general movement

Principles of Power

With thanks and apologies to Sifu Mark, who taught me all this stuff.

When I was training Wan Yi Chuan kung fu, Sifu taught that power generation has five components:

  • Mass towards target
  • Wave-like motion
  • Crack the whip
  • Towards, through, past
  • Drive your intent

Understand that this is one perspective on power generation – every art effectively uses all of these components, but different arts think about them in different ways. The purpose of this post is to lay out my understanding of these components, how they interact, and how they contribute to particular movements.

To be clear, every movement incorporates all of these sources of power to a greater or lesser extent. You might do them badly, even so badly that the net effect is negative (such as striking forward while moving backwards), but they’re always there. The power in your movement is the combination of the contributions from each component, and the amount of each will vary depending on the movement and the circumstances.

Sifu, if you ever see this, feel free to tell me all the ways in which I’m wrong in the comments.

Mass towards target

This one is simple: kinetic energy equals mass times velocity squared. If you want to hit something hard, throw as much mass at it as you can, as fast as possible. When you throw a punch, ideally you should step into the punch so that your entire body mass is moving at the moment of impact.

Wave-like motion

The internal arts teach that all human motion is reciprocal. In order to go down you must go up, in order to go forward you must go back. Many of those motions end up being internal, but they’re no less real for it. When you step forward, there’s a rising-then-falling component to the movement even if your center of mass doesn’t physically change its distance from the ground. That falling component incorporates the force of gravity, so you can use it to increase the power of your movement.

This is true for all movements that exert force, but it’s most obvious if you look at a downward strike such as the pi quan strike in Xing Yi. As you “fall” onto your front foot, the structure of your torso and arm translate that momentum into your strike. With proper timing, you can add this power to all the other sources discussed here.

Crack the whip

This is a concept I’ve seen called “staged activation”. The idea is that each joint is attached to muscles that can create power in that joint. If you want to maximize your power the best thing you can do is to activate each of those sets of muscles in sequence so that each can add their power to the movement. The longer the distance over which you can do this, the more joints are involved and the more power you can generate.

Let’s look at a movement like a full-arm slap. I’m specifically not using a punch here because it’s a bit more complicated – in particular, I disagree with the idea of moving your hand last when you’re punching, but that’s beyond the scope of this post. In order to execute a slap with maximum power, you would start at the foot pushing away from the ground, turn the knee, then the hip, then the shoulder, then the arm (without letting the arm move outside of the zone of structure, of course).

By staging the activation of your body in this way, the power builds as it moves from joint to joint. If you time it correctly, it reaches a maximum at the moment of impact and is added to the power from each of the other sources.

Towards, through, past

This is standard advice in every boxing class: “punch through your target”. All the stuff above talks about focusing your power to a specific instant in time, which is also a specific point in space. The idea with this principle is that you don’t want that point to be on the surface of your target, or you’ll expend your energy there and not be able to penetrate – you want to focus your energy within your target so that the energy will be fully transferred.

How deep within the target is a matter of some debate, and it depends on what kind of strike you’re talking about. A short snappy strike like a jab certainly wants to be pretty close to the surface, while a committed strike like a cross should be farther inside. For something like a stomp kick (also called “heel kick”), you might even want it to be beyond the target. Go too far beyond, however, and you lose the focus of your strike; it’ll turn into a push, with no real sharp impact.

Sidebar: there is a concept in Chinese internal arts called fa jin. This has to do with focusing the power of a strike into an extremely small area and then releasing it into a target. This style of striking deliberately focuses a strike close to the surface of a target and relies on the shock wave generated within that target to carry the energy beyond the point where the strike actually ends. This is the secret to the so-called “one-inch punch” – by using all of these principles, it’s possible to generate a lot of power in an extremely small space and to deliver that power beyond the reach of your actual fist.

Drive your intent

To this point, everything has been about the shape and timing of the movement. This one is about your mental state, but of course it has to translate into the physical in order to result in a change in power. The idea here is that you have to intend to hit and to hit hard. If your intent is to strike to the target, that’s what you’ll do. You have to put your mind on the other side of the strike, both in space and in time (that is, both “past” and “after” the point of the strike), in order for the strike to arrive correctly.

I said this has to translate to the physical, so what form does that take? Primarily, driving your intent results in relaxation and lack of hesitation. If you’re striking to a target (and this is much more common than you might think, especially among people who are used to point-sparring), you’ll unconsciously start to tense the opposing muscle groups before you reach the target. For example, if you’re throwing a jab you need to keep your bicep relaxed until (just barely) after the moment of contact, and the way to ensure that happens is to drive your intent through the target.

Related to relaxation but separate, if you’re not sure about striking you may hesitate. Committing mentally to the strike removes this hurdle. You can commit to a strike without actually executing the strike – some arts refer to this as “loading”. Crane-style kung fu uses this principle and describes it (or at least it was described to me) as “leaning on a glass wall” – the idea is that you have decided to attack and you’re being held back by an invisible force. The instant the force is removed, the attack will go off and you don’t have to make a conscious decision to do it.

In kenjutsu, we practice this using an exercise called “sensitivity drills”: you lock swords with your partner, look for an opening, and try to hit it. Each person is trying to maintain a constant committed offense at every moment, so that if the obstruction of their partner’s sword is removed, their strike will occur automatically and without hesitation. If there’s a suki (gap) in your intent, your partner may be able to strike you without being struck in return, because you’ll have to make a conscious decision to move before your attack will occur.


Cutting with the rapier

Here’s something that will make certain people mad: the rapier is a great sword for thrusting, and kind of a terrible sword for cutting. The weight distribution is all wrong, for one thing – swords designed specifically for cutting have more weight towards the tip. The bigger consideration, however, is the shape of the hilt. Here’s a longsword, which is designed primarily for cutting and secondarily for thrusting:

Note that the hilt is long (sized for a hand and a half) and open – the natural grip would hold the blade at roughly ninety degrees to the arm.

For comparison, here’s a rapier:

This is an Italian rapier, but the open guard shows the hand position better than a cup-hilted Spanish rapier would. Note that the hilt is short (with no way to put two hands on it) and closed – the natural grip, with a finger over the cross, is much closer to in line with the arm.

The long hilt on the longsword allows leverage with two hands for increased power on the cut, and the open hilt gives flexibility for the wrist to swing the sword through its natural arc. The rapier’s hilt allows neither of those things, being primarily designed to support the sword in the thrust and only secondarily to allow cutting.

All of that said, cutting with the rapier is important, and more so in Destreza than in most rapier styles. If you want to be able to pull it off, it’s important to understand the mechanics of it.

I want to start by saying that rapier cuts are percussive cuts. That is, you’re striking with the blade moving at more or less 90 degrees to its direction. There’s little or no slicing action like there is with a katana. This is partly because of the structure of the hilt, and partly because the rapier doesn’t have very much mass at the tip. With so little mass to work with, you can’t afford to “sacrifice” any by angling the cut in order to get better edge geometry.

Destreza, as I was taught it, classifies cuts by the pivot point: shoulder cuts, elbow cuts, wrist cuts. Shoulder and elbow cuts are big and powerful but often too slow to use in combat, so wrist cuts get most of the action. I don’t think other rapier styles make this distinction explicitly, but I do think it’s generally useful and applies to all rapier cuts.

When I was taught to cut, Sensei described the wrist cut as an extremely tight motion – from the palm-down right-angle guard, pretty much just turning the hand to palm up and snapping the hand horizontally. That motion never felt good to me no matter how much I practiced it, and it wasn’t obviously the same kind of motion as a shoulder or elbow cut.

This bugged me for years. Now, a couple of decades later, I think I’ve figured it out. In retrospect it should have been obvious, but since it wasn’t at all obvious to me I figured I’d share.

Rather than trying to create a list of types of cuts, I want to describe the fundamental principles at play. A cut has two components to its motion: the circular component and the linear component. The former happens because you’re rotating the blade around a pivot point; the latter happens because you’re pulling the blade’s center of mass towards yourself and then pushing it back out at the moment of impact. It’s possible to minimize the linear component, but there are consequences if you do (more later about that). The angular and linear momentum of the two movements (I’m trying to avoid using the correct-but-pretentious “momenta” or the clearer-but-wrong “momentums” here) combine to create the impact force of the cut.

Let’s start with the circular component. The defining characteristic of a circle is its center; the shoulder/elbow/wrist classification is a start towards this, but you can go further. If you pick any point between your shoulder and the center of rotation of the blade (more on that in a moment), you can use that as the pivot point. For example, choosing a point at the ricasso of the blade, you can drop your pommel as you raise your point so as to pivot the blade around that point.

That term “center of rotation” is an important concept in the rapier, and it’s surprising to me that I haven’t mentioned this before. The center of rotation is the point around which the sword wants to rotate. Intuitively you would think this would be the center of mass, but you would be wrong. You can prove this to yourself: first, find the balance point of the sword – this is the center of mass. Next, hold the pommel of the sword and wiggle it back and forth, and observe the point that stays still – you’ll see that it’s not at all the same point.

If you don’t care about the physics, skip this paragraph. If you do, here’s why the center of rotation isn’t at the center of mass. Any long skinny object can be modeled as a massless rod with a point mass at each end, where the two masses and the length of the rod depend on the mass distribution of the object. In the case of a rapier, most of the mass is at the hilt, so that mass is significantly larger and the center of mass is fairly close to that end. However, you’re holding that end in your hand, so you’re moving that mass directly. This results in the blade pivoting around the mass at the other end, which turns out to be roughly halfway down the blade.

If you choose the center of rotation as the pivot for your cut, the result will be that the sword moves perfectly in balance. Unfortunately, the effect of this is that there’s no net angular momentum in the movement (if you don’t believe me, think about the massless-rod model above), so you won’t have any power in your cut. Any pivot point “above” (closer to your shoulder) the center of rotation will result in net angular momentum, and the farther “up” you go, the more mass you’ll be swinging and the more momentum you’ll get.

There’s an additional factor: the radius of the circle. The larger your circle (distance from the point of the sword to the pivot point), the faster your tip will move and the more angular momentum you’ll have at impact.

The linear portion of the movement is much simpler. The more (and faster) you move, the more linear momentum you’ll have. “Higher” (closer to the shoulder) pivots result in more linear movement due to the only-bends-one-direction structure of the arm joints, so all the components generally agree that a higher pivot means more movement and therefore more power but more time.

OK, so that was a lot of physics, so let’s tie that back to the actual technique. If you pivot at the shoulder, you’ll have the full mass of your arm and sword moving, you’ll have the largest possible radius, and you’ll get the largest linear movement. The result is that shoulder cuts are the most powerful version, but also the slowest because you’re moving the most mass over the longest distance. Elbow cuts are faster but less powerful, wrist cuts even faster and even less powerful.

What’s past wrist cuts? I use the term “snap cuts” for any cut with a pivot point past the wrist. Once you get past the wrist, cuts mostly stop getting faster because you still have to articulate your hand and wrist, but they have a significant tactical advantage in that they keep the sword in the space between you and your opponent. A wrist cut necessarily has a moment where the hand is in line but the sword is not, and in that moment you’re open. A snap cut does not have this moment, because you’re displacing your hand in order to keep some part of the sword in a defensive position.

To execute a snap cut, pick a spot on the sword as your pivot – I like to use a spot about three or four inches past the guard. Starting from the palm-down guard, you slightly bend your elbow (this creates the linear component of the movement), drop your hand, and raise your point, pivoting the sword around the chosen spot. The larger the circle your point makes (that is, the more you deflect your sword from the right-angle position), the more momentum you’ll get in the final cut, but the slower your movement will be.

If you take the snap cut and make it as small as possible, you end up with more or less what Sensei was describing to me twenty years ago. Choose a pivot point close to the center of rotation, use a very small deflection from the right angle, and you’re pretty much just turning your hand over. This is the fastest but least powerful version of the cut, and it’s why that cut never felt good to me. Pulling the pivot point back towards the hand and increasing the deflection angle results in a lot more momentum and makes the cut feel better.

One other point here that applies no matter which pivot point you choose. I said above that minimizing the linear component causes problems. That’s because, at the end of the cut, you want your force going outwards, not along the arc of the cut. If you minimize the linear component, you’ll end up with entirely angular momentum, and if you don’t hit something it will want to fly around the arc and end up far out of line. By introducing the linear component you not only add more power to the cut, you redirect that final momentum to an angle where it will try to pull the sword away from you rather than out of line. This makes it a lot easier to control the followthrough of your cut and lessens the need to reduce your power in order to keep control.


Respect and ego

Ego is a major problem in the martial arts. Any time you put people into a competitive situation you get egos involved, and when it’s as “masculine” as martial arts (which is not to say that there aren’t extremely able female and nonbinary martial artists!) it’s worse. Everyone is familiar with the swagger associated with MMA fighters, boxers, and so forth, and there are lots of people in the traditional martial arts who are just as bad.

TMA often disdains competition and claims to put a priority on humility, so ego there is usually quieter but just as prevalent. It’s the guy telling you that the way you’re doing the technique can’t possibly work, or challenging you to spar “to see which style is better”. It’s the student saying “that’s not the way we did things in my last school,” or “I like to do it this way and not the way you’re teaching.”

Ego is damaging. I was in a class once with my sensei. I was already a black belt in kenjutsu, and he had started teaching a bunch of new students, so I was far and away the most senior person in the room. He was demonstrating shinchokugiri, the vertical cut, something a student usually learns in their first few classes. I was paying only marginal attention and just drilling the cut that I’d done thousands and thousands of times.

Sensei repeated the lesson over and over. Finally, he pulled me aside. He told me that my cuts were terrible, and that the lesson that he’d been giving was for me. I was so sure that I knew what I was doing that I’d been ignoring him as he was explaining what I was doing wrong. If he hadn’t taken the time to specifically whack me on the head (figuratively… at least this time), I would have just kept doing it wrong and probably teaching it wrong.

The flip side of ego is respect. I wasn’t listening because my ego said I had nothing more to learn about that simple technique (for the record, fifteen years later I’m still learning about it), and I didn’t have enough respect for my sensei to assume that what he was saying would be valuable. If you respect someone, you listen to what they have to say and try not to prejudge based on what you think they’re going to say.

Respect, of course, goes both ways. As a teacher, you must have respect for your students. Of course they’re not as skilled as you are – if they were, they’d be in the wrong class. That doesn’t mean that they have nothing to teach you, and it doesn’t mean that when they disagree with you they’re wrong. Listen to what they say and consider it respectfully. More often than not, you’ll go ahead and tell them that they’re wrong and point them towards the right answer, but once in a while they may surprise you.

I do a lot of writing here about kenjutsu, and some about Destreza. I mention kung fu and taiji sometimes, but not to say “this is how to do it”. I pretty much never talk about escrima even though that’s what I’m currently training. Why is that? Because I have respect for my teacher and the art.

I have permission from my sensei to teach the things that he taught me, which includes kenjutsu, Destreza, and taiji. He also taught me some bagua and xingyi, but I don’t have enough training in those arts to speak intelligently about them so I don’t.

I don’t have permission to teach kung fu or escrima, so I don’t, even on those aspects where I do feel qualified to offer an intelligent opinion. To do so would be to disrespect my teachers in those arts, because I’d be implicitly claiming that I could teach it as well as they could. That’s not my decision to make, and I won’t take it upon myself to make it.

At some point, if I stick with escrima, I’ll be asked to teach it to more junior students, and I’ll happily teach what I’m asked to teach. If I’m given permission to teach the entire art, I’ll consider that my “graduation” (which doesn’t mean I’ll stop training and learning, of course!) At that point, you may hear about it here.

kenjutsu weapons


There’s a section of the curriculum in the Suisha Ryu called “muto”, which we inherited from the Yagyu Shinkage Ryu. It deals with disarming the sword – that is, if you’re fighting unarmed against someone with a sword, these are techniques for taking the sword away.

I feel like I’m saying this a lot recently, but obviously this is not something you’re going to do if you have a choice. If you have a sword, use it. If you don’t have a sword and they do, run the hell away if you possibly can. Trying to take a sword out of somebody’s hand while they’re trying to kill you is a good way to… get killed.

OK, now that’s out of the way. If you do need to do this, how do you do it? In Suisha Ryu, we do a lot of one-handed cuts, so we start by doing disarms against one-handed cuts (which are conveniently a good deal easier than against two-handed cuts, as you might imagine). For a one-handed cut, the general steps are these:

  1. Step inside the arc of the cut and stop it, just like I described in the post about jamming (except without a sword, obviously). Also just like jamming, you need to simultaneously provide some threat of offense or you’ll just get hit.
  2. Capture the sword arm so that it doesn’t get away while you’re messing around with the sword. Grab the tsuka (hilt) with the hand that’s not holding the sword arm.
  3. Turn the tsuka over the sword hand and to the outside. For a forehand cut, you’d be turning it counterclockwise; for a backhand cut it would be clockwise. This will force the attacker to let go.
  4. Back away, possibly cutting with the sword as you go, though you’ll probably have a terrible grip on it so your cut will almost certainly be awkward.

Let’s look at a specific cut to see how that works. As usual, I’ll start with forehand kesagiri (downward diagonal cut), because it’s the most common cut. As the attacker begins their cut, you step in and raise your left forearm to stop the cut (making contact with the attacker’s forearm). Simultaneously, you strike to the attacker’s face with your right hand – I like to do this as an open palm to the nose, because the open hand is startling, gives you a lot of room for error, and lets you strike hard without worrying too much about hurting yourself. The point of this strike isn’t so much to hurt them as it is to take their attention away from what’s going on with their hand.

Assuming you pull that much off, you move your right hand to capture their arm at the elbow. Don’t go much below the elbow or they’ll fold their elbow away from you and you won’t have room to manipulate the sword. Move your left hand to the tsuka, palm up with your thumb to the inside (so your fingers are pointing back towards yourself).

Finally, torque the sword counterclockwise, pushing it against their little finger and turning their thumb towards the floor. They don’t have a lot of rotation available in their wrist in that direction, so it will quickly lock up and let you torque the sword against their fingers. You’ll end up holding the sword in a left-handed reverse grip, all the way at the end of the tsuka near the kashira (butt cap). At that point, you can back away, maybe making a clumsy reverse grip cut with your left hand to keep them from chasing you as you do. Once you’re clear, you can fix your grip.

For other cuts, it works more or less identically. For a backhand yokogiri for example, you’d catch the cut on your right forearm and strike with your left hand. You’d capture the arm with your left hand and the tsuka with your right, torque the sword clockwise, and end up in the same awkward reverse grip but right-handed.

OK, so much for easy mode. I haven’t said anything about shinchokugiri (vertical cut) yet, and that’s because it’s pretty much always done two-handed.

For two-handed cuts, you enter the same way, but then you need to use two hands on the actual disarm. The trick is to put your two hands over the attacker’s two hands, with your palms on the backs of their hands. You then press into their hands, which puts you in a stronger position than they are (and also hopefully they’ve just been whacked in the nose and aren’t thinking about fighting you). This will cause their wrists to cross and lock together, and further pressure will let you take the sword out of their hands. You’ll end up in the inverse grip: two-handed, reversed, with your left hand next to the tsuba. This works with any cut, but with forehand cuts it takes a lot more rotation to cross the attacker’s wrists than it does with backhand cuts. It can be done from that side, but it’s not ideal.

Tsuki is a completely different beast in that it’s done with the blade more or less in line with the arm, so there’s no nice convenient lever sticking out. Instead, we’re going to use the blade itself as the lever. When the thrust comes in, you slip to your left (keep in mind that on a standard tsuki, the blade edge is to the attacker’s left or your right, so you’re moving to the dull side of the blade). You put the back of your forearm or wrist on the mune (spine) of the blade and lift it upwards as you move in. The goal here is to flip it entirely over, and ideally you’d like to lay it on the attacker’s right shoulder so that the ha (edge) of the blade is against their neck. You may or may not pull that off, but it’s OK – the point is to torque the attacker’s hands and to put the blade in a position where they can’t easily just turn it into you. You’re still striking towards their face while you do this, for the same reason as in all the other disarms.

Once you get the blade out of the way, you slide both hands down to the tsuka. It really doesn’t matter where or how you grab it, but ideally your hands are over the attacker’s hands in the same palm-to-back-of-hand position as for other two-handed disarms. Once you get your grab, you simply drop your weight and step back, and you’ll find that the attacker is forced to let go. Ideally, you’ll end up in a good two-handed right-hand-lead grip on the sword, though in practice you’ll usually have to adjust the position of your right hand.

One caveat – in most of these disarms, as in almost every other technique where you’re manipulating someone’s body, a prepared and resistant opponent can make the techniques not work. If you try a hip throw on somebody who’s ready for it and who wants to resist, you’ll find that it’s very difficult. Does that mean hip throws don’t work? Well, I have thoughts about that, which I’ll save for a future post, but in general – no, of course they work, you’re just not doing it right. To make throws work in combat, you need atemi (strikes).

Here’s an exercise I like to do with students when they’re learning these techniques. I’ll get to the point where I have a grip on their sword, and then I’ll stop and tell them to resist as hard as they can. I’ll demonstrate that they’re strong enough to prevent me from taking away the sword. Then, while they’re still resisting, I’ll reach out with my other hand and tap them lightly on the forehead. In the moment when they’re figuring out what just happened, I’ll take the sword away.

Disarms, and grappling techniques in general, don’t work if your opponent is prepared and centered. You need to break their intent to make them work; that’s what the strike is for when you’re entering. If you try it without the strike, odds are good that it won’t work. If you land that strike solidly, they may let go of the sword without you having to do any fancy disarming technique.

empty hand


Another point on which different arts disagree: when you punch, which knuckles should make contact? Some arts, notably karate, use the “front two knuckles”, in other words the index and middle fingers. Other arts, such as most kung fu styles I’m familiar with, use the “back three knuckles”. Which one is right?

Like everything else in martial arts, of course the answer is “it’s complicated”, but also like everything else, there are people who will argue to the death that their way is right and everyone else is wrong. Before I give you my answer, let’s look at the considerations that people claim as the rationale for the decision.

First, all else being equal, bigger knuckles are stronger than smaller knuckles. The size of the knuckle depends on the size of the metacarpal bone in the hand, and that’s the bone that will take the force of the strike. In particular, the little-finger knuckle is especially vulnerable, because the last metacarpal bone is very thin; there’s a thing called a “boxer’s fracture” where you break that bone in a specific way because you’re punching with the last knuckle on your hand. Of course, it’s also a question of how many bones you distribute the force across: three smaller bones may be as good or better than two larger bones.

Second, there’s a question of alignment. Both camps will tell you that their preferred hand position creates better alignment of the hand bones to the arm bones. Obviously, better alignment means you’ll be able to send more power out through your hand without risking damage to your wrist, so this is important.

Complicating this issue is that different kinds of punches change the effectiveness of one position or the other. The biggest difference is horizontal versus vertical punches, and in my experience this tends to be where the battle lines are drawn: karate, for example, uses primarily horizontal punches and focuses on front-two-knuckles, whereas Wing Chun uses primarily vertical punches and focuses on back-three-knuckles.

OK, enough setup. My personal opinion is that back-three-knuckles is superior in almost all cases, so from here on out I’m going to be telling you why I think that. First, the strength of bones: the difference in cross-section between the biggest metacarpal and the smallest is not that large, certainly less than 2x. Given that the third metacarpal (the middle finger knuckle) is shared between the two punching positions, if you assume that the second metacarpal (index finger) is twice as strong as both the fourth and fifth, then the total strength is identical between the two positions (2+2 versus 2+1+1). In practice, for most people it’s certainly less than 2x in terms of tensile strength, so three knuckles will almost always have more bone behind them than two as long as you distribute the force correctly.

One caveat, though – the danger of using three knuckles is that it creates more opportunity to screw it up. The boxer’s fracture is much more likely than it would be if you were focusing on the stronger side of your hand.

Alignment is a more interesting question, because it’s very hard to prove objectively which position offers better alignment. It’s fairly easy to prove it empirically though, and I’m going to give you an experiment you can try: do a pushup. Do it on your knuckles. Keep in mind that we normally punch to the center line, so put your fists together beneath your chest or shoulders. Which knuckles are on the ground? The answer is that it’s the back three knuckles, and if you try to put your front two knuckles down you’ll find it extremely difficult. You can try this with your fists horizontal or vertical, and you’ll get the same result in both cases.

Now, I said that back-three-knuckles is superior in almost all cases. The one exception is for a horizontal hook punch. You can do a vertical hook punch if you’re close enough, and in that case I think back-three-knuckles is still the right approach, but if you’re doing a horizontal hook and trying to hit with the outside knuckles you’re likely to either strike a glancing blow or just clip the last knuckle and hurt yourself. In that case, I think using the front two knuckles is preferable (almost inevitable, in fact, given the hand position).

To be very specific, in my experience the best point of contact for a straight punch is on the outside of the third knuckle (the middle finger):

This is where I have found I get the best alignment and the most power, and it’s how I teach my students to punch. If you try leaning your knuckles on a wall and rolling your hand back and forth so that the contact point moves across your knuckles (remembering to put your hand on the center line), you’ll probably feel what I mean.

There are certainly people who will disagree with me on this one. If you’re one of them, please give me your reasons in the comments. If it’s just “because it feels better to me”, consider the possibility that it feels better because that’s the way you’ve practiced it (I’ve technically trained both ways, but my karate days were a very long time ago so I certainly have this type of bias as well). If you have objective or clear empirical evidence that a different hand position is better in some or all situations, please speak up!



“Jamming” is a name for a class of defensive techniques with the katana. These are niche techniques somewhere between passive defenses and infighting; they may be an entry into infighting, or they may just be an aggressive form of blocking.

A “jam” is a passive block performed inside the arc of the attack, so it makes contact with the attacker’s arm rather than with their weapon. This wouldn’t work with most weapons; a short weapon like a knife isn’t swung at arm’s length (if the attacker knows what they’re doing, anyway), and a long weapon like a staff will cover the arm that’s holding it. It works specifically with the katana due to the method in which it’s wielded: one- or two-handed, with arms fully extended out from the body.

Jamming also requires that you be early – you have to have enough time to move inside the attacker’s range and place your block. Generally this works because the attacker is also moving in and expecting you to retreat, so it takes quite a small forward movement to get inside their range. It also requires the attacker to have some sort of tell, though – if they execute their attack properly and without warning, you’ll almost never pull this off.

Mechanically, a jam is similar to a passive block: the sword is placed where it will cross the arc of the attack. There are a few differences, however. We don’t need to worry about edge-to-edge contact – in fact, it’s beneficial to lead with the edge – so we don’t turn the wrist. We don’t need to worry about the attack sliding, so it’s not as important to have the point lower than the hand, though usually we still do just because it’s mechanically better to meet the attack close to 90 degrees.

The biggest difference, though, is in support. Meeting the arm means that you’re going to absorb the entire force of the attack, rather than only the amount that’s transferred through the attacker’s blade. A normal passive block will crumble under that much force unless you have two hands solidly on the hilt, and even then it’s iffy. For jamming, therefore, we want to reinforce the blade with something that can support that much incoming force. The way this done differs based on the angle of attack.

There’s an additional consideration, as well – jamming leaves us very close to the opponent, in infighting range. At that range, you can’t afford to have a moment where you have no threat of offense, or you’re definitely going to get attacked. Getting attacked at close range while your sword is buried (possibly literally…) in a block means you’re probably getting hit. To avoid that, we need to incorporate some offense into the movement.

We’ll start with the most common case: a forehand kesagiri (diagonal downward cut). In this case, the block is done with the kissaki (tip of the sword) pointing left so as to catch the attacker’s arm in the center of the blade. You reinforce the blade with your own forearm against the mune (spine), with your fingers towards the attacker’s eyes. This, combined with the sword edge in their arm, will hopefully make them flinch, giving you the opportunity to follow up with additional attacks.

On the backhand, the approach is similar, but you can’t use your left hand to do it. Instead, you raise your right elbow, brace the mune on your upper arm or shoulder, and drive the elbow into the attacker’s face or chest. Your left hand should be on the hilt as additional stabilization, but as soon as the attack lands you can release and use it to strike.

Yokogiri (horizontal cut) can be handled similarly, using the left arm or right upper arm to brace the sword. In this case, however, because the attack is lower you have less ability to strike towards the attacker’s face during the jamming movement. This makes your counter less effective, so you may want to add a low kick or knee, headbutt, or shoulder strike on your way in. The nice thing about jamming is that it has effectively no minimum range, so you can place the sword and then drive all the way in with a body strike, and if you’re early you’ll just catch them higher on their arm. If you’re late, of course, you’re in trouble.

Kiriage (rising cut) is difficult to jam and generally not worth it. It’s almost always best to step away from a kiriage, but if you must step in you can just interpose your body without worrying about the sword; kiri has so little power that you can largely ignore it once you’re inside the arc. Kiri is also short, however, so you need to be almost body-to-body to be sure you’re inside the arc; the other option is to trap or otherwise control the sword arm on the way in. Again, there’s (comparatively) so little power in the cut that this is fairly straightforward if you get the timing right.

Shinchokugiri (vertical cut) can, as usual, be treated as either kesagiri; pick a side and get yourself slightly off the line on the way in.

Jamming tsuki (thrust) is really not a good idea, for obvious reasons.

I’ll just repeat here that this is a niche technique. You’d do this because you’re trying to get inside and you’re early enough to pull it off, or because the attacker is advancing and you don’t think you can retreat enough for safety. This isn’t a normal defensive technique; given a choice, it’s usually best to stay outside and use your sword as a sword.


Dynamic vs Static Balance

I’ve had lots of conversations with other martial artists about movement and balance, and one stumbling block that we often run into is that not everybody defines the word “balance” in the same way. If you look at the movement of an Olympic fencer, for example, they don’t follow what I consider to be the fundamental principles of footwork. Is that because they’re not good martial artists? If you give a high-level epee fencer a smallsword, I think you’ll find that they’re extremely effective in combat, which pretty much by definition means that they’re a good martial artist. So what’s the disconnect here?

Fundamentally, the difference is whether you’re looking at each instant in time as a separate condition, or whether you’re averaging over time. Let’s look at an oversimplified example: imagine you’re standing with your feet square and spread at shoulder width, and your weight centered. This is a statically-balanced position: you can remain still without any effort.

Now, rock your weight back and forth between your feet without bending your knees, so that one foot is off the ground at each moment. This is not statically-balanced: you can’t freeze at any moment, or you’ll fall back to the center. However, if you average your position over a span of a few seconds, you’ll end up in the same position as the first case. It’s dynamically-balanced; over time, you can remain still on average, even if you can’t actually remain still at any given moment.

Running, even with perfect technique, is a dynamically-balanced movement – you can’t just freeze and expect to remain standing. Walking, done properly, can be statically-balanced – it can be done in a way so that you can freeze at any moment and be stable. This is the fundamental difference between a run and a walk, and it’s why we have trouble with speeds that are in between a fast walk and a slow jog.

All else being equal, static balance is preferable to dynamic balance. If you’re statically-balanced, you have a lot more options – you can change your direction or speed of movement at any moment, by any amount. If you’re dynamically-balanced, your options are limited – usually, you have to come back to a position of static balance before you can change, though there’s some amount of change you can make within the envelope of your movement (e.g., you can run faster or slower by some amount and you can curve to one side or the other, but you can’t stop dead or reverse your movement).

Of course, all else is definitely not equal. There are lots of cases where dynamic balance is what you need. One reason why you might need dynamic rather than static balance is that by its nature a statically-balanced position can’t have much momentum. If you want to strike with your full body mass, you need momentum – there needs to be some sort of ballistic movement, however small. Of course, this is a committed strike, not the sort of thing you’d do every time (even a cross only moves to the front foot, not past it like you’d get with a ballistic movement). The “straight lead” punch in Wing Chun/JKD is an example of this kind of strike.

Another example would be a lunging movement – like, say, what a fencer does. If done correctly, you can recover from the movement quickly, but your opportunity to stop halfway through or redirect it is always going to be limited.

Now, here’s the tricky bit: dynamic balance is what humans are used to doing. When you’re a baby learning to walk, you basically topple forward and catch yourself with a foot. Repeat on alternating sides and you’re walking. Lots of people walk that way throughout their lives, and in fact if you read about the biomechanics of walking you’ll see a lot of people describe that as the canonical and only way to walk. That’s wrong; it’s entirely possible to walk without ever leaving a position of static balance, it’s just not fully intuitive to do so because of the way we learn.

Learning statically-balanced movement takes training, and that’s a large part of any martial arts curriculum. You may think you’re learning how to punch, but what you’re actually learning is how to have your feet and body in the right place to support your punch; any idiot can learn to throw out their fist in two minutes, but that doesn’t mean they’ll have any power or control. If you’re learning to kick, this is even more important; you’re moving so much mass that if you’re not statically-balanced (keeping in mind the force you’re absorbing when you hit), you’ll almost certainly lose control.

This is one of those things that lots of people just internalize without ever thinking about it. In my opinion, however, having these concepts in your mental toolbox gives you more ways to understand movements. If you’re analytical like I am, this will help you figure out the right way to perform techniques. Of course, that doesn’t get you out of doing the work!

mind training


I have a training partner. His name is Bill. He’s about an inch shorter than me but in better shape than I am. He’s white, in his 30s, and has pale blue eyes, short, messy blonde hair, and blonde beard scruff. He wears an old white gi with sleeves and pant legs that are a bit too short, and a black belt that’s a bit frayed at the edges.

Bill doesn’t exist. I invented him in order to have a picture when I’m training solo. These days, given <gestures broadly> we’re all doing a lot more solo training, so visualization has become a really critical skill for improving your abilities. I’ve been working on infighting with my students recently, and visualization is absolutely critical to understand those techniques because at that range you really can’t separate what you’re doing from what your opponent is doing.

Even in more “normal” solo training, though, visualization is still an important tool. If you’re working on, say, punch defenses, it’s important to understand where the punches are coming from and what effect they’ll have on your body. It matters whether your opponent is six inches shorter than you or six inches taller, and it matters whether they’re standing twelve inches away or three feet.

Most people find visualization really difficult, but it’s a skill that can be developed like any other. It helps a lot if you’ve actually had the experience that you’re trying to visualize (blocking a particular punch, for example), but it’s not critical.

So, how do you go about building the skill? Like any other skill, by practice. Pick a simple movement, the simpler the better – say, parrying a jab. Decide where your opponent is standing, how tall they are, how far away they are, which hand they’re punching with, which foot they have forward, etc. Now, visualize the punch coming and execute the parry. What happens?

What probably happened is that as soon as you started thinking about what your own hand was doing, you lost track of what the opponent was doing. The fix for this is to strengthen your visualization “muscles”. Here’s how to do that.

First, add detail to your visualization. The more detail you can add and the more clearly you can picture the opponent, the stronger your ability to visualize will get – having blurry patches in your visualization makes your brain have to work harder when it tries to fill in the bits you really care about. Figure out not just how tall they are and where they’re standing, but what they’re wearing, what color their eyes are, what their breath sounds like, all the “irrelevant” details – see my description of Bill at the top for an example. Work on picturing them just standing in front of you as clearly as possible, without having them actually do anything.

Once you have a clear picture of the opponent, you can start working on making them execute a technique. Start by picturing the technique slowly, and don’t try to respond to it yet. Have them repeat the technique over and over again, until you can clearly picture it without much effort. Don’t speed it up yet. If you’re having trouble picturing the technique, try executing it yourself in a mirror. If the left-right flip causes problems, you can either take video of yourself, or you can do it on the opposite side (so left-handed for a right-handed punch) and pretend that the person in the mirror is doing it right-handed.

At that point, you can start trying to respond to it. Let’s take that punch defense as an example – really see the jab coming, in slow motion, and see where your parry meets it. Where on the opponent’s arm are you making contact? Exactly what is the angle of force applied? How far is their strike deflected, and how far after that does it move before they retract it? Make sure you can picture all of those things, still working slowly.

Once that becomes easy, you can start speeding it up. Bring it up slowly, making sure you’re not losing detail. If your picture starts getting fuzzy, slow it down again until it’s clear and work there for a bit. Once you’ve got that up to full speed, or at least the speed at which you want to train, then (and only then) can you move on to the next technique.

That all sounds like a lot of work, and it is. Like everything else, though, it gets easier with practice. After you’ve done a dozen or so of these, the next one should be fairly straightforward. After you’ve done a hundred of them, you should be able to pretty much pick any technique you’re familiar with and work it at full speed. If you move into a new domain – for example, unarmed to weapons – you may have to start over and build up your visualization muscles again, but it’ll go faster because body mechanics are the same no matter what.

One last point here – even though I’m calling it visualization and talking about pictures, it doesn’t have to be limited to sight. You should definitely include touch, in particular – when you parry that jab, what do you feel on your hand? It’s worth thinking about adding hearing as well, just for the sake of immersion. Smell and taste are probably overkill, but hey, whatever works for you.

If you’re really serious about this, there’s another approach you can take to strengthen your ability to visualize. It’s a meditation technique called saiminjutsu, and it’s basically self-hypnosis. You make up a short journey for yourself, and then visualize yourself taking that journey in as much detail as you can manage, including all the senses.

Traditionally, at least the way I was taught, it starts with something like a staircase – this is a way to get yourself into the visualization mode. After that is a path of some sort, and somewhere in that path there’s a door that you pass through. At the end of the path is a place that you can go and spend time in; mine is an imaginary dojo, which I’ve designed in great detail in my head. The details here aren’t really important, it’s the exercise of visualizing the path intensely, including the feel of the ground, the sights, sounds, and smells of the environment, the exertion of moving yourself along, etc. All of that will contribute to building up your visualization “muscles”.

I’ll say again, visualization is a critical tool for anybody who’s practicing solo. Martial arts is necessarily a multi-person activity, so if you want your techniques to be at all realistic and you’re operating without another person, you’ll have to invent them.

mind training

Mushin, Zanshin and Flow

The term mushin means “no mind”. It refers to a frame of mind that is characterized by a lack of attachment, and it’s an extremely important aspect of all combat arts. Unfortunately, it’s hard to describe and hard to understand if you haven’t actually experienced it – but I’m going to take a shot anyway, so temper your expectations.

Let’s start with that term “attachment”. What does that mean? This isn’t attachment in the Jedi-must-not-marry sense, though that has its basis in the same Buddhist philosophy. Attachment in the mushin sense means paying attention to something beyond the moment in which it’s actually relevant. An example: you’re driving your car, and somebody crosses the street in front of you. You register their presence and hit your brakes so you don’t crash into them, and then you pass by them. Ten seconds later, you’re still thinking about what an idiot they were, so you totally miss the fact that there’s a cop with a radar gun by the side of the road.

The idea of mushin is that you should give everything around you precisely as much attention as it needs, and no more. You should remain open to new things as you’re dealing with the old things, and you should discard things from your attention the instant they’re no longer important. Another example: say you’re in a fight, in a room with pillars supporting the ceiling. You’re moving around the pillars and focusing on your opponent. The pillars are occupying a part of your attention, but as soon as one is more than a couple of steps away you’re no longer worrying about it, because it can no longer affect you. You’re not thinking about what color the pillars are, or how they’re attached to the ceiling, because those things aren’t immediately relevant, and spending any attention on those things will probably get you killed.

Now, a counterexample: you’re in a fight, and you get hit in the side of the head. You’re not stunned or badly injured and the fight continues, but it hurts and you can feel blood trickling down your scalp. Even though the injury doesn’t affect your ability to fight in any way, you let it distract you and so you take a much more serious hit.

A second counterexample: you’re in a fight, and your opponent suddenly shouts and stamps their foot. You’re momentarily distracted thinking about what they did and why, and in that moment they hit you. This, by the way, is one of my favorite tactics, but it can work the other way as well – in the moment when I’m trying to distract you, I am myself distracted. If you’re sensitive enough, you could move into that suki (gap) and strike me while I’m setting you up.

Next, zanshin – this is a strongly related but distinct concept. Zanshin, or “remaining mind”, is a state of open awareness of your surroundings. Without mushin, it’s extremely difficult to maintain zanshin, because as soon as anything interesting occurs your mind will attach to it and you’ll lose your awareness. Without zanshin, mushin is still possible but less useful; you can be free of unnecessary attachment but still have tunnel vision.

Having defined our terms, then, let’s talk about flow. Flow is the application of mushin and zanshin in combat, and fundamentally it’s about the elimination of surprise. Surprise is what happens when the state of the world doesn’t match your expectation. Because in normal life we’re all operating on our expectation of the world rather than the reality (have you ever expected there to be one more step on a flight of stairs and been uncomfortably surprised when there wasn’t?), when there’s a mismatch it takes a moment to adapt your mental model to reality.

If you have mushin, you’ve thinned your mental model as much as possible, so you’re operating as closely as possible on actual reality rather than what you expect reality to be. With zanshin, you’re open to all the reality around you, rather than restricting your attention to only where you think it should go. With both, you can shift fluidly (thus “flow”) to adapt to your surroundings from moment to moment, no matter what happens. If a second opponent jumps out of the bushes at you, there’s no moment of surprise – you simply adapt as best you can to the situation of the moment, just like you would to the pillars in the room.

(A side note: I have, during meditation, experienced being able to literally see the curtain of my preconceptions and to flip back and forth between what I expect to see and what is actually hitting my eyes. It’s a really odd experience and I have no idea if it corresponds to anything real or if it’s just an illusion my brain came up with. I’ve never been able to do it to that degree in any other situation.)

Of course, this is all much, much easier said than done. I have experienced mushin in training, but I can’t do it reliably when I want to (or possibly my standards for what constitutes mushin go up as I get better at it – see The “Oh God, I suck” moment). I think I’m OK at zanshin, but that’s a hard one to evaluate – it’s tricky to be aware of not being aware. If you want to get there, the only tool I know for doing it is meditation, consistently and over a long period.

Sensei, who was much, much better at mushin than I’m ever likely to be, used it to distinguish between a “real” fighter and what he called a “technician”. A technician, in his lexicon, is somebody who’s technically proficient – they can execute all the movements correctly and effectively – but who doesn’t have the real combat mindset. Watching him work was a real education; you could never surprise him with anything, and whatever you did was countered in the moment when you did it. It doesn’t mean he was always perfect – you could just move faster than he was physically capable of, or execute an attack that he was out of position to counter – it just meant that there wasn’t the usual lag between your action and his counter that normal humans have.

This, to me, is the mark of a master. It doesn’t matter if you only know one technique – if you can consistently apply that technique in the right way at the perfect moment, you’ll win most fights.

general training

Levels of abstraction

When I start teaching kesagiri (the downward diagonal cut) to a new student, I describe it as a rotation from the shoulder through more or less 180 degrees. This lets the student focus on keeping the cut straight, maintaining the structure of their hand, and so forth.

After they can do that version reliably, which usually doesn’t take long, I introduce the weight shift that goes along with the cut. The weight shift causes a body turn, which reduces the necessary arm movement (because more of the movement is coming from the body). This lets them think about synchronizing the cut with the step and using the body movement to create movement in the sword.

Once that much is understood (which usually takes quite a while), I start talking about structure in the body and the shoulder. This doesn’t produce a large change in the student’s movement, but it gives them another way to think about the technique. In order to maintain structure, they end up reducing their arm movement even further, keeping their hand on their center line, and creating more of the movement from the body.

Finally, I talk about circles and moving from the hara (center). The energy in kesagiri moves down and back to a first approximation, but if done properly it ends up circling in such a way as to throw the body weight forward at the end of the cut; this allows you to move smoothly into the next technique without losing energy in the transition. If you have a proper connection between your hara and your sword, you effectively remove the arm movement from the equation – not that it doesn’t happen, but rather that it happens as a consequence of the body movement and not as a separate action.

That’s the end of the formal lessons on how to execute a kesagiri. Getting this far normally takes at least a year, and that’s for a student who practices diligently. For a more typical student (😩), it’s several years at least. That doesn’t mean it’s the end of understanding kesagiri, though. The circles can be refined and further integrated, there is variable acceleration during the cut in order to focus your power into the moment of impact, and a thousand other tiny details to be understood and cleaned up.

Kesagiri is usually literally the very first technique that a new student learns. I want to say something like “to master it takes years”, but I think it’s not reasonable to even think about “mastering” a technique like this – all you can do is refine it further and further. The refinements get smaller as you go on, but they never end.

So, here’s the question: which of those versions of the technique is actually kesagiri? The answer, of course, is all of them, at different levels of abstraction. The initial version from the shoulder is a highly abstract form of kesagiri, without any of the aspects that make it actually effective, yet it’s still recognizably kesagiri.

When you learn a technique, you’re inevitably learning the abstract version, whether you know it or not. It may not be as obviously abstract as the from-the-shoulder version of the cut, but your teacher is definitely leaving things out. They may be doing it deliberately in order to simplify the technique for you, they may be doing it unconsciously because there are things they’ve learned to do without realizing they’re doing them, or they may be forced to do it because there are things that are just too difficult to describe in words.

It’s very easy to “learn” a technique, add it to your list of techniques, and call it done. The quotes are there, though, because “knowing” a technique is not a binary thing. If you know how to execute a cut, or a round kick, or a hip throw, you don’t get to just check it off and call it done – there’s a continuous and never-ending process of refinement to be done on every technique.

The lesson here is that you can’t ever assume that you understand something. You always need to be looking for that next level of abstraction down, where something that was previously glossed over becomes specific. Interestingly, often when you break techniques down far enough they stop being distinct – for example, if you know how to throw a jab and a cross and a straight punch and a vertical punch and an uppercut, practice them enough and eventually they all just turn into “punch” with the same set of principles behind them.

Bruce Lee famously spoke of understanding the form of techniques and then discarding the form. This is an example of what I’m talking about: you can’t discard the form without understanding it first. You start with the crude outline of a technique, refine it until it’s effective, refine it more until it’s just an application of principle, and then you can discard the technique and just apply the principle.

Whatever you’re doing, work to understand it “all the way to the bottom”. Anything else is cheating yourself.