“Cunning of Hand”: Joseph Moxon’s Waving Engine (1678)

“Cunning of Hand”: Joseph Moxon’s Waving Engine (1678)

In 1678, a London printer and instrument maker named Joseph Moxon began publishing what would become one of the most important technical books in the English language. Mechanick Exercises: or, the Doctrine of Handy-works appeared in monthly installments — an early form of serial publication — and covered trades from smithing to joinery to printing. In the joinery section, Moxon devoted several pages and an engraving to a device he called the “Waving Engine.”

It was the first detailed description in English of a machine for making ripple moldings, and it reveals something that the later, more sophisticated machines would gradually erase: the central role of human skill in producing beautiful work.

Who Was Joseph Moxon?

Moxon (1627–1691) was not a joiner. He was a printer, globe-maker, and Fellow of the Royal Society — the first tradesman admitted to that learned body. His interest in documenting the “mechanick arts” was part of the broader Enlightenment impulse to record, systematize, and make public the practical knowledge that had traditionally been locked inside guild workshops and passed from master to apprentice through demonstration rather than text.

Mechanick Exercises was revolutionary in this respect. Guild traditions were oral and closely guarded. A Nuremberg Kunstschreiner who understood the art of geflammtes Hobeln (flamed planing) had no incentive to publish his methods — his livelihood depended on others not knowing them. Moxon, operating outside the guild system as a writer and instrument maker, had no such constraint. He observed, asked questions, drew diagrams, and published.

The Waving Engine section is characteristically precise in its description of the physical device and characteristically honest about the limits of written instruction when it came to using it. Moxon understood — as many technical writers since have failed to — that there is a category of knowledge that resists textual transmission.

How the Waving Engine Worked

Moxon’s device represented a fundamentally different approach from the earlier Kaseman wriggling plane. Where the Kaseman plane moved the cutter over a fixed piece of stock, Moxon’s engine moved the stock through a stationary cutter. This inversion would prove to be the enduring design principle — every subsequent wave-molding machine, including those documented by Diderot and Roubo a century later, would follow Moxon’s approach rather than Kaseman’s.

The device consisted of several components working together:

The slide-board: A flat board to which the stock piece (the strip of wood to be rippled) was clamped. This slide-board ran in a track, guided by rails, and the operator pulled it back and forth beneath the cutter by hand.

The cutter-head: A fixed blade, ground to the desired molding profile, mounted above the track. The blade was stationary — it didn’t move during operation. The stock piece, carried on its slide-board, passed beneath it.

The guide template: This was the key element. A shaped wooden template — essentially a sinuous strip with the desired wave pattern carved into it — was mounted alongside the slide-board. A follower pin or roller engaged this template so that as the operator pulled the slide-board back and forth, the template forced the board (and the stock piece clamped to it) to move laterally, vertically, or both, depending on how the template was shaped and oriented.

The depth mechanism: A screw or wedge arrangement that allowed the operator to raise the slide-board slightly between passes, bringing the stock progressively closer to the blade. Each pass removed a thin shaving, gradually deepening the waveform until the desired profile was achieved.

The entire device was compact enough to clamp into the end-vise of a cabinetmaker’s workbench. It was not a factory machine. It was a bench tool, scaled for an individual craftsman’s workshop.

What It Could Do That Kaseman’s Couldn’t

The critical advantage of Moxon’s design over the Kaseman wriggling plane was versatility. Because the guide template was separate from the device itself, it could be swapped out. Different templates produced different wave patterns — tighter or looser undulations, different amplitudes, different combinations of lateral and vertical movement.

Most importantly, Moxon’s engine could produce both Flammleisten (side-to-side waves) and Wellenleisten (vertical undulations), depending on which axis the template controlled. The Kaseman plane, with its guides mounted on the track walls, was mechanically limited to side-to-side movement only. Moxon’s device broke through that limitation.

Jonathan Thornton, who built a close reproduction of the Moxon engine in 1994 during a semester sabbatical and has been exploring its capabilities ever since, has demonstrated the remarkable range of patterns a skilled operator can produce. By varying the template, the blade profile, the depth of cut, and the skip pattern (how many wave cycles the operator advances between passes), a single Moxon-type device can generate dozens of distinct surface textures — from fine, delicate ripples to bold, aggressive undulations, from pure side-to-side flames to pure vertical waves to complex diagonal basketweaves.

Thornton’s Fig. 7 in his 2002 paper shows his reproduction device set up specifically to make Flammleisten, demonstrating that even a Moxon-type engine could replicate the output of the earlier Kaseman approach while also going far beyond it.

“Cunning or Sleight or Craft of Hand”

But here is where Moxon’s account becomes most revealing — and most relevant to anyone who has ever struggled to transfer skill from one medium to another.

Moxon was meticulous in describing the physical construction of the Waving Engine. He provided dimensions, materials, and an engraving clear enough that Thornton was able to build a working reproduction from it three centuries later. But when it came to using the device, Moxon hit a wall:

He acknowledged that the quality of the finished molding depended enormously on the operator’s hand — on the speed and steadiness of the pull, on the judgment of when to advance the depth, on the feel of the blade engaging the grain. These were skills that could only be acquired through practice, through the body’s gradual calibration to the tool. No amount of written description could substitute for the hands-on learning that a master joiner would impart to an apprentice standing beside him at the bench.

This was not false modesty. It was an honest admission of the limits of codified knowledge when applied to a craft that lived partly in the muscles and the ear and the fingertips. Anyone who has taught (or tried to learn) a hand-tool technique recognizes this gap immediately. You can describe how to sharpen a plane iron in exhaustive detail, but the moment when the student’s hands actually feel the wire edge form on the back of the blade — that moment cannot be written down.

Moxon’s Waving Engine was, in this sense, a machine that required a craftsman. The device was relatively simple. The intelligence was in the operator.

The Device in Context

Moxon published his description in 1678, roughly fifty years after the Kaseman engraving and roughly eighty years after Schwanhardt’s invention of the technique. By this time, wave-molding production was well established across northern Europe. Ebony ripple frames were standard in Dutch, Flemish, and German workshops. The technique was not new — but its documentation in English was.

What Moxon captured, perhaps without fully realizing it, was a snapshot of a technology at a particular moment in its evolution — the moment before mechanization began to replace skill. His Waving Engine sits at the hinge point. It is more versatile than the Kaseman plane but still fundamentally dependent on the operator’s hands. Within two years of Moxon’s publication, André Félibien would describe a device in France that took the first step toward eliminating that dependency, replacing the hand-pulled slide-board with a crank-driven rack-and-pinion mechanism.

Moxon’s device was the last of the hand-skill machines. Everything that came after it traded operator expertise for mechanical complexity. The direction of that trade — more machine, less craftsman — would define the next three hundred years of woodworking, and indeed of manufacturing generally.

But there’s a reason Thornton built a Moxon engine and not a Diderot machine. There’s a reason the moldings that come off Thornton’s reproduction have a quality that factory machines don’t quite match. The hand-pulled pass, with its subtle variations in speed and pressure, leaves a surface that is almost but not quite perfectly regular — and that slight irregularity is precisely what makes hand-scraped ripple moldings so alive under changing light.

The machine that requires a craftsman produces work that bears the mark of a craftsman. That is not a flaw. It is the point.

 

Sources:

• Joseph Moxon, Mechanick Exercises: or, the Doctrine of Handy-works (1678–80). Available in multiple digital editions.
• Jonathan Thornton, “The History and Technology of Waveform Moldings: Reproducing and Using Moxon’s ‘Waving Engine,’” WAG Postprints, 2002. Free PDF at wag-aic.org.
• Jonathan Thornton, essay in With All the Precision Possible: Roubo on Furniture (Lost Art Press).
• Lost Art Press blog, 2025 post by Thornton on reproducing the waving engine.
• 1642 Designs, “About” page (1642designs.com/about) — description of waving engine operation and CNC evolution.