The importance of flow

Lessons from China

After my recent intensive training with Master Chen, a fellow karateka asked me the following:
    ”I hope this isn’t a stupid question, but did you learn anything of relevance to karate?”
My answer was, “yes”. There were many applications of forms and other techniques which I couldn’t help notice existed in precisely the same form in Okinawan karate.

But more relevantly, this trip has reaffirmed one of my core beliefs about the relationship between karate and Chinese martial arts – in particular the internal or “soft” arts: they are really part of the same continuum. What differentiates them on a surface level evaporates once one starts to use biomechanically sound and efficient principles of movement (see my article “Goju as an internal art”).

One of the most important of these principles is the need for “flow” or connectivity between techniques.

Many Chinese martial artists lament that karate seems to have lost any semblance of "flow". While this might be true of a lot of karate one sees today, I don't think this is necessarily a function of karate technique, but rather a function of the emphasis of individual karate practitioners.

What I mean by "flow"

It is important to note that "flow" need not take the form of a dance-like wu shu performance or perhaps the soft flow of taijiquan. Flow exists, for example, in the rather short, sharp and brutal internal art of xingyiquan. The flow in this art occurs between deflection and its related counter - not necessarily between longer strings of techniques (as is the case in taijiquan, for example). It has long been my contention that a "xingyi-type" of flow is meant to underlie karate kata - just as all martial solo forms exist to provide a means of learning connectivity between techniques (as opposed to the practice of isolated basics).

A video where I demonstrate karate techniques (taken from Aragaki seisan kata) performed using “xingyi-like” principles of “flow”.

What critics say

Some would disagree and argue that “flow” is a product of aesthetics more than function and can (and should) be dispensed with in practising an art for civilian defence purposes (as opposed to learning an art “for art’s sake”).

In this vein, a practitioner of Koryu Uchinadi (“KU” - the system of respected karate historian, researcher and innovator Patrick McCarthy Hanshi) said the following in response to my series of articles beginning with “Whole lotta shakin’: preloading the hips”:
    “Whether the [flow] approach is great or not it could be questioned how relevant 'perfect alignment' is for self defence (and thus karate) and I think it could be rightly stated that modern arts, including karate, have consciously chucked out these characteristics in favour of a curriculum drawn from a mindset based on... brutality and simplicity.

    In KU these exact lessons, as we know, are delivered through template based learning (kata).

    Template based learning means the emphasis lies on training defensive applications of its individual templates, without these being connected to the next template or move in kata, and certainly not with an emphasis on the flow between each template
    The templates are connected in a geometrical configuration to serve as mnemonics so the whole lot of techniques can be remembered more easily.

    For practising the flow of techniques or the 'dynamic environment of combat' as Mr Djurdjevic impressively calls it, we use a list of other exercises (not surprisingly called ‘flow drills’ in KU)

    The concept Mr Djurdjevic projects is clearly a different one (perhaps derived from the internal systems he also practices?) and I’m afraid that if he is wrong (which I believe he is) about his initial premise that kata is 'a method for practising flow of techniques', he’s not making a very good point.”
There is a lot to agree with in what the above correspondent says. Kata do indeed provide a template-based learning method – a mnemonic, or otherwise a means by which techniques are inculcated into your subconscious so that they emerge reflexively.

And I would agree that much of modern karate has “chucked out” aspects like flow in favour of “brutality and simplicity”. This is, I believe, highly representative of the “external” martial tradition, and the correspondent is right that I adhere to a more “internal” viewpoint.

KU yudansha Erik Angerhofer demonstrates Aragaki seisan kata as researched by Patrick McCarthy Hanshi. Note the powerful hip use – a trademark of the KU system.

Efficiency: why flow is necessary

So what is my reason for adhering to this “internal” paradigm (as opposed to a more “brutal”, powerful, straightforward approach)? It comes down to what I believe is a more efficient use of body mechanics. Consider the following animated gif as an example of movement using “internal arts type” flow:

In this gif I perform a series of steps taken from Aragaki seisan kata (as researched by Patrick McCarthy Hanshi). I have used this gif as an example because it provides a good contrast to how the same movement is performed in KU. You will note that in my version I use xingyi principles to connect the footwork and hand techniques into what is essentially one movement. I strongly feel that these xingyi principles are equally a part of karate training.

Note in particular the connectivity of between the deflection and the counter. As a matter of interest, the feature that most distinguishes it as an internal "xingyi-like" movement is this:
    The counter strike is completed as the front foot lands - not as the back foot slides up (or even later!).
[I encourage readers to watch the video embedded at the start of this article for a better view of what I am doing. I chose the series of frames for the above gif because the footwork was more apparent, however the video demonstrates more clearly, I believe, the way in which the "connectivity" harnesses the forward momentum of the step to add force to the counter.]

The adjacent gif demonstrates this flow applied against a partner using the closed hand version of the same deflection (the principle is the same). Again, note the connectivity between the deflection and the counter - how one feeds into the other, creating one seamless move using contextually appropriate1staged activation”.

Kata: a template for grooving connectivity without pressure

In my experience it is important to learn this connectivity in isolation (eg. in kata) because it is hard enough to learn it in the first place - never mind when someone is trying to knock your block off! A beginner under pressure will default to a more basic "1-2" - ie. a deflection, followed by a separate, disconnected counter. It is my strong view that "1-2" movements are really quite inapplicable in fighting - the pause between "1" and "2" gives your opponent time to move away from your line of counter - or just hit you! People don't "just stand there" as they often do in ippon kumite (one step sparring). This is what I mean when I refer to the "dynamic environment of combat".

Disconnections within a template: what happens when you dispense with flow

Just as many Chinese martial artists seem to have lost the concept of "kime" (focus) (see my article "Kime: the soul of the karate punch") in pursuing a kind of "flow for the sake of performance", so many karateka seem to have lost the concept of "flow" in their pursuit of kime behind each technique. In fact, both are vital.

I feel that the loss of "flow" is a principal factor in explaining why deflections (and karate techniques generally) are not applied by many karateka in sparring/fighting; the "1-2" approach of basic karate simply doesn't fit in a flowing paradigm of dynamic combat.

Compare the preceding gifs to this one of Erik Angerhofer, a (very able) KU yudansha demonstrating steps from Aragaki seisan (the same steps I demonstrate in the first gif). While I admire Erik's karate skill and greatly respect the research and depth of knowledge of the KU founder, Patrick McCarthy Hanshi, as mentioned previously I disagree with this approach to kata performance principally because of my view that the "internal arts" concepts of flow can (and should) be applied to karate. I believe the above gifs serve as good examples for highlighting the differences between our points of view.

Note in particular the pause between the deflection and the counter. The movement is clearly different from the internal arts in as much as the counter strike is not completed as the front foot lands. It is not even completed as the back foot slides up (which would be standard for karate).
    Rather, the pre-loading of the hip2 delays the strike until well after the back foot slides up.
From an internal arts perspective this pause disconnects these techniques into the "1-2" pattern. This is, I believe is inconsistent with the goal of providing a mnemonic for, or otherwise inculcating movements that comprise, "templates" for the dynamic environment that is combat (if one assumes, for example, that a template must consist of at least a deflection and its related counter - not just one or the other).

An internal artist would be inclined to view the kind of kata practise demonstrated by Erik in the above gif as less a "form" and more a sequence of isolated techniques - ie. a series of deflections and strikes that are not connected to each other.3

Accordingly, in answer to the above correspondent, I must stress that I am not talking about disconnections between templates (which might be an issue in, say, taijiquan or wu shu but is irrelevant to karate), but disconnections within the templates themselves.

Loss of speed and introduction of telegraphing

It is worth noting that when I created this gif of Erik from screen captures, up until the mid-point (ie. when the legs are crossing and the deflection is at the "interception point") I used the exact same number of frames (7) as I did for the gif where I perform the movement.

After the mid-point, the gif showing me performing the movement uses another 7 frames. The gif of Erik uses 14 frames after the mid-point - ie. it takes Erik exactly twice as long to deliver his strike after his deflection as it took me.4

Furthermore, in addition to grooving a slower response to an attack, this non-contextual hip-load also telegraphs your intention well ahead of time...

But what about "power"?

Ostensibly the KU approach is to give greater "power" (ie. force) to the counter. Leaving aside the extra time taken to load the hips and the fact that the load telegraphs your intention, I think that what force you gain with hip use, you lose (at least in part) through not using (sufficiently) the forward momentum of your body. In other words, while Erik is generating a lot of force, I'm not entirely certain that the full potential of that force (available in the context of the kata movement) is being harnessed or applied.5

My reason for saying this is that the failure to connect the movements into a cohesive whole interferes with the naturally occuring1 process of staged activation. Staged activation is essential to transferring momentum (ie. "impulse" - see my article “Hitting harder: physics made easy”) and accordingly increasing the force applied to your target.

Put another way, Erik's strike, while forceful, does not benefit much (if at all) from having a step preceding it; it is about as forceful as a stationary, isolated reverse thrust would be2 (which in Erik's case is clearly very forceful!). By contrast, in this kata move I prefer to use the contextual forward momentum of the step as a major component of the force.1

In this regard it is important to note the following:
    There is a difference between force generated by your body and the force applied to your target.
And on a related note, "forceful looking" is not always the same as "forceful in effect" (see my article: "Visible force vs. applied force").

Differences between "external" and "internal" methodologies

It is my view that the KU approach is representative of an “external” methodology. I have previously attempted to describe the differences between what are known as the “internal” and “external” martial traditions in China/Okinawa (see for example “Understanding the internal arts” and “Internal vs. external martial arts”). However the above gifs provide what is possibly one of the clearest examples of this difference.

The external tradition places emphasis on generating as much force as possible. This concept is instantly accessible/understandable to students and is also very quickly absorbed, lending its use to practical training for civilian defence.

The internal artist will attempt to apply force through efficient body mechanics, not through the “brutal simplicity” of generating more force. Now there are drawbacks to this approach, the most obvious being the time it takes to develop the subtle and sophisticated kinaesthetics that underpin efficient body mechanics (hence the saying in taiji: “15 years before you leave the training hall”). On the other hand, with any physical activity there is a point at which mastery takes the form of both an art and a science.

For example a good golf swing requires subtle kinaesthetics utilising staged activiation. It would be pointless hacking away at the ball with greater and greater force without coaching in the art/science that underpins golf. Ultimately a good golf swing relies very little on simple force. Rather, an efficient swing can appear relatively effortless, yet propel the ball straight down the length of the fairway. For obvious reasons, simple or brute force plays a far bigger role in fighting, but ultimately it faces the same kinaesthetic issues: when all things (experience, aggression, strength, mental toughness etc.) are equal, more efficient technique wins the day.

What it all means

So what does all this mean? Am I suggesting that KU practitioners are not as effective as karate practitioners using an “internal arts” approach of the kind to which I refer? Far from it. I have trained with KU instructor Chris Mazzali and been most impressed by his obvious effectiveness as a martial artist. It does however highlight the very different approaches we take to martial arts training.

KU practitioners like Erik Angerhofer and Chris Mazzali clearly generate more force than I do. They might even apply more force than I do; however I’d like to think that I’m no slouch at applying force – and I think I do so by expending a great deal less effort (ie. I don’t have to generate as much force overall - see the video at the start of the article which I think illustrates my point). This is important to me as I age and as my body succumbs to a progressive immune-related arthritis. I simply can no longer do the sorts of “power” movements that people like Chris can. I have no option but to adopt a “softer” methodology.

Sequential relativism: a bridge between hard and soft

Is there a “happy medium” between the “internal” and “external” approaches? As I have discussed in my article “My quest for the martial ‘holy grail’”, I believe the answer lies in a pedagogy I call “sequential relativism” – ie. a syllabus that starts off with an external pedagogy and gradually shifts to an internal one as the student advances.

Again, the above gifs provide some idea of how one might apply this pedagogy of “sequential relativism”. For example, when I teach beginners a deflection/counter combination of the kind used in the Aragaki seisan sequence, I initially tell them to do a “1-2” movement. I focus on an effective deflection and a powerful counter. I don’t spend any time linking the 2 movements in some sort of “flow” – not until they have a good grasp of the fundamentals of both movements separately.

Once my students have a grasp of these fundamentals, I gradually introduce concepts of “flow” – the connectivity between movements. This is a long process and there are varying levels of “flow”. The flow in xingyiquan is quite applicable to an external art like karate. But, as I’ve discussed in my article “Can karate become taiji?”, the flow in taijiquan is something altogether different.

A drill I teach to senior karate students to get them linking deflections and counters into one cohesive, "flowing" technique. Note however that it is important with such drills not to lose "kime" or focus at the expense of "flow". Both need to be present in equal measure. In my experience the subtle kinaesthetics required to achieve this take many years to develop.


In the end I cannot say that my approach is “better” than others. I can only offer what I believe to be cogent reasons for why I prefer to do things the way that I do (ie. why it suits my particular needs at this point in my life and martial arts "career"). These reasons are based on my own understanding and research into martial techniques – in particular my cross-referencing of karate and the Chinese internal arts. As I commented to Patrick McCarthy Hanshi, I might be wrong – time will probably tell.

In the meantime I have not arrived at my position through (as some have speculated) ignorance of the importance of things like “koshi” (hip use). And it is to deflect such assertions that I have explained my position here in relation to "flow" (not to denigrate obviously effective martial artists who choose to follow a different path).


1. It is important to note that I am not opposed to the use of hips in kata. Hip use is a very important part of staged activation. However I feel that when hips are used, this use should be appropriate in the context. In the case of the move from Aragaki seisan illustrated in this article, I do use my hips. The hip loads naturally at the beginning of the move - ie. as you first step (see the adjacent images)!

These images are of me doing the move on the other side from that represented in the animated gif eariler in the article; I have chosen them as they illustrate the point more clearly.

The hip load occurs naturally as you step (you can see from the first image that my hips are "torqued"); this hip load doesn't need to be "forced" into the technique. I strongly feel that when executing this kata move you can and should use your hip to harness the full force of your forward momentum. And if you do so there will be no need to have a separate/additional hip load later on.

To quote a colleague of mine at
    "When stepping, there is an inherent motion to the hips. If this is utilized to load a technique, then there is no telegraphing or slowdown."
2. That Erik is using a kind of staged activation is quite clear; his hips load up and deliver a powerful blow. However this staged activation is not contextually appropriate. His hip use has been artificially "forced into" a move that already has a naturally occurring hip load - ie. the hip load illustrated in note 1 above.

So what Erik is doing is:

(a) not using the initial, contextually appropriate hip load; and
(b) not so much "pre-loading" as "re-loading" his hip - long after the contextually appropriate hip load moment has passed.

The net result is that his strike is divorced from the context of the deflection and step.

3. Accordingly it follows that as a "template" for a deflection/counter, the "koshi" way of doing kata offers no advantage over practising isolated basics. At best this way of doing the kata is a sequential rendering of 2 different, "stand-alone" templates, namely:

(a) one for a deflection; and
(b) another for a reverse thrust using the hip.

The step in between those techniques has no function other than to create a sequence "for the sake of a sequence".

With respect to the correspondent referred to in my article, I see no valid "template" purpose in doing kata this way. One might as well practise reverse thrusts on their own or against a makiwara; the reason for putting the techniques into a kata (ie. a dynamic context that utilises the step and the naturally occuring hip-load) has been lost.

4. The images below detail the extra time taken to do the "koshi" version of the Aragaki seisan movement discussed in this article (click on each image to enlarge it). I've ascertained that the frame rate in the 2 performances is the same, so the number of frames provides a very clear indicator of how long things are taking.

Here are the first 7 frames that take both Erik and me up to what I have called the "mid-point" - ie. the point at which the deflection would intercept the attack and when our feet are starting to cross in the course of stepping.

At this point Erik and I are "neck and neck". You will note however that my hips are torqued, ready to deliver the counter.

Here is the second set of 7 frames.

You will note that by the 3rd frame both Erik and I have almost completed the step forward. My strike has landed, but Erik's has not even started. In the last 3 frames my back leg shuffles up (which can be used with a foot stamp to add extra percussive moment to the counter - another "xingy-like" element) and my technique is complete. Erik however delays his slide up while he loads his hip for his strike.

Here are the last 7 frames.

You will note that only Erik is represented in these frames since I finished my technique in the preceding set.

Erik only commences his counter strike in the 4th frame of this set - that's a total of 11 frames after my own counter strike commenced. Similarly he lands his strike 11 frames after I have landed mine. These 11 frames comprise approximatley half a second. That might not seem like much, but in the course of a fight, half a second might as well be half a minute or even half an hour. Ask yourself: would you stand still for that half a second, waiting for your opponent to hit you?

5. I want to make it clear that I am not in any way commenting adversely on Erik's skill as a karateka: On the contrary, I chose Erik's example because he moves in a very powerful and skilled way. If I chose a poor example, my argument would be easily dismissed.

What I am attempting to highlight is that the insertion of additional, non-contextual hip movement will occasion a delay in delivering a strike - a delay that I think is unacceptable. In other words, it is the pedagogy (in particular the yamane ryu hip use) I am arguing against, not the individual. It just so happens that Erik is moving at almost an identical speed to me, so his example seemed particularly apposite in framing my argument.

I have since ascertained that Erik was suffering from a significant knee injury at the time his video was taken. I am also reliably informed that the video was intended to isolate individual movements rather than represent a full-speed, flowing version. In this light, the number of frames by which Erik's performance is delayed by the hip action might be considerably less had he been doing the kata "full speed" and "full force" (with the benefit of healthy knees).

Regardless of how fast the kata is done, I still think it is inescapable that the hip action delays the strike (relative to the "non-koshi" way of doing the movement). I also still need to be convinced that the hip action is appropriate and contextual in terms of maximising delivery of force.

Copyright © 2010 Dejan Djurdjevic


  1. As a practitioner of Krav Maga (who's striking component is very similar to kickboxing), I find the that flow is very important in a series of techniques in that some strikes flow more naturally from certain blocks, or after other strikes, depending on the relative position to my opponent, and being able to flow well is a huge advantage.

    I really enjoyed this post, but judging from what I just I think I might have missed something...

  2. Thanks Elias and Jorge for your support! I suspect I'll need it, given that this post will likely draw the ire of some practitioners who focus strongly on "koshi".

    But on the other hand, I've endured the last 20 or so years listening to how I "don't have depth" to my karate because I choose not to do this kind of action in my kata. There comes a time when you have to set out, in clear logical terms, your argument. And if you can't do so, then admit you are wrong. I think I've done the former so I'm not admitting anything yet!

    In the end, I don't know if it matters a great deal whether you have the "koshi" focus or not - it just means that you are assigning a different role to your kata practice. Given my "assigned use" of kata, the "koshi" principle has no role to play.

  3. I only recently started reading your blog. This post interests me because some of the things you discuss remind me strongly of RyuTe as I am being taught it (I am a very junior practioner). More than once since I began my study, I have said that karate as I am being taught it very nearly approaches being an internal art.

    Although I am reasonably sure that so well read an author already knows this, Taika Oyata says that the name "RyuTe" was chosen as a somewhat short-handed way of saying both "Ryukyu Hands" (don't ask me how--I neither read nor speak Japanese, let alone the Okinawan dialect) and "Flowing Hand," or "Flow of the Hand," reflecting his contention that karate techniques are intended to be performed in a flowing manner.

    I also can't help but note that when I asked my RyuTe teacher about generating power from the hips (I was originally trained in ITF Taekwon-do), he instantly made the same objection you did: that doing so telegraphed my intentions.

    One last thing: my own instructor is a rheumatoid arthritis patient, currently in remission. He credits the combined efforts of a particularly cranky but experienced rheumatologist and an acupuncturist for the remission. Perhaps that is a helpful tip, perhaps not, but I thought I'd offer it up.

  4. Thanks Man of the West.

    Yes, the character "流" means a flow or current.

    I'm very glad to hear that karate with flow in your school.

    My arthritis is mostly under control (however it is not in remission) thanks to my own marvelous rheumatologist to whom I am eternally grateful. I have tried acupuncture without success. Thanks for thinking of me.

  5. The main question about the flow is how to structure it so that your attack/couterattack delivers sufficient power while leaving no room for the opponent to defend or recover from it. The former needs more time for performing techniques and changing between them while the latter wants to cut the necessary time down as much as possible. I practice jujutsu (not BJJ)and our approach is to take the opponent to the ground, pin it down and then have all the time you need to deliver that final blow. Frankly speaking I don't get what karate's stance on this issue is.

  6. The late jujutsu master Jan de Jong once told me:

    "A lock is the same as a punch. It is no use saying 'how do I get out of this lock' when the lock has been properly applied. It is too late. In this respect it is no different from saying 'how do I deflect/dodge this blow now that it has landed.'"

    Extrapolating from this, I can tell that de Jong Shihan would have answered your query thus:

    Your take down and control on the ground IS your finishing technique. The punch you add to "finish off" is part and parcel of this technique. It just has more steps to the "finishing" aspect.

    So what is the karate stance on your jujutsu position? The same: avoid or deflect his incoming attack and apply a counter. Your counter comprises potentially more moves (a longer gameplan). But, properly executed, your gameplan is as irresistible from the first entry (eg. your opening grapple) as a counter punch to the face. In other words, like a game of chess, the outcome is predetermined (assuming correct application of technique).

  7. Mr. Dan,

    You said you trained with Hong Yi Xiang?

    I have a question:

    when talking about hips, do you turn them in the same direction as the force or in the reverse?

  8. Hi there.

    I didn't train directly with Hong - I trained with one of his students.

    As to hip use, I discuss this issue in this article.

    Thanks for reading!

  9. Hi Dan,

    Thanks for the redirect. I read the article, and was wondering if you would mind explaining just a little more detail about hip use for power in strikes.

    Specifically, whether you do snap them with the direction of force, or in the reverse.

    thank you!

  10. I forgot that the article didn't really address your query.

    There are 2 broad categories of lateral hip use: jun kaiten (moving the hips into the technique) and gyaku kaiten (away from the technique).

    I illustrate them in this video.

    Essentially there is no golden rule about when to use which one.

    Broadly speaking, gyaku kaiten is more powerful because it is a reverse movement and takes advantage of the natural counterswing your arms and hips make as you move. But having said that, there are many instances where jun kaiten is more appropriate given the rotation of your body relative to your attack etc.

    In the end, it all comes down to tangents and angles of interception. At some point I'll compile a dossier of different circumstances and see if some sort of calculus can be applied. But right now I see it as a "case by case" affair.

    As a matter of interest, I don't agree with this fellow in his use of gyaku kaiten. He is stepping into the attack and his hips should be turning into the attack with the counter following almost immediately. His method means that he is stepping in a way that is weak and setting him up for a footsweep; all because he feels more "power" in the block. In truth, it feels more powerful to him but his foundation is weaker - and he is chopping what is essentially one move (block and counter) into 2 disjointed "power" moves. In my experience this doesn't work in application against a resistant opponent.

    If you want to do what he has done, then your first step is off line (not into the attack) and you use a different stance (kiba dachi or Okinawan kokutsu dachi). It is only necessary to do this when the technique is very powerful and you can't wedge it easily - eg. a penetrating front kick.

    These are just some rough thoughts on the issue. In the future I'll examine this in more detail!

    All the best.


Post a Comment

Popular posts from this blog

Karate punches vs. boxing punches

Zhan zhuang: grounding, structure, intention and qi

"Combat tai chi"? Seriously?