the income and the training needed to learn to guide a braid under a specialised machine.

It can keep a wooden block. It does not create a firm able to design a collection, develop a sample, calculate the cost and sell it on an international market. Transmission takes working workshops, orders, schools, machines kept in repair, people willing to teach and apprentices who can afford to stay long enough to become competent. A craft does not survive merely by being remembered: it survives when it goes on solving problems for clients willing to pay what that competence costs.

In July 2026 the Comune di Signa, the town council, announced a project to recover and multiply a seed held to correspond to the grano marzuolo historically grown for straw. It brings attention back to the agricultural link the whole manufacture came out of, and it tries to give the territory back a material that for more than a century has no longer been grown in the form the sources describe. But a seed is not yet a hat.

First it has to be verified how the variety behaves in today’s soils and climate, the available quantity has to be multiplied, the time and the density of sowing defined, the harvest and the drying organised. Then people able to plait that straw are needed, and firms willing to try it, to compare it with the semi-finished materials already in use and to build products whose price can carry every stage. Recovering a plant genetically is different from recovering a supply chain economically: the first concerns researchers, agronomists and farms, the second requires field, manufacture and market to recognise one another again.

The project will really have rebuilt something when that straw is not merely collected in a museum or used for a few commemorative pieces. It will have to be able to become work.

The point where the line becomes volume

The contemporary Florentine hat is not produced exactly as it was in the eighteenth century. The historic raw material has become rare, the supply chain has opened to international fibres and semi-finished bodies, the machines have changed, fashion has altered proportions and functions. This transformation does not prove the craft has disappeared: it proves the craft has had to go on choosing which gestures to keep, which tools to update, which materials to take in and which promises to stop making.

Tradition becomes false when it describes as unchanging something that has changed, and it becomes useless when it refuses every change for fear of losing its own image. The real continuity lies elsewhere: in the ability to read a fibre, in the control of tension, in the relation between model and head, in the point where a surface has to begin to curve. These are the skills that connect a nineteenth-century trecciaiola, a machinist at a specialised sewing machine and a pattern-maker working today on a three-dimensional design. They are not doing the same work with the same means: they are facing the same problem, giving a flexible material a shape that looks inevitable.

At the start there is a stem, then a set of strands, then a long, almost flat braid. The hat begins when that line has to occupy space. Every turn changes its relation with the one before: the tip widens, the crown descends, the brim opens. Moisture, heat and pressure fix what the sewing has prepared, and the sweatband adapts the shape to a real head.

When the work succeeds, the construction disappears. We do not see a strip forced to change direction: we see a hat, and the curve seems to belong to the material by nature. It does not. It was negotiated turn after turn, stitch after stitch, in the knowledge that a fibre too dry would have snapped, a wrong tension would have distorted the volume and an unsuitable block would have turned precision into discomfort.

Straw does not already contain the hat. It contains the possibility of bending without breaking. The craft is everything needed to recognise that possibility and carry it far enough for it to stay.

Straw is born straight. The hatter builds the point where it can become curve.