Thursday, April 14, 2011

A new look!

I thought that it was time for a new look for this blog - one for the 2010s. I hope you like it!

Monday, April 11, 2011

"Boards don't hit back": Part 2

Continued from Part 1.

So, in order to learn to strike a "live" opponent the way we would strike a "dead" target, we need more than bags and shields. We need to learn timing skills - skills that comprise appropriate, if not optimal, reactions.

Sparring is going to test these skills, but won't necessarily teach them. What will teach these skills are drills: drills comprising elements of techniques isolated for practise. But as we've seen, such drills will probably not teach literal fighting techniques (ie. "when he does this, you do that").

Principles vs. techniques

So if martial arts drills don't teach literal fighting techniques, what do they teach? As one of my colleagues at the Traditional Fighting Arts Forums is fond of saying, martial arts drills teach principles - not techniques.

When you are taught a martial arts drill you learn a general principle of movement; an appropriate reaction using the correct biomechanics, optimal positioning and other tactics. You learn things like methods of evasion and other body movement. You learn to use friction grips. You develop a productive flinch reflex. You learn angles of deflection. You learn principles of stepping. You learn not to over-commit to techniques. You strive to eliminate inefficient movement. You learn concepts of "flow". You learn when to tense and when to relax. You learn the secret of an effective guard. And so on.

There are so many, many principles that drills develop that this list is endless. And these principles are conveyed in martial drills in much the same way as principles are conveyed by tennis and golf drills. The difference is that in the latter activities, the drills also teach literal technique (ie. "if he does this, you do that"). This level of particularity is rarely possible in the exponentially variable and chaotic world of fighting.

Okay, so now I've learned some principles of unarmed combat. But which of these are going to teach me what I really want to know - how to time my strikes so that I "get" him and he doesn't "get" me?

Limitations of the stimulus/response model

Most drills today - whether traditional or eclectic/reality-based - follow the same model. There is an attack and you evade/deflect it, then counter. These are the classic "stimulus/response" drills to which I referred previously. They teach both evasion and/or deflection and counter attack. The counter attack should, ideally, be delivered as soon after the interception/deflection/evasion of the attack as possible (if not "simultaneously").

The stimulus/response model seems quite appropriate, particularly given my earlier discussion about the primary goal of civilian defence being "not to get hit". But, I hear you ask, why is it that this still doesn't really provide me with a ready "bridge to reality" - however much contact and protective gear my partners and I might use? Why is it that these drills still fail to give me a satisfying, scientific approach to "hitting him before he hits me"? After all, if we are honest with ourselves we will acknowledge that the moment we go into sparring, most of these drills never see the light of day - neither literally nor "in principle". In other words, at the first sign of a dynamic environment - of chaos, if you will - the drills are forgotten, and people default to "faux boxing".

And it's no use asserting (as one fellow said to me in one of my radio interviews) "that's because I'd have to kill you". Yeah - right! This is nothing if not a a big cop-out. After more than 30 years I know well enough when someone is in position in sparring to have dealt me a serious injury, but he/she chose not to. And I know all too well that many "grand techniques" taught in drills fail to make an appearance in sparring - even in "principle". Sparring might not be real fighting - but if you can't even start to apply a principle from a drill against someone who is "play fighting" you, what chance do you have when he is really fighting you? We're back to the start: what the heck do you do if the board hits back?

"Standing starts": the fundamental problem

The biggest issue with stimulus/response drills is that they usually start from a stationary position. This is quite understandable. Confrontations have to start somewhere, and often it is with the typical "push and shove" which are largely (though not necessarily) committed from a "standing start".

While it is certainly important to learn how to deal with "standing start" attacks, it is worth remembering that the "push and shove" comprises only a very small portion of most fights. Yes, the initial attack can often determine the outcome of the conflict, but regardless we need to keep in mind that the first attack occurs in less than a second. So even if you fight only for another 30 seconds, this will mean that the "standing start" portion will comprise less than 1/30 of the total "fight time". Yet many martial arts drills are focus exclusively on the "standing start" component.

Furthermore, while the first blow can often determine the outcome of the fight, you have to remember that you might not be the one to land it. You might be the victim of a "sucker punch". Or your attacker might be faster. Or you might just be "caught napping".

Even if you get the first punch in, you can't rely on it being determinative. It might not do the kind of "damage" you imagine. I have personally hit an attacker, only to be surprised at how little it did to "stop" him. The grand theories of dim mak (vital point striking) and "ikken hitto" (one punch, certain defeat) all seem to fall by the wayside when you are facing a determined, powerful opponent.

Don't get me wrong: I'm more than happy to train so as to deal with the first attack decisively. Stimulus/response drills are excellent for this. They develop the right flinch reflex, to some extent they groove automatic responses that can function despite the "adrenaline dump" etc. But I wouldn't be putting all my eggs in the basket labelled "I'll always beat him on the first attack".

My fundamental question remains: what is it that will give me the skills I need to ensure that if I can't end the fight on the first attack, I will on the second or third? What drill can teach me better timing than my opponent - ie. optimal responses in a dynamic environment?

Side-track No. 1: denying late initiative

As I discuss in my article "The flinch reflex" I think it is likely, if not inevitable, that during an attack you will be "surprised" enough to be "on the back foot". In other words, you will have to deflect, evade or both - and only then counter. A "simultaneous" and decisive defence/counter won't always be open to you.

I hold this to be self-evident. None of is are perfect. No training device can help you avoid completely the possibility (if not the probability) that you'll be dealing with an attack using "late initiative" and that you'll have to "claw back a position of dominance" after that.

Yet many martial artists today will imply the opposite. They assert the (obvious) preferableness of maintaining a position of dominance from the start - as if this assertion were some kind of guarantee that your preference will be realized. I hold this argument to be nothing more than an obfuscation.

The question I posed remains: how in the world do I learn to hit him before he hits me? Okay, you've shown me how to do it from a "standing start" where I have sufficient opportunity to "seize initiative" immediately. But what if I've just been surprised or overwhelmed? How do I "claw back a position of dominance" if a melee exchange has already begun and I'm not dictating the terms (shutting down his attacks, etc.)? What element of "timing" allows a good fighter to "turn the tables" where others succumb?

"Standing start" drills don't even begin to address this - especially when they assume that you'll always be able to "seize the initiative" from the "get-go". Rather than answer this question, the "late initiative deniers" are simply avoiding it.

Side-track No. 2: string attacks

Many martial artists put all their eggs into another basket, namely what I call "string attacks".

Typically string attack drills (also called "complexes" or "chain attacks") involve a furious response to an initial "standing start" attack - ie. a string of counter attacks that overwhelm or "bulldoze" the opponent into submission.

Like the "late initiative deniers", some proponents of string attacks assume that initiative can always be seized for the outset. Accordingly all of my comments in the previous paragraphs apply equally here. Certainly "string attack" drills don't address the question of "turning the tables" or "clawing back" initiative.

Otherwise, these drills are very impressive. But, like my old man's bag punching, impressive doesn't necessarily equal applicable against a "live" target. In my article on "string attacks" I discuss issues with these drills, the foremost being lack of predictive ability: The drills often assume a stationary opponent, where in reality, a resistive opponent in dynamic environment is going to be moving - doing anything but standing there, taking your "string attacks" on the chin, nose, groin etc.

American kempo is big on string attacks or "complexes"

And not only is the opponent moving, but he will almost certainly be actively resisting your attacks. In other words, he will fight back. By now it should be apparent how I came to choose the title to this essay; the "board" won't just stand there. It will "hit back".

The most sophisticated string attack drills attempt to address this issue by "shutting down" responses. And I have some time for this. Such drills can be an invaluable part of civilian defence strategy. But in reality, chaos theory should tell you that you can't predict with any certainty that you'll be doing "X" at "Y" time during a fight. You'll be lucky to predict the next move, never mind the next 10. Even the best designed string attack drill can't guarantee any kind of "shut down". Different people will respond in different ways. Chaos theory tells us this.

To give you just one example, I once had a student in a sleeper hold on the ground. He was face-down and I was on his back. Unbeknown to me, the student was a contortionist; he bent his back double and cracked me on the temple with his heels - sharp blows that had me lying dazed on the ground next to him a second later.

And you needn't just look to the above, extreme, case for evidence of "predictive failure". Because I have yet to see a "string attack" drill that featured optimal responses to your string attacks. Most of the responses built into the drill are token. Others are realistic enough - but certainly not optimal.

Attack as the sole form of defence: what's wrong with both side-tracks

By now the central problem with both of the above "side-tracks" should be obvious: they base an entire strategy on attack. In other words, they use attack as defence. Blocks/deflections and evasions - anything that is not directly part of the attack - are denigrated as "impossible to apply" or "inefficient" or "ineffective" or "vastly inferior technology". Both the "late initiative denier" and the "string attack proponent" argue that "attack is the best form of defence".

I can't help but think back to my old man, slugging away at the bag. The only thing that is different is the form of drill. Otherwise the core premise is the same: just concentrate on hitting hard and hitting fast. Everything else will work itself out. But it won't. All the above approaches are valid and indeed vital forms of training. But the fact remains that boards, bags, "standing start attackers" and "string drill victims" don't hit back. But real attackers do.

So can we identify the "missing link" that helps us address this issue?

Rephrasing Bruce Lee's quote

I mentioned at the start that civilian defence had a subtly, but profoundly, different goal to sport or military defence strategies. Instead of "hurting", civilian defence focuses on "not being hurt". And that, dear reader, is the key to resolving this conundrum.

In this article I have deliberately done what most martial artists do: focus on attack - attack from a "standing start", attack as defence, attack in a dynamic environment and a counter to an attack.

But the real issue, one that has slipped into the background, is defence - both your own and your attacker's.

I think that the best way to express my point is to rephrase Bruce Lee's initial quote. What if he hadn't said what he did. What if, after O'Hara broke the board, Lee had slowly, and emphatically, said:
    "Boards... don't thwart attacks."
The actor Bob Wall probably would have burst into laughter. And the dialog would have been truly crappy. But maybe, just maybe, Lee would have said something that wasn't trite (for a change).

Because the issue isn't that dead targets "don't hit back". The problem is that they don't stop you hitting them. In other words, what stops you applying your bagwork and other drills to resistant opponents doesn't depend on whether your attacker is going to hit you; that issue impacts on your defensive strategies - not your offensive ones. The main thing that differentiates your ability to attack "dead" targets as opposed to "live" ones is that the latter generally don't let you hit them! In other words, they resist or thwart you with defensive measures as much as they might attack you.

Sure, some attacks function as both defences and attacks "simultaneously". But not all defences and counters are "simultaneous"; watch any fight - on a street or in a ring - and you'll see many, many purely defensive measures. And, as I've discussed previously, even when defences and counters are "simultaneous" the defence component will almost always precede the attack, if even by a millisecond. In other words, to understand the dynamics of a fight we need to understand that defence and attack can't just be lumped together. Conflating them simply buys into the sort of false assumption propagated by "late initiative deniers" and obscures important data.

So it is worth repeating: "hitting back" isn't what stops you from landing your punches against a resistant opponent. It's a problem, for sure. You have to watch out that you don't get smacked. But what stops you from landing your punches is the fact that your opponent is not letting them land! It's really that simple.

Accordingly, if you want to land blows on a "live" target, you need to understand how your opponent will thwart them. Inevitably, this will be followed or accompanied by a counter. But the first thing your opponent will do in a dynamic, melee range environment will be to thwart your attack. He/she will stop it, evade it, deflect it or neutralize it. The counter comes second - even if by a millisecond.

Again, I'm not talking about your opponent simply hitting you first; as I've said, that is a question of your defence failing (whether this is because you launched an inappropriate attack or otherwise). I'm talking about your opponent using a defensive measure to prevent your attack succeeding. And that, ladies and gentlemen, is the crux of the problem. It is the "forgotten elephant in the room".

They say that if you define a problem, you are half-way to solving it. So what's the other half?

A part answer: drills that address what happens when your attacks fail

I don't have a panacea for the eternal question of how to bridge the gap between the dojo and the street. But I have at least one method of addressing timing in a melee exchange. It might not be much, but it is a method that trains appropriate, even optimal, reactions for when your opponent has thwarted your attack. What is it?

I have outlined this approach in my article "Really USING your kata". Essentially it is (somewhat strangely) the exact opposite of the "decisive", "bulldozing" drills where you overwhelm your opponent. It involves drills where your partner thwarts your attacks - and thwarts them in the most optimal way. In other words, it involves putting you in a situation where your attack fails.

The next part of the drill is just as important. You need to find the way to thwart your opponent's (optimal) counter. Just because your attack has failed, doesn't mean that you have to get hit, or that you can't turn the tables (ie. "claw back a position of dominance"). In doing so, your response also needs to be optimal (because there's no reason to groove sub-optimal reactions).

In fact, you don't want either side "losing" - because that side will be knowingly launching a doomed attack; he/she will be "learning to lose"! In order to work, both attacker and defender must have optimal responses.

What should be obvious by now is that this set-up poses one immediate "problem": how can such a drill ever end if both sides are thwarting attacks and launching optimal counters?

The answer, as I detail in the article referred to above, is to create a "looping" drill; one where the sides swap - ie. what I've previously described as a "rock, paper, scissors" format. It is a format that is as old as the most ancient martial art of the Far East - xingyi (see my article "Cracking the xingyi code").

I can see why this isn't popular. It doesn't look "sexy". It isn't impressive. Next to a "reality-based" practitioner's simple "standing start" drills it looks overly complex, detailed and pre-arranged. Next to a "steamrolling" attack string it looks positively pedestrian and contrived. Generally speaking, it looks "unrealistic" and formal.

But I really don't care about "appearances". I don't care that it is "unimpressive" compared to other, simpler drills. It doesn't teach literal techniques. It teaches principles - principles that are not being taught anywhere else. It teaches you principles that apply when your attack fails.

Our gekisai embu which teaches you principles relevant to when your attacks fail


I started this article by posing a question: how do you train the kind of timing that allows you to land blows on your opponent, but not to have blows land on you? How can one train to do to a "live" target what one typically does to a "dead" one - be it a board, a bag or a shield?

We all know that striking things is important. So is sparring. We know that "standing start" drills can teach us vital principles, if not literal techniques. Furthermore learning consecutive "string attacks" or "complexes" gives you great practice in combinations.

But overwhelmingly today, the trend is toward training attacks. Attacks against bags, attacks as defences, attacks in combination. This is great and I'm not saying this isn't important. But what's missing? Defence is what is missing - in particular the defence required immediately after your own attack has failed.

A few weeks ago a fellow wrote to me by email and asked the following question:
    "What is it that you exactly teach? In your articles, you talk about "Civilian self defense" and how Karate is good for self defense in most cases i agree, but in your video's i see you doing ( Very good) traditional Karate. What i don't see is scenario training, like one student is the attacker and starts of with verbal agression, then starts pushing and then takes a swing at the defender of if a guy grabs you by your clothing and slams you in the wall, that kind of training is for me a must for self defense training."
What this correspondent wanted to know is why I don't appear to promote "standing start" drills of the kind one sees in "reality-based" schools (I do practise them - I just don't push them to the fore in my videos or in this blog). The answer, as you can see, is not a short one. Why? Because in explaining my perspective I've had to address the many, many unwritten, unspoken assumptions martial artists routinely make.

The truth is that if you want to do to a "live" opponent what you can do to a "dead" target, you need more than attacking skills. You need to know what to do when your attack fails. For this you need drills that have you both attacking and defending. And your partner needs to be doing the same thing.

It might not be "satisfying" going without a "finishing blow". But there are other drills for that. "Finishing blows" are, after all, attacks. At the Academy of Traditional Fighting Arts, we train them every day, against bags, shields, in kata, in bunkai, in "standing start" drills, etc. On the other hand most martial artists never train to deal with the failure of their attacks.

And this is something you have to do if you ever want to figure out why your carefully wrought attack plans don't play out in sparring or in the street. You can't just ignore this issue, hoping that it will somehow work itself out through a mixture of bagwork, sparring and attack-centred drills. You might fumble your way to a part-solution through the school of hard-knocks, but it sure as heck ain't a scientific or efficient method of training. If you're serious about it, you need to tackle the issue head-on.

Because boards... don't thwart attacks.

Further reading: Pre-arranged Sparring: Definition, Purpose and Value by SooShimKwan

Copyright © 2011 Dejan Djurdjevic

Friday, April 8, 2011

“Boards don’t hit back”: Part 1

The missing link between practice and application


There is a famous scene in Bruce Lee’s “Enter the Dragon” where the character O’Hara (played by Bob Wall) holds up a board in front of Lee’s face and breaks it with a punch. Lee stares back unblinkingly and says, slowly and emphatically:
    “Boards… don’t hit back”
That line is a fairly typical example of Bruce Lee’s philosophy. Nowadays people would say that it was trite. However in its day the statement was quite novel, at least in the wider public’s eye. You have to remember that up until the end of the ’70s traditional Asian martial arts were regarded by many in the West as exotic and mysterious - if not supernatural. People were deeply impressed by traditional demonstrations. And board-breaking was common in these, especially in the case of karate which was “king” of the martial arts until Bruce Lee started everyone “kung fu fighting”.

Of course, even back in the early ’70s experienced fighters knew what most people know today: hitting a “dead” target, however forcefully, doesn’t necessarily mean you will be able to hit someone who is actively resisting your attack. In order to be a good fighter, you need to do more than practise hitting things that “don’t hit back”.

“Boards... don’t hit back” - Bruce Lee’s famous line from Enter the Dragon

And so, the question still arises for martial artists of all styles and methods: What should you practise in order to bridge the gap between the dojo and the street?

The goal

In order to examine this question, we first need to define the goal shared by civilian defence oriented martial artists. It is, I suppose, quite obvious, but it is still worth articulating clearly. The goal is not to get hurt.

As I discuss in my article “Civilian defence systems”, this is subtly, but significantly, different from combat sports (where the goal is "to win") and military engagement (where the goal is to "neutralise").

In relation to both sport and military defence one could broadly say that the objective is "to hurt", rather than merely "not to get hurt". Yes, civilian defence might well necessitate "hurting" your opponent. But it isn't your goal. In civilian defence, you "win" if you walk away unscathed - even if your opponent does too. If you can talk or run your way out of trouble, so much the better. If not, then you do what you have to do (but no more than that, according to the laws of most modern societies).

To the extent that this distinction matters, I'll expand on this concept later. But for the time being however, let us assume that there is no real difference between civilian defence and sport or military defence. Let us assume that the common goal is "to strike and not be struck".

Training for the goal

So now that we have established our goal, what do we need to do to achieve it? Clearly breaking boards and hitting targets generally is not going to "cut it".

Bruce Lee's statement gives you an insight into his "solution". Boards might not hit back - but you should. Lee had a very proactive system of civilian defence which he called "Jeet Kune Do" meaning "the way of the intercepting fist". The philosophy behind his system was/is that you should intercept your opponent's attack with your own. In other words, it was geared at "hitting your opponent before he hits you". As Lee once said: "If someone grabs you, punch him."

Given his famous speed in movement, it is unsurprising that this should have been the cornerstone of Bruce Lee's system. But what does this mean for those of us who don't have natural ability in speed? How do we train to "hit first"?

Sports scientists will tell you that no matter how hard you train, you can't change the nature of your muscle fibres. Some people are born to be fast, others for endurance, others for nothing in particular! Yes, you can improve your physical strength and fitness, and this is important. But unless we're comparing a couch potato with a professional athlete, the differences in speed attributable to training are fairly minimal.

Put another way, you should do what you can to be better, however in the end you can't change what you were born to be. Unless you're born with the right muscle fibre you're never going to be as fast as Bruce Lee.

Furthermore there is little you can do about improving your reaction time. It is what it is. And it declines with age. If you want to test your reaction time, try this website.

Developing better timing

So if hitting faster isn't your training priority what is? I suppose the answer to that is "hitting smarter". This means improving your timing. In turn, this means grooving appropriate reactions within whatever reaction time, and with whatever handspeed, God gave you.

We know good timing when we see it: 2 people are in a fight, one takes a big swing, but the other uses a straight punch that lands before the swing is even half way to its destination. Or one dives at the other's feet only to find himself pushed down and placed in a disabling lock. Or in a brutal exchange of punches one misses each time while the other inexplicably connects each time, leading to a knockout.

But this still leaves the big question unanswered: How in the world do I go about improving my "timing" so that I "get" my opponent but he doesn't "get" me? Haven't I just restated my goal and given it a clever label - "timing"? There has to be more to it than that; my training depends on it. I want to know how I can improve my "timing" so that mine is better than my attacker's. I want to know how to make my reactions "appropriate" and, ideally, optimal.

Bags and shields - the usual first "port of call"

Many people leave this question unanswered: they go straight back to the bags and shields. My late father used to roll his eyes when I mentioned my karate training.

"Humph - karate," he would scoff as he donned his mitts and went out to where the heavy bag was hung in our back yard.

There, he would pound the bag with blows that seemed powerful enough to fell an ox.

But as useful and important as "hitting things" is, that doesn't get us any closer to answering our question. Because, clearly, bags and shields "don't hit back". What was my old man missing?

Sparring and the "dyanmic environment"

Typically most people's answer to this conundrum lies in sparring - what is sometimes called "live training", or training against "resistant partners".

In sparring you face your partner and start "fighting". Not "real fighting", but a kind of "play fighting" where the rules of engagement are sufficient to protect you both from real injury, while leaving sufficient physical stress to "test your skill".

Setting aside the obvious points that people make ("sparring is not real fighting" or "your sparring doesn't have enought contact", etc.), I think it is safe to say that, generally speaking, sparring is an excellent training device. It gives you a chance to apply some of your skills in a dynamic environment.

I stess "dynamic" because this is a very important point.

Most training drills have a high degree of predictability. They start from a stationary position and end in a stationary postion. In reality, fighting (of any kind) is a process of constant, chaotic change. Usually both (or if there are multiple opponents, all) of you are in motion.

More than that, you or your opponent(s) might be accelerating. In other words, you don't just have the vectors and velocity of your own movement and your attackers' to consider, but also the way in which these vectors and velocities are changing. Last, you don't know how or when new vectors and other variables might come into play (including such things as obstacles, terrain, etc.).

In short, nothing about real fighting is predictable. In a drill, barring any silly mistakes or carelessness, the situation is entirely predictable.

So what is sparring useful for? You can take the skills you've learned in "static" (ie. non-dynamic) drills and see if they will work (at least in part) in a dynamic environment - ie. where the board you're "breaking" not only moves, but also "hits back". Even if it isn't "real fighting" it gives you a test environment that has at least some level of unpredictability.

So far, so good. But the fundamental question remains unanswered: how do you develop the "skills" that sparring tests?

Sparring tests skills - it doesn't teach them

You really do need a certain level of skill before you start sparring. Yet most folks I know in the martial arts start sparring from day one, before they have learned a single fighting technique or principle. God only knows what they are doing when they spar - but it usually isn't "skillful". [In the Academy of Traditional Fighting Arts we don't start free-sparring until a certain basic level of skill is acquired - I'll expand on this topic another time.]

Invariably beginner sparring comprises an awful mess of inappropriate reactions. There's nothing being "tested" here except how good or bad your "base-level" reactions are. Some folks have a better base-level reaction than others. Some folks are "naturals". But in the end, you need to learn some skill before you start "testing" anything.

If you doubt me, take 2 people who've never played tennis before and put them on a court. Watch the resulting disaster and ask if it bears any resemblance to what you might see at Wimbledon or Roland Garros. Why should martial arts be any different? My tennis coach had me training for a full 2 months before he said I was ready to start playing tennis games. The same will apply to learning golf, ballroom dancing - any physical activity that requires skill and coordination. Yes, some beginners might be able to "jump right in" and look "okay". Just. But what they exhibit is just a naturally high "base-level". And it is still a far cry from what is appropriate, never mind optimal, movement.

And before you say that comparing fighting and sport/dance is like comparing apples and oranges, think of this: sport movements are inherently limited. You have a ball and/or racquet and a limited number of strokes/movements. In hand-to-hand combat the variables are exponentially greater. The greater the number of variables, the greater the need for training in fundamental skills before they are "tested" or applied. Ask a fighter pilot how much training he or she has had to undergo to prepare for real combat. Hand-to-hand combat might not be fighter plane combat, but it sure as hell ain't tennis neither.

What confuses people is that size and brute strength play a much bigger role in fighting than in a sport like tennis. But what if your opponent is at least as big and strong as you are? Then you need skill in timing and technique - ie. the appropriate or optimal reactions we discussed earlier. The bigger and stronger your opponent is, the more skill becomes essential.

So, in summary, sparring is a useful and, I think, necessary training device. But it doesn't teach you the appropriate or optimal reactions. It merely tests them. How do we learn the right reactions?

Drills - a forum for learning fundamental skills

The answer for most people in both the traditional and modern/eclectic martial arts or systems lies in drills. What are these?

A blocking drill from goju ryu karate

In their simplest sense, drills are means of inculcating a particular, appropriate move until it is reflexive. Solo drills typically involve a person repeating a movement, in a very particular form, again and again.

Two person drills however involve stimulus/response training: one person provides the stimulus, the other the response. Typically the drill loops so that a sequence is repeated - either with both sides doing the same action over and over, or with both sides swapping roles.

The concept is that a particular movement is taken out of the relevant art/sport/activity (eg. grappling) and isolated into the drill. But is this strictly true? In martial arts and systems of all descriptions there are countless drills that bear little or no resemblance to the actual activity. What's going on here?

Sometimes the difference is because the portion of the movement taken from the activity is very small. So, for example, some drills in Brazilian jujitsu involve a discreet ground movement repeated endlessly because it is a critical point in maneuvering into the guard. On its own it has little resemblance to the activity, but when seen in context, it does.

The similarity of drills to their eventual application

However in many other cases there is little to no apparent nexus between a drill and the "real thing". Examples range from the speedball in boxing through to traditional "blocking" exercises in karate. Even shadow boxing seldom looks anything like what happens in the ring (as much as people would have you believe otherwise). Look at Muhammad Ali's shadow boxing in preparation for the "Rumble in the Jungle" with George Foreman and then compare it to the untidy, sluggish and ultimately chaotic rope-a-dope tactics he actually employed. What accounts for such differences?

The reason comes back to the number of variables contained in the relevant art/sport/activity. As I have foreshadowed, martial arts and sports have many more than ball sports.

Generally speaking, the fewer the variables involved in the activity, the closer the drill movement will resemble its manifestation in the activity.

An activity like golf, difficult though it might be to master, has very few variables. For example, the ball is stationary; all you have to do is swing at it. No one is trying to thwart or otherwise obstructing your swing. Though you are competing against others, when it comes to the drive or putt it's just you and the ball. Accordingly it should come as no surprise that the "drills" you employ at the driving range or on the putting green comprise exactly the same movements you employ in a game of golf.

With something like tennis, a small disparity between drill and activity starts to creep in. Tennis has greater variables than golf, but they are still comparatively small, and accordingly so are the disparities. There is only one ball in play at any time. It takes a certain amount of time to cross the net. You and your opponent may only employ one arm and in doing so you can only use the tennis racquet (not your forearm or hand). Only a limited number of strokes are feasible or possible (ground stroke, volley, half-volley).

As a result the drills my tennis coach had me practising were pretty darn close to what happened in a game. I'd have to return repeated ground strokes from the baseline, or I'd stand at the net and volley repeatedly. These drills resembled the movements employed in a game however there weren't identical (as per golf). Rather there were important differences: I knew my coach would be giving me appropriate shots for the baseline or the net. Accordingly I wasn't forced to make any real guesses as to my court positioning, other than a little shift to the left or a little shift to the right. And my stroke was pre-chosen, leaving me with a choice of forehand or backhand only.

Martial drills

By the time we get to martial activities the variables increase exponentially. You're not dealing with a handful of racquet strokes and a ball travelling on a fairly limited range of trajectories over a net. There is no defined "start" or "finish". And in true civilian defence encounters there are no rules. Rather, in fighting any portion of the body can do anything at all at any time and from any angle. There might even be more than one opponent.

Accordingly I think it should come as no surprise that combat drills might look far less like the actual activity. After all, there is no such thing as a "typical fight". Given the almost infinite variables it is very hard to isolate drills representative of something that might actually happen in the particular form and sequence. This contrasts heavily with golf, where executing a drive or putt will happen in precisely that form and sequence, albeit with different terrain. Or tennis, where you might well drill a volley, ground stroke or overhead smash in more or less the same form as you would execute it in a game.

Again, compare this to your martial arts training: when is the last time you executed a kata application in free-sparring in anything approaching the "when he does this, you do that" methodology employed in standard bunkai study (typically involving a single step and a single attack)? I'll wager that your answer is "never".

Yet this does not diminish the importance of drills. Professional combat sports practitioners, the military and pragmatic or "reality-based" civilian defence instructors would not practise them if they weren't an effective training tool. When training for a complex activity such as fighting, a drill used in that training need not look exactly like the activity itself in order fulfil a useful (if not vital) purpose.

Continued in Part 2.

Copyright © 2011 Dejan Djurdjevic

Tuesday, April 5, 2011

Body movement in kata - what does it mean?

Embusen: a kata's movement "floorplan"

“Embusen” is a Japanese term that literally means “demonstration line(s)”. It is commonly used to describe the pattern of lines along which a kata is performed.

So, for example, the kata pinan nidan / heian shodan has a sideways “H” embusen (meaning that when performed, the kata will, broadly speaking, trace an “H” - or if you prefer, a capital I with serifs - on the ground).

The kata shisochin has both an “x” and a “+” embusen (ie. it moves in 4 directions from the centre - which fits well with its name, meaning “4 direction battle”).

The variations on embusen are almost endless - particularly when one looks at other arts.

The “embusen” of the internal art of xingyiquan is essentially a straight line forward and a straight line back. Baguazhang’s “embusen” comprises endless circles, like eddies in a pond or stream. I even know of one bak mei (white eyebrow) form where the “embusen” traces the Chinese character that is the name of the form.

[Note that for the purposes of this article I shall, for convenience, use the term “embusen” even when referring to Chinese arts.]

The significance of embusen

So traditional martial arts are often defined by the embusen of their forms. But what significance does embusen have for kata/forms?

Goju ryu founder Chojun Miyagi is quoted as having said: “Don’t be fooled by the rule of embusen”. By this I think he meant to say that one should not put too much stock in the pattern of lines drawn by a kata. I would hazard a guess and say that Miyagi’s reasoning was as follows:

Embusen functions principally to maintain symmetry and allow for moves to be repeated (for the sake of practice).

A well-designed pattern also allows for a fairly complex and lengthy sequence to be practised in a short space. So, for example, xingyi and xingyi-derived forms can require an inordinate amount of practise room because they follow a straight line “embusen”. Consider the form Da Peng Zhan Chi which is modeled on a xingyi platform: it requires the entire diagonal length of our dojo (about 12 m), and even then the movements require a little “shortening” to fit.

Da peng zhan chi - a xingyi-related form of Hong Yi Xiang which has a linear "embusen"

Baguazhang has solved this particular issue by being practised in a circle. It is my view that this is the principal reason for the circular practice of bagua - not any particular application; the circle allows continuous stepping without the need for an enormous practice space. Indeed, bagua can be practised in a straight line if you so choose - consider the Gao style of Zhang Junfeng, for example.

So embusen is pertinent to kata/forms for pedagogic reasons. But does it also have some function in relation to application?

The difference between kata embusen and body movement in kata

Arguably embusen plays no role at all in application. It is purely a pedagogic device. However it is important here to distinguish the overall pattern traced by the performer of a kata and the body movement, if any, employed in individual techniques (eg. whether a particular kata technique is performed with a step forward, backwards or sideways or during a turn). I shall call the latter “body movement” ("taisabaki" in Japanese). I could have called it "evasive movement" ("tenshin") since a substantial amount of kata movement is concerned with evading attacks, however it is often more about "intercepting" attacks rather than simply evading them - more on this later.

While the overall pattern of your embusen has no significance to application (bunkai), the individual movements comprising the embusen were, in my view, chosen with care: in other words, they are relevant to the application of the particular techniques because they are part of what makes them applicable.

So if body movement comprises a step forward, I think it is self-evident that the primary application intended by the kata designer is an “entering” type application; in other words the kata designer intended that the technique be applied with a “forwards evasion” or interception. Correspondingly, if the body movement comprises a step back, then the primary bunkai involves a backward evasion. [By “forwards” or “backwards” I don’t necessarily mean “in a straight line”. I shall expand on this in a minute.]

If the body movement includes a turn, then some aspect of body turning (either with or without a step) will be involved. Again, as I shall expand upon in a moment, this needn’t involve a full turn as executed in the kata, but will usually involve a smaller movement.

Forward body movement

Most karate kata feature forward stepping. The same is true of karate’s southern Chinese counterparts. Even the northern shaolin schools and the internal arts of xingyi, bagua and taiji feature very little “retreating” movement.

So how is a forward step to be interpreted?

It is my view that “forward steps” comprise 2 types - movement in attack and movement in defence. The latter usually features both a defence and a counter-attack.

A forward attack on its own is self-explanatory: you want to use your bodyweight when attacking, so this means either lunging or stepping forward.

The forward defence is a little more complicated. Why would you defend by moving forwards? The answer lies in catching the attack early: by moving into your opponent you are trying to intercept the attack before it has reached its full speed. The earlier you can catch the attack, the slower its speed and the smaller the angle needed to deflect the attack. Some people call this approach “slowing down the attack”.

Typically a forward defensive step is not straight into the attack, but off-line. So where a kata features a forward step with a block, it is usually safe to assume that the step should be applied at an angle towards the front.

I typically refer to such a step as “45° forward” because it is a generic descriptor of an off-line forward movement. But in reality the angle might be 10° or 15° or 30°, etc: you choose the finest angle that will permit you to deflect and evade the attack while putting you in a position to counter. The smaller the angle, the better your position for an immediate counter, but the greater the risk of your deflection/evasion failing.

The advantages in “evading forwards” are obvious: it not only “slows down” the attack, it also places you within immediate range for a counter (where evading backwards might put you out of range, requiring you to move forward again to counter).

It should come as no surprise therefore that the overwhelming majority of body movements in kata comprise forward steps. In the video below I give 2 examples of such body movement, namely the penultimate technique from saifa where you step and pivot (see 0:59 to 1:23) and the penultimate technique from shisochin where you step and pivot (1:24 to 1:46)

A video where I discuss body movement in kata and its relationship to bunkai (applications)

Backward body movement

Despite the fact that “forward evasion” is more advantageous, sometimes kata will feature a retreat instead. Why? Backward evasion might not confer the same advantages as forward movement, but it is a fact of life. As I discuss in my article “The flinch reflex” it is human nature to withdraw from danger. While we can train to modify our flinch reflex so that it propels us forward, a surprise attack can often put us (literally) “on the back foot”. In other words, the lateness of your own reaction might leave you with no option but to evade the attack by moving backward. Moving forward at such a point might simply put you directly into the “firing line”.

There are 2 principal reasons why people suffer this kind of “late initiative”. Beginners typically haven’t grooved both a quick and productive reaction to an attack. However even very experienced fighters can be surprised; surprised in such a way that backward evasion is their only option.

Moving away from an attack means that you give the attack the chance of reaching its full speed. But if the attack is pretty much at full speed at the moment you react to it you don’t exactly lose anything by moving away from it. In fact, you can use this to your advantage. Because by allowing your attack to “speed up” you can “exhaust” its forward momentum - then counter when the attacker is “resetting” him or herself.

Rarely is such an evasion effected straight back. To do this will simply permit (or even encourage) the attacker to continue to attacking along the same line. Rather, as with the forward evasion, you should step off-line.

Again, I typically refer to such a step as “45° back” because it is a generic descriptor of an offline backward movement. But in reality the angle might be 10° or 15° or 30°, etc: you choose the finest angle that will permit you to deflect and evade the attack while putting you in a position to counter. And again, the smaller the angle, the better your position for an immediate counter, but the greater the risk of your deflection/evasion failing.

Backward steps are found in a number of kata. Predictably, kata designed as more “basic” teaching tools (for example gekisai dai ichi and ni) contain a higher proportion of backward body movement. I don’t believe this is an accident. To me it is quite clear that the gekisai forms were designed by Chojun Miyagi with beginners (and the beginner tendency to react very late in the piece - probably with a withdrawal “flinch reflex”) in mind.

In this regard it is important to consider the gekisai forms in the context in which they were invented:

Chojun Miyagi and Shoshin Nagamine were respectively called upon to create practical beginner kata in the early 1940s, ostensibly/officially for teaching in schools and for the wider dissemination of the art (but in all probability, more as a reflection of Japan's increasing militarism during that era).

As a result Nagamine created a kata he called fukyugata ichi (meaning “a kata to be spread or disseminated - No. 1”). This kata’s embusen is based heavily on the pinan floor pattern, ie. a sideways “H”. But more importantly than this, the body movement is also taken from the pinan series: namely the defender moves directly into the oncoming attack.

This occurs right from the word go: The first movement is against an attack from the left. The defender steps with his left leg directly to the left, turns the hips and blocks the attack.

When faced with his job of designing fukyugata ni (what is now gekisai dai ichi) Miyagi also had the defender responding to an attack from the left. However his approach to stepping was significantly different. Rather than have the defender move his left leg to the left, his kata has the defender moving the right leg forward. In practical terms this amounts to a “slightly back” off-line step.

Why did Miyagi do this against the trend of Nagamine’s creation, Itosu’s pinan series and, indeed, every other goju kata (all of which overwhelmingly favour forward steps)? And why did he do so against the prevailing political pressure to create forms with “no retreat”? [I can’t help but notice that gekisai dai ichi even features a step forward at the end!]

I believe this was no accident. Miyagi wanted to create a form that was usable by beginners. To expect beginners to have quick and appropriate “forward-biased” reflex reactions is to expect too much. In my view Miyagi wanted something more immediately applicable.

Further to this train of thought, I observe that Miyagi also included a backward step into shiko with a gedan barai (a low sweeping block or strike). Again, I think this was no accident. When defending against attacks such as kicks (which are faster and heavier than arm attacks) beginners are prone to panic - and flinch away. Miyagi decided to harness this natural reaction and make the backward step the primary application.

A video in which I discuss the gedan barai uke in gekisai kata

The step in gekisai is to be contrasted with the similar movement in seiunchin - where first step into shiko dachi is forward.

Sideways body movement

Of course movement needn’t be forwards or backwards. It could be sideways. I won’t focus on this to any real extent because sideways evasion is, quite clearly, a hybrid of forward and backward evasion. It confers some advantages, but some disadvantages too. Like any other type of evasion, its use is determined by context.

For example, when facing a very fast, penetrating attack, you might have very little time to move despite reacting early. In that case moving forward or back off-line might be insufficient to take you out of the firing line. Instead you might suddenly need to increase the amount of off-line movement by stepping directly to the side (an angled off-line movement being clearly a smaller off-line movement).

Sideways movements in kata (such as the sideways step in cat stance in gekisai dai ni) are typically much larger than those needed in real civilian defence encounters, but that is the case with any kata technique in my experience: you practice larger movement because it is easier to correct and harder to perform.

After all, if you can effect a big step quickly, a smaller step will inevitably be much easier to perform quickly.

I discuss sideways evasion from gekisai kata

Turning body movement

Sometimes the body turns or pivots during the forward, backward or sideways step. [These are compound movements which I will examine another time.]

Indeed, sometimes there is no real “step” component to the body movement. In that case, the body turn or pivot is itself the principal body movement.

The most common example that springs to mind is what I call the “open door evasion” - where you turn your hips sideways. This can be done with a small step forwards or backwards - or, as I’ve foreshadowed, without any step at all. The turning of the hips sideways is the key. Such movements can be found in the gekisai kata, for example. Consider the following videos showing how this body movement can be applied:

I discuss the “open back” evasion from gekisai kata

I discuss the “open forward” evasion from gekisai kata

Another quote attributed to Miyagi is: “a turn doesn't mean you are facing a new opponent.” This has been interpreted to mean that turns are really nothing more than a device to keep you within the boundaries of the embusen pattern; to permit the equal repetition of techniques within a confined area.

There is a lot of merit to this thought, as I outlined at the start. But I believe there is more to turns than just this. How you turn holds important lessons because it often teaches you the kind of turning body movement to which I refer in this part of the article. Which leg does what in a turn is critical: you might never apply a body turn as a full turn-around (180° or 270°) but the exact same principle is applied to a smaller body turn - even something like 10°. In short, turns can be “slips”.

I propose to deal with the subject of "slips" at length another time.

So far, so good: but what if a kata has “no body movement”; ie. it features techniques that are executed while the performer is largely stationary?

“Stationary” techniques generally

The first argument I encounter when I raise the issue of body movement is that some southern Chinese and Okinawan forms have very little. Some comprise only a couple of steps forward and back, or side-to-side, with a whole lot of technique being executed while the performer is stationary.

Common examples would be the various sanchin/sanzhan forms of karate and Fujian province in southern China. Consider the following clip of yong chun baihe sanzhan, and in particular note the hand techniques executed while stationary from 0:40 to 0:50 - that’s 10 seconds of stationary technique:

Pan Shifu of Yong Chun baihe (white crane) demonstrating sanzhan. Note the (relative) lack of body movement, in particular the lack of stepping from 0:40 to 0:50.

This low number of steps in the above performance of sanzhan seems to be at odds with the reality of combat. So why is this form practiced with so few steps?

One answer to this is to be found in understanding the focus of the particular form: sanzhan is a foundational practice method that teaches such things as grounding. In order to have good grounding skills one must spend a lot of time practising “zhan zhuang” or “standing post” practice. Southern Chinese forms emphasize grounding much more than their northern counterparts which emphasize mobility, so it is hardly surprising to find that the latter feature forms with a great deal more stepping and body movement generally.

A typical northern form from taizu changquan. Note the focus on mobility and in particular the side-to-side movement upon which I shall expand in a moment.

Once foundational forms like sanzhan are mastered however, the southern Chinese forms become far more mobile. Consider the yong chun baihe form below and note the dynamic forward movement:

A dynamic white crane form. Note the forward movement.

How to interpret “stationary” techniques

So, assuming you have a form that is largely stationary, how, if at all, does one infer body movement from it?

The kata will usually provide a clue. In the case of sanchin/sanzhan (and the southern Chinese systems generally), the forms might have few steps. But they do have steps nonetheless. And these steps are generally initially forward steps (followed by either a turn and more forward steps or, occasionally, backward steps). Either way, the forward steps come first. This suggests to me that the “stationary” techniques were primarily intended to be applied with forwards body movement.

Second, the movements, while performed without stepping, might well involve weight transfer or body twisting of a kind that utilises a smaller, more subtle variation of stepping or turning body movement.

But what if there is no meaningful stepping - either forward or back - and no other, more subtle, body movement either? The most obvious example given to me in this regard is the kata naihanchi/tekki (which we call "naifunchin" at the Academy of Traditional Fighting Arts). This kata moves with side-to-side cross steps, but none of these are readily interpreted as being some sort of body movement intended to augment the techniques themselves. Mostly the techniques are executed in a stationary kiba dachi (horse stance). What’s going on here?

It is my view that in such a case the kata designer has essentially given you, the student, carte blanche to use whatever evasion you feel is appropriate: forwards, backwards, sideways, body turning etc. All of these are options and the kata designer is not specifying which you should use. The kata’s arm and leg techniques are capable of being executed on any of these “platforms” of body movement. I demonstrate this principle in the video below:

I demonstrate evasion using the kiba dachi of naihanchi kata

But the question remains, why doesn’t naihanchi feature some of this applied taisabaki? On one view it would be limiting. Because once you execute a technique with a particular body movement (eg. a step forwards) this becomes your focus - often to the exclusion of another movement (eg. a body turn). Could it be that a kata as fundamental as naihanchi (or sanchin) was designed not to be so “limiting”? I think so.

Southern Chinese quan fa is, in particular, very reticent about body movement for this reason and the other reasons I’ve discussed before (eg. grounding). The more fundamental the kata, the more the steps are reduced or abbreviated.

Some have argued that naihanchi is of northern origins and that such considerations do not apply. Indeed, when you consider the northern taizu changquan video I have embedded above, there is some superficial resemblance. Both forms move side-to-side and use horse stance. The main difference is, of course, that the “sideways” movement in the taizu form turns the whole body - so although you are moving side-to-side relative to someone watching from the front, you’re actually moving forwards and backwards along that line.

I think this is a fascinating topic that deserves more attention another time. Suffice it to say that even if naihanchi began life as a northern form, it has long abandoned its origins and subscribed to the southern paradigm of grounding and less movement. It is certainly used in this context by shorin-based karate practitioners.


Embusen of a kata is indeed primarily about facilitating your performance of certain techniques within a defined area. It also assists in enabling such things as technique repetition for practise. But the nature of the body movement in each technique, including the steps and turns, is not just a by-product of the embusen “floor pattern”. Every such movement must have some meaning, otherwise kata is pointless.

Why "pointless"? Because kata should be about putting techniques into a meaningful, dynamic context. Isolated techniques strung together consecutively create nothing more than a pattern for the sake of a pattern. If you want to practise isolated techniques, you might as well do so without arranging them into an elaborate (meaningless) sequence. Otherwise you risk setting yourself up for ridicule of the kind Baz Rutten effects below:

Baz Rutten ridicules kata

In other words, the body movement in kata must match the technique - it must be appropriate to it and complement it. It must make the technique work.

In this regard it should be clear that it is impossible to separate a hand or foot technique from taisabaki (the movement of the body): the 2 are inextricably linked. They can be separated in analysis - which might lead to "oyo" (extrapolations or variations of kata techniques) - but fundamentally they are equal parts of the same application.

So in order to understand the bunkai (applications) of your kata, don’t ignore the context in which they occur in kata. It’s all very well not to be “fooled by the rule of embusen” but it is quite another to be fooled by the assumption that the body movement comprising embusen is irrelevant to kata application. I think it is quite reverse. The taisabaki shown, hinted at or accommodated in the embusen provides the key to understanding kata applications in a dynamic environment.

Copyright © 2011 Dejan Djurdjevic