Occam's Hurdled Katana: The logic of kata application

Occam's Hurdled Katana: The logic of kata application

In this article I want to look at the logical basis for the various approaches to kata application. I’m of the view that analysing things from a logical perspective makes a compelling case for the best way to approach kata. From there, the article then looks at how the discussion about kata and bunkai needs to change. And it does need to change!

As a basis for this logical analysis we will be putting an emphasis on Occam’s Razor. Before we get into what Occam’s razor is, and it’s relevance to kata, I’d first like to define the problem we will be applying it to.

As all readers will no doubt know, there are many approaches to kata application and many competing theories all claiming to be the “right one”. From my perspective the “right one” will be the right one for the individual and their training goals. Not everyone practises the martial arts with a view to developing functional combative skill; some people practise for artistic, aesthetic or cultural purposes and hence the “right one” for them will be a value judgement.

I’m totally OK with people choosing an approach based on personal preference. However, when claims are made about which is functionally the best approach, the discussion moves away from personal option because what works best is what works best. Functionality is not in the eye of the beholder.

So for me, and I’m sure most reading this – while I see value in the artistic, aesthetic and cultural aspects of kata – my main concern is to develop functional combative skills and to fully understand the kata from that perspective. The “right approach” for me is therefore the one that is most functional, and also the one that best explains the nature and structure of kata.

We can broadly split competing kata explanations into three main groups:

1 - Firstly, we have the “what you see is what you get” view that was the standard from the 1940s onwards but has gone into steep decline over the last decade or so. This is where the kata is viewed as being karate vs. karate punches, kicks and blocks. For ease of reference we will call this the “block, kick, punch approach”.

2 – Secondly, we have the “it’s all about motion” view. This is where the motions of the kata are not viewed as having a direct combative function, but instead are a means to develop general movement skills that can be applied in conflict. For ease of reference we will call this the “motion approach”.

3 – Finally, we have the “civilian self-protection” view. This is where kata is viewed as being a record of combative techniques and principles that are directly applicable. For ease of reference we will call this the “civilian self-protection approach”. To me, when I say “bunkai” it is this approach I am referring to.

So they are our three approaches. I accept that there are many variations within each approach and that they do not represent three uniform views. But as broad categories into which we can fit most of the views out there I’m sure you’ll agree that they will suffice.

So which of these three approaches is the right one for the martial artist who seeks functional skills and to understand the nature of kata?

When deciding which approach is the right one, there are three main measures one can make use of: The historical measure, the practical measure and the logical measure.

Now it should come as no surprise that I hold the view that it is the “civilian self-protection” approach that is the right one, that is the most practical, and that provides the best explanation of kata. I’ve been publicly arguing the case for this approach for many years. I feel that it is “civilian self-protection” that has the strongest support historically, practically and logically. See everything I’ve ever written, filmed, recorded, taught and said for the last fourteen years or so for more details! In this article though, I want to look at the logical case via Occam’s razor and “the burden of proof”.

OK, so what is Occam’s razor? Occam's Razor is attributed to the 14th-century English Franciscan friar called William of Occam who wrote extensively about logic. Essentially, Occam's razor is a principle which recommends selecting the hypothesis that makes the fewest assumptions … but only when the hypotheses are equal in all other respects.

Occam’s razor is often incorrectly summarized as "the simplest explanation is most likely the correct one". However, that’s a little misleading as the principle is actually about establishing burden of proof in discussions.

By “burden of proof” we mean who has the greatest obligation to provide evidence for their position. A widely used example is “Russell's teapot”. If I were to say that there is a teapot orbiting the earth, it is up to me to provide evidence that that is the case. I can’t argue from a position of “if you doubt me, then prove that there isn’t a teapot orbiting the earth!”

Occam’s razor is "entities must not be multiplied beyond necessity” or "plurality should not be posited without necessity". The word “razor” is employed because Occam’s razor is about “cutting away” unnecessary assumptions. Occam’s razor is widely used because it has a very good track record.

Returning to kata, for our proposes Occam’s razor means that the approach to kata that makes the least assumptions is to be the favoured one. It also means that all approaches need to explain why any assumptions they make are necessary. The burden of proof would therefore fall on the approaches making the most unnecessary assumptions.

To keep things as simple as possible was can paraphrase all this as:

“The approach to kata that makes the least assumptions should be taken to be the best approach. Hence the burden of proof would be on the other approaches to explain why their assumptions are needed to explain kata.”

When I was thinking about how to structure this article, I thought it would be good to compare some conflicting examples of kata explanations using Occam’s razor. As soon as I thought of “razor”, the motion that immediately jumped to mind was the jump in Pinan Godan (Heian Godan in Shotokan). This motion is often explained as leap over an attack to the legs with a bo or a katana. Hence the title of this article as “Occam's hurdled katana”

Earlier in the article, I said, that “essentially, Occam's razor is a principle which recommends selecting the hypothesis that makes the fewest assumptions … but only when the hypotheses are equal in other respects.” It’s that “equal in other respects” that I’d like to briefly explore now.

It’s my view that the three completing categories of kata explanation (“block, kick punch”, “motion”, and “civilian self-protection”) are far from equal! I’m going to totally leave aside the historical and practical strengths and weaknesses of each approach because I want to concentrate purely on logic and burden of proof in this article.

Whist I believe that the “civilian self-protection approach” is strongest both historically and practically, for the sake of the article I’m going to assume they are all equal in those regards so we discuss things purely from a logical position. The approaches are still not equal though because it is only “civilian self-protection” that explains all the data. To help illustrate this, let’s now look at the example for Pinan Godan.

The motions we are going to look at is where the head turns, the hand shoots up behind the head as the other arm is pulled across the chest (knuckles up). There is then a jump where the body turns 180 degrees and lands in reverse cat stance (some styles have one knee down) and the arms are crossed: in what is often referred to as a “lower cross block”.

The “block, kick, punch approach” most commonly explains this sequence as a punch, a leap over a low weapon attack (typically a staff or a katana) followed by a “cross block” to stop a kick. Ignoring, as I said we would, all the practical problems with this explanation, this approach fails to explain all the data! If the motion is a punch why is the head turned away? Why is the arm placed across the chest with the knuckles up? Why is the “lower cross block” done in reverse cat stance? Why not any other stance? The turned head, the hand position, and the stances are not explained.

The “motion approach” has similar problems but on an even bigger scale.  In this approach this whole motion is not explained. Why is any of it the way that it is? Why use that set of motions as opposed to any other? If any motion will do, then why bother to get kata right? And if any motion won’t do, then those who hold to this approach need to explain why the kata is the way it is?

Those who hold to this approach need to explain why is every part of the kata is the way it is. They also need to explain the mechanism by which the motions of kata – which they have divorced from direct combative function because they say the movements have “no bunkai”– can later on be reconnected to combative function. If a convincing explanation has been put forward, or even seriously attempted, I have yet to come across it. Why the kata is the way it is not explained.

The “civilian self-protection” approach (specifically my take on it) would explain the motion as follows: The arm rising shoots under the enemy’s right arm. The left arm (the one across the chest) takes hold of the back of the arm and then twists and pulls that arm in preparation for a throw. The head turn initiates the rotation of the following shoulder throw. The 180 degree turn will see the enemy thrown to the floor. The crossed arm position is the same arm position used for an entangled shoulder-lock. The natural response for this lock is for the enemy to twist out of it. The “jump” prevents that by taking the right leg over the enemy and blocking their path. The reverse cat stance puts the thigh against the enemy’s back and hence ensures they are trapped and the lock can have full effect.

Always hard to describe these things in text, but my DVD “The Pinan / Heian system: the complete fighting system volume 2” explains this in detail. I also hope the pictures down the side are useful in making things a little clearer.

Every single part of the motion is explained with the “civilian self-protection approach” … and that is not true of the others. You can apply this style of analysis to any kata motion and consistently the “civilian self-protection” approach can explain why every single part of the motion is as it is (which is after all a fundamental requirement of that approach), but the “block, kick, punch approach” and the “motion approach” cannot explain why the kata is the way it is and not some other way.

So Occam’s razor cannot really be applied because the three competing hypothesises are not equal as only one of them can explain the data and why the kata is the way it is and not some other way. However, for the purposes of this article we will ignore all this and pretend, for now, that these huge gaps do not exist. Although only “civilian self-protection” meets the entry requirements (being the only one to explain the data), I’d now like to move on and apply Occam’s razor to the three competing approaches using the same example.

You’ll recall that earlier I said that to keep things as simple as possible we would apply Occam’s razor to kata by saying:

“The approach to kata that makes the least assumptions should be taken to be the best approach. Hence the burden of proof would be on the other approaches to explain why their assumptions are needed to explain kata.”

So let’s look at the assumptions each approach makes:

The “block, kick, punch” approach makes the assumption that you will be attacked by three people in a specific manner and sequence. For it to “work” you need to hit one opponent in the face (without looking), then, immediately after, a katana wielding opponent needs to attack you from your left with a low cut. Any other attack out of an infinite number of possibilities would render the sequence invalid.

There is then the assumption that the katana wielding opponent will stop their attack as no further attempt is made to neutralise them. There then needs to be a low kick which is delivered to your assumed place of landing (not where you were) at exactly the right time to make the sequence work. Again, anything other than that specific motion, out of the infinite number possible, will render the whole sequence invalid.

Without very specific assumptions about when and how the enemies will act, the whole sequence cannot be said to explain the sequence; and that is leaving to one side all the things that are seemingly done for no reason (i.e. the head turn, the hand across the chest, etc) and the obvious practical failings. The odds of this actually happening in the right time, place and sequence are astronomical!

I should probably say that there will be some who will say I’m presenting a “straw man argument” here; which is when you misrepresent a contrary position to make it seem like a very weak and easily defeated position. While this is a very weak position, it is nevertheless the position that is legitimately put forward by innumerable karate books and instructors who subscribe to the “block, kick, punch approach”! As mentioned earlier, there may be variations (i.e. a bo and not a katana) but the approach is as literarily as weak as presented.

Let’s now look at the “motion approach”. The big problem here is that the process by which the supposedly non-combative motions of kata become applicable in combat has never been identified. The failed attempts to do so normally employ the argument that it’s about “principles” not techniques.

Now as anyone who is familiar with my material will know, I am also a great believer in principles over techniques. The crucial difference is that I say the methods of the kata are directly combative. It is therefore simple to see how the combative techniques of the kata are based on combative principles; and how the study of those combative techniques can lead to an understanding of combative principles. Simple!

What is much harder to grasp is how non-combative techniques can ever be based on combative principles! And how studying non-combative motions can lead to combative skill! It’s simply not logical.

The big assumption is therefore that the supposedly non-combative motions of kata can lead to combative skill through an unidentified and unexplained process. This process has never been explained or identified and is often argued from a position of “faith” as opposed to logic i.e. “if you have faith and stick with the process all will become clear”. This really does not hold water though and we’ll discuss why in a moment.

In the interests of thoroughly exploring the topic, I should say here that the “motion approach” is largely based on modern revisionism in response to the “civilian self-protection approach” to bunkai.

Whist the view that “kata is not about bunkai, or direct application, but motion” is now very prevalent in certain groups; those same groups were not saying that before practical bunkai became more widely practised. It was the “block, kick, punch approach” that was what was subscribed to in those groups. A simple look at what the most senior instructors of these groups and styles were presenting in their books and videos fifteen to twenty years ago will confirm that.

Those same groups that now say “in our approach kata doesn’t have bunkai as they are all about motion” need to remember that the most senior instructors that lead their groups where not saying that until comparatively recently. Look at their material and you see kata being explained as karateka vs. karateka block, kicks and punches and no mention that kata was anything other than what they were presenting.

As the “civilian self-protection approach” to kata has grown in popularity – and as it has become increasingly obvious to all that the “block, kick, punch” is seriously deficient – people started to ask, “Why don’t we do that?”, “Why don’t we do bunkai?”, and “Why are our methods not workable in real situations?”

Those that don’t subscribe to the “civilian self-protection” approach have three ways to address these questions:

1 - To adopt the “civilian self-protection” approach.

2 – To ignore it and carry on with “block, kick, punch” regardless.

3 - To say that neither “civilian self-protection” or “block, kick, punch” are valid and in the absence of an alternative valid approach plump for “motion”. 

The “motion approach” has the seeming advantage of not having to stick with the obviously flawed “block, kick, punch” but also of not having to accept the “civilian self-protection approach” to bunkai and the learning curve associated with it. It also protects the ego from having to admit there may be a side to karate that they have yet to explore and know little about.

It’s also worth saying that I think the “motion approach” is a temporary holding position that will eventually die away. People will either switch to the “civilian self-protection approach” – and we see a lot of that happening – or, after accepting that the “motion approach” does not really have any validity, they will abandon kata altogether. In fact, you can see the beginnings of that already as the groups most strongly tied to the “motion approach” are now frequently remarking that kata is not really that important: which is a huge shame when embracing it in the right context can bring so much.

As we have discussed the main failing with the “motion approach” is that the process for how it develops combative skill has never been explained. It’s a matter of faith where the exemplary practitioners of the style as pointed to as “proof”. The trouble with that is, no matter how superbly skilled the practitioner, there is no proof that their approach to kata had anything at all to do with developing that skill because the process remains unidentified.

We have seen that the “block, kick, punch approach” needs to make impossible assumptions about the unpredictable actions of their enemies. The “motion approach” is even worse because it makes the massive assumption kata will develop combative skill through an unidentified and unexplained process. Let’s now contrast them with the “civilian self-protection approach”.

When we look at how the “civilian self-protection approach” explains the motion in Pinan Godan, no assumption about the enemy is made. They are thrown and locked. They are not required to do anything to make the technique work. It makes infinitely fewer assumptions than the “block, kick, punch approach” and hence according to Occam’s razor the “civilian self-protection approach” should be the preferred hypothesis.

The “motion approach” is assumed to develop combative skills through an unidentified process. It is very easy to see how the “civilian self-protection approach” develops combative skill … it is through the practise of combative methods! We can also look at the principles employed in those methods (balance breaking, leverage, exploiting weaknesses of the human body, etc) and expand our study beyond the specific examples in the kata to the underlying principles. There is no “assumed process” but a direct and demonstrable one. On this basis, Occam’s razor would again have the “civilian self-protection approach” as the preferred hypothesis.

Although I believe otherwise, for the sake of argument let’s assume that the unidentified process supposedly used in the “motion approach” works. If you practise kata for a long time it will somehow lead to the development of combative skill. Occam also wrote about the “principle of economy” which basically states:

“It is pointless to do with more what can be done with fewer”.

The “civilian self-protection approach” starts to develop combative skill immediately through the practice of directly applicable methods. The “motion approach” is said to develop combative skill after time. But even if it does (and it doesn’t!) the “civilian self-protection approach” should still be the preferred hypothesis via Occam’s principle of economy, because it develops combative skill with less. You are learning to deal with violence immediately and directly through a readily identifiable process.

Whichever way you look at it, and however you stretch the alternative approaches, the “civilian self-protection approach” is logically the hypothesis to favour.

The bottom line is that the “civilian self-protection approach” works and the other approaches do not. However, what I hope this discussion has shown is that even when we put practicality aside, the other approaches cannot explain the nature of kata, and need to make many improbable or unidentifiable assumptions. Under Occam’s razor, the “civilian self-protection approach” is the favoured one … and therefore the burden of proof lies with those who hold the other views!

The “civilian self-protection approach” can explain every part of the kata, it makes no improbable assumptions, it does not rely on improbable or unidentified processes, and it develops combative skill from day one. If people believe the other approaches are more valid then the burden of proof lies with them.

Those who truly believe in “block, kick punch” need to explain all the elements they currently ignore. If the motion is Pinan Godan really is a punch, then explain why they turn away from the target when they deliver the punch? I’m genuinely interested why anyone would think that is better than looking at the target? Please put forward your case as I’m all ears?

And that’s not enough. They also need to explain why every part of the kata is the way it is and not some other way? They also need to explain why their seeming improbable scenarios are in fact better explanations than the direct and obvious explanations put forward by the “civilian self-protection approach”.

Those who truly believe in the “motion approach” need to explain the process whereby supposedly non-combative motions can develop combative skill? That seems like an impossible ask to me, but having boldly made the claim, I want to see them back it up with solid explanations.

They need to show how the “motion approach” can develop combative skill quicker and more effectively than the “civilian self-protection approach” if it is to be preferred over it?

They also need to explain why every part of the kata is the way it is and not some other way? Why that motion and not some other motion? This needs to be done for every single movement in all the kata practised.

That may seem like a tall order, but the “civilian self-protection approach” can explain why every move in kata is the way it is, so if the “motion approach” is to be considered as a valid alternative it needs to be able to do the same.

We need to see the process explained and all the details of any given kata explained in accordance with it. Myself and those like me have done that in our books, DVDs, articles, etc. Where is the material that puts forward all the details of the motion approach and logically explains its effectiveness? I genuinely want to read that information.

This brings me to my reason for producing this article. We who subscribe to the “civilian self-protection approach” have thoroughly explained our views and put forward the case for our approach. This is something we should continue to work hard to do because I believe it is through our collective efforts that karate has a strong future.

But while we work hard to further explore the kata and its bunkai for the benefit of those who share our approach, we need to remember that when people challenge that approach that the burden of proof does not lie with us! By any measure you choose to use, the burden of proof lies with those who hold the alternative views.

The “civilian self-protection approach” is simple, it works and it has no need for unexplained processes or massive assumptions. The “block, kick, punch” and “motion” approaches simply can’t say that.

Occam’s razor therefore says that the “civilian self-protection approach” should be the preferred one. The “civilian self-protection approach” is the logically preferred hypothesis and hence it should be the hypothesis adopted by all people who approach kata logically.

If people believe otherwise, it is up to them to come forward with convincing answers to the questions asked in this article. If they can’t do that then they need to accept that their reasons for holding to the approaches they do are illogical i.e. too invested to change, dislike the idea of something they don’t know having validity (ego), kowtowing to authority (Master X says so and that’s good enough for me) and so on.

The burden of proof points firmly one way! Although it is very easy to do so, it not up to us to prove that those who subscribe to “block, kick, punch” or “motion” are wrong: it’s up to them to prove themselves right!

I genuinely look forward to seeing any attempts at doing so as it will move the discussion along. However, I won’t be holding my breath on the basis that none have been forthcoming so far.

As I said at the beginning of the article, I’m totally OK with people choosing an approach based on personal preference. And if people prefer “block, kick, punch” or “motion” as a personal preference then more power to them. However, when claims are made about which is functionally the best approach, the discussion moves away from personal option and they need to thoroughly explain their position in order to enter the discussion. Simply saying, “because it is” or because “Master X said” are not logically tenable positions.

To have a meaningful discussion, all competing approaches need to be thoroughly explained. Those who subscribe to the “civilian self-protection approach” have done that. Subscribers to the other approaches, to date, have not explained their approach enough for it to be meaningfully discussed.

For those who subscribe to the “civilian self-protection approach”, our efforts should be 100% focused on advancing ourselves and that approach. It’s not our job to convince those who disagree about the merits of what we do. That’s a waste of our time as we are so far ahead on all fronts. They need to catch up to be part of the discussion!

When they have put forward their approach in as much detail as we have, and when their theory can explain the data, and be shown to work as well as ours does, then, at that point, the discussion will be worthwhile. Until then, we should leave them behind and get on with shaping karate’s future.

We bunkai / “civilian self-protection approach” types need to remember that we are the ones in the strongest, most logical position. We need to understand that the burden of proof is not ours. We need to remember that the discussion has moved on and it’s up to those who chose to stick to the “block, kick, punch” and “motion” approaches to explain their position in as much detail as we have and show that it can work every bit as well.

We are way ahead, and it’s not up to us to slow down or look back. If subscribers to the “block, kick, punch” or “motion” approaches want to be part of karate’s future, they need to catch us up if they are to avoid being cut away by Occam’s ever sharp razor.