Pirsig writes, “But to my surprise he didn’t see the cleverness of this at all. In fact he got noticeably haughty about the whole thing. Pretty soon he was dodging and filling with all kinds of excuses and, before I realized what his real attitude was, we had decided not to fix the handlebars after all.”

“As far as I know those handlebars are still loose. And I believe now that he was actually offended at the time. I had had the nerve to propose repair of his new eighteen-hundred dollar BMW [the book was published in 1974 so don’t be pokin’ around looking for any $1,800 new BMW bikes today, my friends], the pride of a half-century of German mechanical finesse, with a piece of old beer can!”

To me, one of the (many) larger take-aways from the anecdote is that someone who buys a fine mechanical watch, especially from brands which invest heavily in core timekeeping technology like metallurgy and escapement design, may actually be averse to understanding how the watch works – or, if not averse, at least uninterested, as having a grasp of the ins and outs of the mechanism is not the necessarily the owner’s primary source of enjoyment in having the watch. There is nothing wrong per se with this, obviously. Plenty of people enjoy their cars immensely and feel perfectly legitimate pride of ownership without understanding how an automatic transmission works.

Moreover, the possibility of tweaking something mechanical is often part of what drives interest in mechanics, and a watch is not something that the average owner generally feels an impulse to tweak. Even back in 1974, when Pirsig lamented his friend’s disinterest in the art of motorcycle maintenance, and when probably a lot more people were working on their own motorcycles, cars, and what have you than do so today, fiddling with the gubbins, as they say, was considered something best left to a watchmaker.

All this tends to set up watch owners and enthusiasts for a certain degree of detachment from the technical, and apparently stone-cold boring, aspects of horology – and especially the rather arcane world of escapements. But there is a great deal of pleasure, if you are inclined to find it there, in understanding escapements, because the principles behind escapement design are universal principles of physics and mechanics. Once you have a sense of what escapements are doing, ticking away in the darkness under the dial, you really don’t look at your watch in quite the same way ever again and indeed, you may even look differently at the world. Miracles are usually defined partly by their singular nature. But though the technology of the escapement strives for, and at its best achieves, ubiquity (historically, the most successful escapements are widely adopted), it is nonetheless miraculous for that. Appreciating this fact is by far the most democratic pleasure fine watchmaking offers – why, you don’t even have to own a watch to get in on the fun.

In examining the modern watch escapement, we’ll focus on those which are most widely adopted (we will not attempt a general history of the development of escapements; HODINKEE’s Nick Manousos has, however, provided a useful general overview) . All industrialized watch escapements are attempting to solve the same basic set of problems, and in understanding how they resemble, and differ from, each other, we can come to a new and deeper understanding of what makes a watch a watch.

Now the balance is a sensitive little thing – I should say the balance and spring because, without the spring, the balance is useless (and the other way around too, come to think of it). The balance is best understood, I’ve always thought, by looking at what it’s trying to imitate, which is the pendulum.

Think of a pendulum that isn’t swinging. The pendulum hangs straight up and down; it is fixed at its point of equilibrium and, like all equilibria, this is a rather boring situation. Leave it alone, and it will hang there for all eternity (or at least until the proton decays, whichever comes first).

Give a pendulum a push, however, and it begins to swing – how long each swing takes depends on one thing only, which is the length of the pendulum. The pendulum will pass through its point of equilibrium at each swing. How far the pendulum swings depends on how hard you push it, natch. But because how hard gravity pulls the pendulum back towards its equilibrium point, depends on how far it swings away from its equilibrium point, the pendulum should be isochronous – the time of each oscillation should be the same, regardless of the amplitude. Isochronism is an essential property for any oscillator expected to act as a timekeeper. As it turns out, practically speaking, the pendulum is only isochronous for small amplitude swings but this basic description is enough to go on with. If the pendulum loses no energy, once set swinging, it will swing forever.

It will, at this point, have occurred to the alert reader that pendulums rather noticeably do not swing forever. Why? The short answer is friction. Friction can occur in two ways in a pendulum – the first is through air resistance, and the second is at the point where the pendulum is attached to its frame. (There are other extremely small losses – for instance, the frame in which the pendulum is mounted will never be perfectly rigid and, as it flexes, it drains energy from the pendulum into the ground, albeit in homeopathic amounts). You can reduce both to almost nothing, and makers of high precision pendulum clocks mounted their pendulums on knife-edge crystal suspensions in vacuum-sealed canisters, but no practical solution is perfect. Even in these situations, pendulums will still gradually lose energy to the environment as friction converts kinetic energy into minute amounts of heat. So, to keep the pendulum swinging, you must give it a push now and again.

Now you have created another problem. When you push the pendulum, you interfere with its natural frequency – a little, or possibly a lot, depending on when and how hard you push it. Ideally, the pendulum would be given impulse instantaneously, at its equilibrium point, and there would be no variation in rate. However, anything that actually physically impulses a physical pendulum will introduce errors. The problem is made even more severe if you want to have an actual clock. Now you not only have to keep the pendulum swinging, but you must also count each swing mechanically. You therefore need some sort of mechanism that both gives impulse to the pendulum and which, in doing so, advances a gear train. Such devices do exist – they are called escapements.

The anchor escapement is an example of this simple but wonderful device.  The escape wheel is made to rotate – maybe by a weight attached to a pulley, or maybe by a mainspring barrel. As the escape wheel rotates, it is alternately locked and unlocked by the anchor, under the impetus of the pendulum. Every time the anchor unlocks the escape wheel, the escape wheel tooth slides along the curved impulse face of the anchor, giving the pendulum a push. You can see the beauty of it – the escape wheel advances one tooth; the pendulum gets an impulse; this is all you need, really, to have a clock. The name “escapement” is apt – with each oscillation, it allows one tooth of the driving wheel to “escape,” or advance.

If we look at the anchor escapement, however, we can see that it does not fully fit the definition of an ideal escapement. In an ideal escapement (and I owe a lot of this analysis to Daniels’ Watchmaking, which for a lucid explanation of the principles of a practical watch escapement is very hard to beat), impulse would be applied instantaneously at the equilibrium point, in both directions, with equal force each time in order to ensure perfect symmetry of motion (especially important in a watch). There would also be no friction involved as this dissipates energy and affects the motion of the oscillator. The anchor escapement fails on both counts – not badly, by the way; you can get excellent performance out of it – but it is not an ideal solution. Moreover, the sliding friction at the escape wheel teeth and pallets, as the curved projections of the anchor are called, requires oil, and any oil will eventually thicken and evaporate over time. The viscosity of oils will also change with temperature, and this means that the ideal escapement would be oil-free as well. A watch escapement should be self-starting – that is, its design should be such that the watch will spontaneously begin to run once a certain amount of energy is wound into the mainspring. The escapement must have good safety – that is, it should not unlock accidentally if the watch is given a shock. And overall, of course, in giving impulse and counting oscillations, the escapement should interfere with the natural harmonic motion of the oscillator as little as possible. So we have a little checklist:

• Impulse as close to the equilibrium point as possible, in both directions
• Minimal friction and, ideally, no oil
• Self-starting
• Good safety
• Minimal interference with the natural motion of the balance

All this means that designing a watch escapement that fits as closely as possible the requirements of an ideal escapement is a very tall order indeed, and if you think about it, you begin to understand why successful, practical escapements are very few and far between. Escapement design is something horologists have been fiddling with for 500 or so years, but while many are called, few are chosen, and the timeline of horology is littered with the sad, silent, inert corpses of escapements which enjoyed, as it were, a brief moment in the sun before fading and falling with all the poignant finality of a cherry blossom (oh, chaff-cutter escapement, we hardly knew ye). With this in mind, we can now look at some examples of escapements in modern watches.

So how does the lever escapement stack up when you look at it against our checklist? Not bad, my friends, pas mal. It delivers impulse in both directions, and it is also self-starting. Moreover, the lever has excellent safety. The angle of the impulse and locking faces of the ruby pallets, and the escape wheel teeth, interact in such a way as to press the shaft of the lever firmly against its bankings, which is the term for the pins that prevent the lever from moving any further at either extreme of its swing. The fact that it takes quite a jolt to cause the lever to unlock accidentally gives the escapement great reliability and is a big factor in lever escapements having found their way into wristwatches that have all sorts of adventures, from mountaintop to sea bottom and everything in between. (As a side note, banking pins are usually adjustable, but they can also just be the solid walls of the well in the movement plate in which the escapement sits; these are so-called solid bankings, and they are one of the requirements of the Geneva Seal).

So why go to the trouble of developing any new escapement at all? Well, the lever’s not perfect. For one thing, it doesn’t deliver impulse completely symmetrically (you’ll notice the lever arm is longer on the right than the left). Like any escapement, it introduces its own, characteristic escapement error – impulse is delivered as the impulse jewel on the balance passes through the notch in the upper tip of the lever, and there is a loss of energy as this happens. This, combined with other aspects of the escapement’s geometry, tends to introduce a losing error – this losing escapement error is an inherent feature of the lever escapement which must be taken into account in the design and setting up of the rest of the watch.

The most formidable problem, though, is the sliding friction between the escape wheel teeth and the ruby pallets. Those teeth are scraping along those jewels, and there are no two ways about it. And although the friction isn’t all that high – friction is proportional to load, and the loads in a mechanical watch are pretty low – it’s not nothing. In a modern lever watch running at 28,800 vph, that scraping friction happens eight times per second. That is 252,288,000 times per year … scrape scrape scrape. In five years, that’s 1,261,440,000 times … scrape scrape scrape. You need oil, and if the lever has an Achilles’ heel, it’s that it needs oil on those impulse surfaces, and oil, even the best, breaks down after a while.

Still, the lever is tried and true. Think about it – just about every watch on Earth (except for some exotics, and of course, Omega, which we’ll get to in a minute) uses a lever escapement. Even in watches with silicon pallets and escape wheels, the basic principle is the same. It is a somewhat humbling reminder to not get too up on your high horse about your watch – that 5711 you blew the kid’s tuition payment this year on is using exactly the same basic mechanism as a Seiko 5. Okay, I’m being a little rhetorical there – after all, there are enormous differences in craft and execution, across the board – but given how many hundreds of escapements have been tried over the centuries, it is a remarkable testimony to the lever escapement that it is, if not the only game in town, still the biggest after so many centuries.

Now, if we are being exact about things, and we should be, this is obviously not a new escapement per se. It is, however, an indication that refinements to the lever escapement can and are still being made, even today, and that research into how to refine the escapement further is an active project both at Rolex and elsewhere. The Chronergy escapement introduces a modified lever geometry, which allows more efficient delivery of energy, and the size of the pallet stones is reduced by half in comparison to the standard Swiss lever. The escape wheel is skeletonized to reduce inertial energy losses.

The lever and escape wheel are both made of a nickel phosphorus alloy, to reduce vulnerability to magnetic fields. One interesting feature of the escapement geometry is the angle of the lever with respect to the escape wheel. In a conventional lever escapement, the lever’s centerline is on a direct radial line from the axis of the escape wheel, but in the Chronergy escapement, it’s slightly offset. All these changes to the standard lever were made with a view to increasing efficiency – the Chronergy escapement is, according to Rolex, about 15% more efficient than a standard lever. Overall, if we stack up the Chronergy against our checklist, we see that it’s still got essentially all the strengths and weaknesses of the standard Swiss lever, but with stronger strengths and diminished weaknesses. It is a testimony to the basic soundness of the lever escapement that working on refining it is still very much a valid strategy horologically.

Okay, so this is a tricky one. The detent escapement is often also called a chronometer escapement, as you often find them in boxed ship’s chronometers. Their history is long and complex and has to be left for another time, but for our purposes, it is enough to understand that the detent escapement is nearly ideal in principle. Most significantly, there is no sliding friction in a detent escapement – in fact, there is no lever or anything like a lever. The escape wheel delivers impulse directly to the balance. This means that the detent escapement does not need to be oiled and so should have superior long-term rate stability in comparison with the lever.

So if the detent escapement is the best thing since sliced bread, how come you don’t find one in every watch? Well, a cursory inspection of the detent escapement reveals its major weakness – it unlocks if you look at it crosseyed, and for that reason, it is unsuitable in general for use in a wristwatch (now, there have been modern watches made with detent escapements, but these are in the minority). George Daniels was not the first watchmaker to whom it had occurred to try and combine the best properties of the lever and detent escapements, but he was the first to make it work in a design that eventually could be adapted, with some modifications, to large scale production.

Today, there are just two places to get a co-axial escapement – one is Omega, of course, which has been continuing to create new versions of the co-axial escapement since bringing out the first co-axial watch – a limited edition – in 1999, and the other is Roger Smith, who has also been making continual updates and modifications to the original co-axial escapement. Omega’s alterations to the design involve modern materials science solutions, while Roger Smith continues to follow an approach which emphasizes classic watchmaking materials and construction, as well as, of course, an extremely artisanal approach to watchmaking vs. the industrial-scale watchmaking taking place at Omega. I would emphasize as well that each has its place and each requires its own idiosyncratic, and I think admirable, set of skills and ways of thinking (that said, if some unknown rich uncle ever dies and leaves me untold lucre, I will bespeak a watch from Roger quicker than you can say lift angle and die a happy man, or at least, less unhappy).

The action of the co-axial escapement can seem quite confusing at first, believe me, you’re not alone. The escapement has two escape wheels – one smaller with half-ogive teeth, and one larger. Both are on the same axis, hence the name “co-axial” escapement. As the balance swings clockwise, the smaller jewel on the balance roller unlocks the escape wheels while, at the same time, a tooth on the larger of the two escape wheels impulses the balance directly, via the larger impulse jewel on the balance.

As the balance swings counterclockwise, the smaller jewel unlocks the escape wheels again but this time, the smaller wheel with half-ogive teeth pushes on the center pallet on the lever, and the lever gives impulse to the balance indirectly. Interestingly, the only lever pallet which actually transmits impulse to the balance is the central one – the two outer pallets on the lever are there just to lock the escape wheel. The lower pallet locks the escape wheels on the counterclockwise swing of the balance and the upper, on the clockwise swing of the balance. As you can see, it is quite literally half lever escapement, half chronometer escapement – in one direction impulse is indirect, via the lever and in the other, direct via an escape wheel tooth.

In general outline, the story of the industrialization of the co-axial is well known – it took quite a bit of time and quite a bit of money but, today, the escapement is found in almost the entire Omega product lineup, which is to say, hundreds of thousands of watches a year. Apparently, Omega oiled the first version of the co-axial – a modified ETA 2892 – very lightly on one driving surface, but it was not at the impulse surfaces, and when Walt Odets looked at the escapement back in 2002, he remarked that he felt it was not a critical issue.

How does the co-axial escapement look, stacked up against our little list of criteria? It looks pretty damned good. The escapement does not require oil, impulse is given in both directions (albeit directly in one and indirectly in the other – I wonder if getting the amount of energy delivered in each direction to match up wasn’t a big part of the challenge in making the escapement work, especially given the fact that the two impulse pallets on the balance roller are so far apart) and it is self-starting. Safety appears to be good as well, and long-term rate stability ought to be excellent. Despite some teething problems (which you have to expect) with early versions, Omega seems to have had the the kinks worked out for some time now. And, of course, if you want a different expression entirely of the co-axial, in something which connects more directly to George Daniels’ philosophy of watchmaking, there is always Roger Smith. I think the only downside to the co-axial from a broader horological perspective is its complexity. You could make an argument, perhaps, that having an oil-free escapement is of diminished importance when you bear in mind that the watch will have to be serviced sooner or later anyway. But if a wristwatch can run for eight to ten years, say, with unchanged rate stability right up until it’s serviced – well, that ain’t too shabby.

If we take a look at the escapement more closely, we see a number of interesting features as well as a number of respects in which it relates, from a conceptual if not an actual engineering perspective, to both the lever and co-axial escapements.

The escape wheel can be seen on the left, and you’ll immediately see that it is quite different from either the co-axial or lever escape wheels. The escapement gives impulse in two directions – in one direction indirectly, via sliding friction between an escape wheel tooth and a lever pallet, and in the other direction directly, via an escape wheel tooth engaging with the impulse jewel on the balance roller. (For an animation showing the action, check out our earlier coverage of the movement, right here). The movement also features the use of a freesprung, adjustable mass balance and an overcoil balance spring, and this, in combination with the high frequency and unique escapement design, clearly seems to represent an ambitious step on the part of Grand Seiko to up its horological game and become even more competitive with the major players in Switzerland. It’s an ambitious move.

Judged against our list of criteria for a modern watch escapement, the GS Dual Impulse escapement seems very promising. It is self-starting (this is per Grand Seiko) and gives impulse in two directions; it offers a high-beat solution to the rate stability problem, and it shares the co-axial escapement’s basic strategy of giving impulse directly in one direction, and indirectly in the other. One of the many respects in which it differs from the co-axial escapement is in the use of lubricant – the indirect impulse is via sliding friction between an escape wheel tooth and a lever pallet, requiring oiling. However, the escapement is overall more efficient than a standard Swiss lever, and it will be interesting to see how widely it is deployed across the larger GS product lineup – between, this, the MEMS-fabricated escapements used in other mechanical GS watches, and Spring Drive, Grand Seiko is now bringing a diverse range of technologies to the table.

A far-from-comprehensive list would include a number of escapements from Ulysse Nardin, since the introduction of the Freak 20 years ago; the Zenith Oscillator in the Zenith Defy Inventor; the Genequand oscillator which Parmigiani Fleurier had under development at one time, and many others. Many of these escapements rely on the elastic properties of silicon. Watchmakers have also produced timepieces that update older escapement designs, using modern engineering methods – one such escapement is Breguet’s “natural” escapement, with versions made by Kari Voutilainen, Laurent Ferrier, and F. P. Journe. Makers such as Frodsham, Urban Jurgensen, and Christophe Claret have created timepieces that adapt the chronometer detent escapement to wristwatches. The number of such experimental and concept-watch escapements is bigger than many of us might suspect, and while they tend to be announced with enormous fanfare, for any of them to result in large scale series production is a much rarer event, but there is always a chance a breakthrough might be made and a new escapement that eclipses the lever and its derivatives will take center stage. In the meantime, smaller production watches with non-lever escapements continue to add an enormous amount of interest to the horological landscape.

I would like to especially acknowledge the work of the various individuals who created the animations in this article and made them available on Wikipedia. I am also indebted very much to the many watchmakers and horologists who have patiently explained their work to me over the years. To the extent this article is correct, it is thanks to them; any errors are mine alone.