The brailler that has not needed redesigning since 1951

Six keys, one for each dot

A Perkins Brailler has a key for each of the six dots of the braille cell, plus a space key, a backspace key and a line-space key. You press the keys for a cell together, the way a chord is played, and the machine embosses that cell in one stroke. Two knobs on the sides move the paper; a lever above the keys returns the carriage.

The detail that matters is what the rollers do. They are grooved, so that paper already carrying raised dots can pass back through the machine without the dots being flattened. That is what makes the brailler correctable: a page can be reinserted and a single dot added to a specific cell without damaging any of the work already on it.

Everything moves inside the case. There is no carriage sticking out of either side, because the embossing head travels across the paper rather than the paper travelling past a fixed head — which is why the machine survives being knocked off a desk.

What came before

Braille itself dates from the 1820s, but for most of the nineteenth century writing it meant a slate and a stylus — pressing dots one at a time, right to left, so they read left to right when the sheet is turned over. It is precise and portable and slow.

Machines were attempted throughout the second half of the century, and none of them lasted. The first that found general acceptance came in 1892, from Frank H. Hall, Superintendent of the Illinois School for the Blind. Hall’s machine was shaped like a typewriter, with the six keys mounted on the front and a roll-up carriage that took eleven-inch paper. That layout — six keys, one per dot — is the one every mechanical brailler has used since.

Perkins School for the Blind built its own braillewriters in the early twentieth century, and they were not good. Made individually rather than to a tolerance, they needed frequent repair, they were noisy and expensive, and their cast-iron frames broke when dropped. In 1931 a new director, Dr Gabriel Farrell, ordered Howe Press to stop making them altogether — while intending that Perkins would one day produce a machine worth having.

The man who designed it

David Abraham was from Liverpool, had served in the Royal Flying Corps in the First World War, and had worked in his family’s stair-rail business, where he built machines for turning wooden rods. He moved his family to the United States just as the Great Depression began, and took what work he could find.

In the early 1930s that meant a road crew on Charles River Road, which runs past Perkins School. He saw the sign, walked in, and was hired as an instructor in the Manual Training Department. His skill as a machinist was noticed by his supervisor, who suggested to Farrell that here was the person to design the brailler.

Abraham worked on it in his spare time, in his basement, for years. The specification was drawn up with Dr Edward J. Waterhouse, then assistant manager of Howe Press. In 1941 he presented the prototype. Perkins’ own history of the machine describes it as essentially identical to the brailler still being produced half a century later.

What the specification asked for

Abraham and Waterhouse wrote down what would make the best machine available, and the list is unusual in how much of it is about the person using it rather than the engineering.

• A light touch — so that very young children, and people with little strength in their hands, could use it without strain.
• Quiet. Braillers of the period were not, and a classroom of them was punishing.
• Hard to break, and quick to load paper into.
• Spacing accurate enough to reinsert a page and add one dot to one cell without disturbing anything else.
• Paper that does not fall out when the operator reaches the bottom of the sheet.

Abraham managed all of it, and said of the result that it had more precision parts than a wristwatch. That claim is the whole story of the machine: simple to operate meant complicated to construct, and the complication is what made it last.

Getting it built

The prototype arrived in 1941 and the war took the materials, so nothing could be made. When it ended, the Perkins trustees agreed to subsidise a run of one thousand braillers — with no evidence yet that anyone would buy them.

It took five more years to reach the public. The Howe Press workshop had to be moved from cramped quarters in South Boston out to the Watertown campus; tooling to Abraham’s tolerances added years; post-war materials were scarce; and Abraham himself held production up, unwilling to settle on a method until he was satisfied it was the best one.

Waterhouse spent 1946 to 1950 touring conferences with the prototype, showing it mainly to teachers and students. By the time manufacturing began there were 1,500 orders waiting — more than the run the trustees had authorised. Getting to that point had consumed more than half of Howe Press’s capital endowment.

The machine went on sale in 1951, and the press could not make them fast enough.

Still made

Seventy-five years on, Perkins still sells mechanical braillers, in several variants: the standard machine, a large-cell version, a unimanual model for one-handed use, an electric model, and the SMART Brailler, which added a screen and audio feedback when it appeared in 2012. A 2008 revision of the classic machine made it lighter and quieter and added an erase key and a carrying handle.

The reason a mechanical device survives alongside refreshable displays and embossers is not nostalgia. It needs no power, no software and no updates; it is quiet enough for a classroom and tough enough for a school bag; and it gives a child learning to write braille the same immediate, physical relationship to the page that a pencil gives a sighted child. What comes out is the same code either way.

Where this comes from

The dates, names and figures on this page are taken from History of the Perkins Brailler, published by Perkins School for the Blind, and from Perkins’ own description of the models it currently sells. Where this page says a thing was claimed — the wristwatch comparison, for instance — it is because Perkins records it as a claim rather than a measurement.