Searching for the "ancestral" naihanchi: Part 1


Hot on the heels of my articles about "stem cell" movements, I came across this version of naihanchi shodan kata recently, courtesy of a Facebook group to which I belong.

Naihanchi (in particular, naihanchi shodan) is one of the most popular and widely known and practised karate kata on the planet, so I think it is one that is truly worth examining more closely. And in my opinion, the version below provides one of the best benchmarks for analysing the various differences and similarities.

Naihanchi shodan by Zenpo Shimabukuro

This is a performance by Zenpo Shimabukuro, head of the Shorin-ryu Seibukan karate organisation.

He is the son of the founder, 10th Dan Zenryo Shimabukuro, the longest-seving student of Chotoku Kyan who, along with people like Choyu and Choki Motobu, Kentsu Yabu, Chomo Hanashiro, Gichin Funakoshi, Shinpan Shiroma, Choshin Chibana and Kenwa Mabuni, studied directly under the legendary grandmaster Anko Itosu, who was himself a direct student of the even more legendary Sokon Matsumura, regarded as the father of Shuri te (the "suidi" school of karate).

However, as my friend K-Hirakis notes, Kyan did not in fact teach naihanchi to Zenryo Shimabukuro. Instead, it seems that the younger Zenpo Shimabukuro learned those kata (as well as the pinan series, among others) from Chozo Nakama whose primary teacher was Chosin Chibana (although he also trained with many of Itosu's other top students). [See this interview with Zenpo Shimabukuro in the Dragon Times.]

I have posted this video not just because it comprises what I think is clearly an exceptionally powerful and economical performance of naihanchi shodan, but also because I believe it exhibits the greatest combination of the disparate features, at least in embryonic/stem-cell form, many of which are then brought into sharper focus by the various ryu (often to the exclusion of others).

In this respect I see it as the most "composite" version of naihanchi shodan in existence today (and in that sense, possibly most reflective of all of the "ancestral options", at least dating back to Itosu's time – we'll see about the latter as the article progresses!).

I thought I'd go through some of the features that I feel are "ancestral" but which are left out of many modern forms. (I hasten to add that I do not think their absence from some of today's forms automatically disqualifies those forms from being effective).

I'll also be pointing out a couple of "ancestral movements" that might be missing from, or altered in, Zenpo Shimabukuro's version.

The stance

Okay, I know I'm a bit of a pain to some folks on this one. Readers will recall that I've written about this extensively in the past (see my article "The naihanchi stance"). Others have debated this topic with me extensively by email. Yet more recently I've had even more discussions about it on Facebook and other social media sites. But really, as far as history goes, I hold it to be self-evident that there simply is no debate. Practically every single photograph and (rare) motion picture performance by Okinawan masters who were direct students of Itosu reveals what is unmistakably just "mabu" (horse step) or "kiba dachi" in both Okinawa and China alike.

The only possible exception I can think of is, oddly enough, Choshin Chibana's naihanchi as shown below:

Choshin Chibana's naihanchi kata. Note the ultra narrow stance.

[I say "oddly enough" because Zenpo Shimabukuro's naihanchi came from Chozo Nakama who presumably learned it from Chibana. Yet Shimabukuro's and Chibana's kata are very dissimilar, indicating that either Nakama or Shimabukuro, or both, studied naihanchi with other instructors. Indeed, this almost certainly accounts for the mix of different features.]

In his kata, Chibana seems to be doing nothing more than a heiko dachi (shoulder width parallel stance). But then again, this footage was taken when he was already quite old, and the narrow stance might have been a reflection of this.

If Chibana's is just an "ultra narrow horse stance" (especially accommodating his age), it is worth noting that the Chinese mabu varies greatly - from a very wide and deep stance (as per kiba dachi of shotokan) to the fairly narrow and higher stances of the kind I've been studying in the Chinese internal arts and hybrid internal/external arts (eg. of the late Hong Yi Xiang of Taiwan).

But when it comes to differences, at least Choshin Chibana's stance seems to differ only in terms of width. By contrast, in today's world there is a tendency (particularly in some schools of isshin ryu, but also in many other systems, eg. wado ryu) to adopt a "new beast" - a "naihanchi dachi" comprising a narrow stance that also has the feet overly turned-in.

In historical terms, I am of the view that this bears no resemblance to any old footage or photographs of naihanchi.

To me, this is clearly a modern innovation: an intention to create a "sanchin-like or wing chun-like" stance. If anyone needs confirmation of this, they need only look directly at the kata performances of the founder of isshin ryu, Tatsuo Shimabuku. These show that his stance was not a "turned-in" stance reminiscent of sanchin or wing chun. Instead it is manifestly just a horse stance – albeit a slightly narrower one.

Tatsuo Shimabuku's naihanchi kata. Ignore the fact that he elected to start his kata to the left, as per Motobu (the kata is perfectly symmetrical, so it really doesn't matter in which direction you start) and focus instead on his stance.

The same goes for wado ryu: have a look at Hinori Otsuka's naihanchi below:

Hinori Otsuka's naihanchi - again, note the stance and compare it to the "naihanchi dachi" proposed above.

Now it's true that in naihanchi your feet might (some would say "should") be oriented so that the outside edge points forwards. This results in a slight inward emphasis (visible in the photographs of both Funakoshi and Motobu and even Otsuka's video, for example). But it is most definitely not the turned-in posture of the preceding video purporting to show the "Isshin ryu naihanchi stance". That is simply not consistent with either isshin ryu history nor the history of naihanchi.

By the way, I'm not saying that there is no point in practising a narrow, turned-in stance. There are good reasons for turning feet in a little more than the "outside edge forwards", namely grounding/rooting as per wing chun, the other Hakka systems, sanchin/sanzhan and zhang zhuan postures. I'll get into this another time.

For now I will just note that I don't think naihanchi is the place for this sort of overtly "pigeon-toed" stance. And I say this for one simply reason: such a stance does not favour sideways movement. In fact, it is the antithesis of sideways movement.

If you doubt me, adopt a pigeon-toed stance, then try to move quickly from side to side using kosa dachi (cross step). Then try it with feet forward. Then try it with feet pointing out. I'm sure you'll agree that the "feet forward" option is by far the most natural.

Slight variation (outside or inside edge forwards) doesn't really make much difference. Adopting an overtly pigeon-toed stance (or, for that matter, a ballet-like plie!) does. Why? Apart from the fact that you might trip over your own feet, you'll note that turning your feet has significant consequences for your hip alignment - impeding sideways mobility in both cases.

In my view, it is for the latter reason that arts that feature "pigeon-toed" stances (eg. wing chun, various Fujian-originated systems such as ngo cho kun, goju ryu karate etc.) do not have forms with side to side embusen (at least not while holding a pigeon-toed stance).

Instead, inward-turned stances are used for moving forwards and backwards (see sanchin kata in Naha te, for example).

Opening inverted ridge hand vs. back hand

The opening strike/deflection/throw is one that is generally performed in two different ways:

* as a haishu uchi/uke – ie. a backhand strike, projection or deflection; or
* as a haito uchi/uke – ie. an inverted (palm up) ridge hand strike, projection or deflection.

Which one is the "ancestral" one? Here I must confess that I don't really care that much.

I actually do the kata with a backhand. But it is important to note that I also perform applications with the move as an inverted ridge hand.

I think the "palm up" is the fuller movement. Now the "fuller movement" doesn't always represent the movement chosen by the kata designer: it might be that a smaller part of the movement is more consistent with the general utility of the kata sequence.

And an argument along these lines can certainly be made to prefer the backhand versus the "palm up".

For this reason, if I were to express a view as to which was the "original" I would have to go with what Funakoshi and Motobu did - ie. the backhand - and against Zenpo Shimabukuro's "palm up" ridge hand. Let me explain my thinking:

First, when two instructors with such different perspectives, such as Funakoshi and Motobu, choose the same hand shape, you have to take note.

But second, let's examine applications for the two options (ie. palm up vs backhand) and see which is really the primary one that we might use in karate, given the dominant tactics of that art:

For the purposes of a throw, you really want the full turn of the forearm all the way to palm up – as I illustrate in the following video:

Some applications of naihanchi's opening moves

But for a deflection or strike (as I demonstrate immediately after the above throw application), I think you will probably never go beyond the "backhand".


First the deflection: In this case, the backhand allows you to use the rotation of the forearm to "slip" even the most momentum-heavy punches and other upper body attacks. Yes, a further rotation to "palm up" would theoretically help even more, however in practice I've found that the deflection is invariably completed early in the piece, making further rotation redundant.

This deflection is a very powerful, yet "soft" technique – a bit like a crane's wing (indeed, I think it might be originally related to one of the crane schools – albeit from northern China (near Beijing) not southern (eg. Fuzhou, Yong Chun, etc).

And this "soft" wing movement also sets you up beautifully for the empi uchi (elbow strike) – particularly if you haven't rotated your palm all the way around to face upwards (because your opponent's head then falls naturally into your hand following the deflection).

Second, as a strike I feel that once you rotate fully over to the ridge hand you've gone a bit too far and you start to lose impact force. I think this has a lot to do with extending the arm out fully "palm up"; this is not a strong or "natural" position (see my discussion in "Why corkscrew your punch" as an example). If you think of most "power" backhand moves (eg. a baseline tennis shot) you'll notice that at the point of impact the hand is holding the racquet so that back of the palm lines up with the strike. The wrist only twists over "palm up" as a follow-through.

Without an extension such as a racquet or stick, humans are even unlikely to execute the "palm up" as a follow-through. If you think of the standard uraken (backfist) it is not executed with a "twist". Nor have I ever noticed a backhand slap being executed with such a "turn over".

Last, one telling feature of the "palm up" naihanchi executions is that the arm is invariably overly bent at the elbow when the strike is completed. This tells me that the movement feels weaker and less stable to the practitioner (who is subconsciously drawing the arm in so as to bring it closer to the core).

Essentially, I hold it to be self-evident that karate is more about "striking" than it ever was about "throwing" and grappling. So for a simple power technique, used either as a strike or as a defence against a strike, you can't go past the haishu (backhand) in performing naihanchi shodan. The haito (inverted ridge hand) is, to me, a secondary option, primarily for throwing and off-balancing generally.

Correct use of forearm rotation and pullback arm

But whichever way it goes (whether you use a palm or backhand finish), Zenpo Shimabukuro's kata is more notable for two other details accompanying this hand movement, namely:

* an optimal rotation of the forearm through to that finish position; and
* the correct the use of the pullback arm to assist the movement.

These details are nothing short of crucial, yet they are often omitted or simply glossed over.

Let's examine details more closely:

First, in order to correctly rotate the forearm to add force to the movement, the deflecting/striking hand should be loaded, palm down and move to a palm vertical or palm up position.

Second, it should be loaded underneath your armpit of your pullback hand, not above it.

In this respect it is the polar opposite of the downward palm gyaku shuto uchi (reverse chop) in Miyagi's gekisai dai ichi and ni (and many shorin kata that feature the gyaku shuto), because this technique starts palm towards the ear.

It then operates on a falling moment, finishing facing palm down).

This naihanchi technique is the exact opposite (ie. it has a rising moment, finishing facing palm up). In that circumstance, it shouldn't be at all surprising that the technique also requires an upward moment with more or less the inverse action to the gyaku shuto (chop).

Now this "upward roll" of the forearm is critical for a number of reasons: It not only allows the forearm rotation to deflect a punch but it also exerts force on the attacker in the form of a off-balancing manoeuvre.

Last, the pullback arm acts to protect you, then assist with any follow-up technique.

Occasionally I see naihanchi palm technique being performed with a "downward" action starting above the pullback arm. In biomechanical terms, I think this is plainly at odds with what how the kata was designed.

Here is one example I've found – and I hate to mention it because the practitioner is not one I wish to be seen to "criticise": his karate (and naihanchi) is nothing short of excellent. And just because he's not doing it the way I think it should be done, doesn't mean it doesn't "work": of course it does. It's just that I don't believe it is being done in a way that optimally uses the movement in the context of this kata – in particular the rotation of the forearm in the particular kata move.

For example, he starts palm down, yet in his application his technique finishes "palm down" in a kind of chop to the opponent's neck. This would imply that almost no circular rotation of the forearm is being employed at all - or that a kind of rotation is being employed that is more like gekisai than naihanchi.

An example of the "above shoulder" version of naihanchi's opening move

If one wants to chop with the palm downwards (as per the gekisai kata), I believe this is best done starting high, palm twisted up near the ear, and rotating downwards (again, as done in gekisai). Now this is a very effective movement, for sure. But in my respectful opinion, this is not what is happening in naihanchi. In that kata the hand starts low, palm down, and has to rise. As it rises, it will naturally rotate. And in that context it cannot rotate to a gyaku shuto (chop) (unless it goes one way, then the other).

It is important to note that just because I feel that the "backhand finish" is more "ancestral" than the "inverted palm finish" doesn't mean that I can realistically believe that the backhand was ever intended to turn over first, then downward into a gyaku shuto (chop). Or, perhaps, that the strike would feature no forearm rotation whatsoever, simply swinging horizontally.

In this regard it is salient that I have not found even one version of naihanchi shodan where anything like this occurs.

They all start palm down, then turn over - either to a vertical (palm sideways) backhand strike or a horizontal (palm up) ridge hand strike. Either way, the rotation happens in the same direction as a natural "completion" of the movement (and the hand starts relatively low!). To me, this speaks volumes.

Otherwise, I have also found a number of variations where the practitioner uses a kind of "pancake flip" motion on the primary arm (and doesn't use the pullback arm at all). Most notably this occurs in Tatsuo Shimabuku's isshin ryu system (see above).

It is also done, rather strangely in an almost identical way, by Choshin Chibana (see my earlier comments about the lack of similarity between Chibana's naihanchi kata and Zenpo Shimabukuro's, despite the suggestion of at least some lineal connection).

Regardless, the school that most features the "single arm" movement today appears to be the isshin ryu school and its offshoots. Consider the adjacent images with the single-arm motion, taken from the video below:

Another isshin ryu performance of naihanchi

[Continued in Part 2.]

Copyright © 2013 Dejan Djurdjevic


Great article. If you can find the book, Javiar Martinez did a book on Isshin-Ryu's Naihanchi kata. He went back to the old chinese sources and found the same technique patterns in their katas and pulled out the "chin na" grappling aspects of the kata. I think it's probably the closest to what the original "may" have looked like.

Also, of note is that Chotoku Kyan did NOT teach Naihanchi kata. The Shimabukuro family brought it in from Chozo Nakama. Which also explains why Isshin Ryu's Naihanchi is different than the other Shorin style schools. It is unknown who taught him the hourglass stance more akin to southern chinese systems. Motobu started his Naihanchi to the left, and the IR version is said to go left as homage to Choki Motobu.
Dan Djurdjevic said…
Thanks mate. I've made some changes already based on your comments!

I'll have to look up the Martinez book!
David (RenegadeMonk) said…
Dan, you have failed to consider the unequivocal written account of Choki Motobu in his Watashi no Karate Jutsu. He is quite clear about that Ankoh Itosu performed Naihanchi was his toes pointed inwards. He states that neither he nor Matsumura agreed with this practice. Chosin Chibana was a disciple of Itosu, the stance you see is a clear fit with Motobu's account. Difference's in method between different generations of a lineage are accounted for by the shu-ha-ri ideology of Okinawan karateka: each generation makes its own judgements about what works best and kata are adjusted accordingly. It is IMO a function of pirate's exportation that the idea of one fixed correct method has taken hold.

I suggest that Itosu's Chinese teacher was the inspiration for either introducing or returning to this method. That it is not the best method for sideways movement would only be relevant if the techniques were intended to be applied sideways. Most people who study Naihanchi do not come to that conclusion.

Dan Djurdjevic said…
Hi David!

As always, some very good points.

Ryan Parker and I were discussing the other day this question of "inward toes" and we both concluded that when you have the outside edge of your feet pointing forwards, this already means your toes are "inwards facing". By contrast, people like Motobu clearly had the inside edge of their feet pointing forwards (ie. there was a slight "toes outward" bias).

In my view there is no indication from historical records or photographs to show that naihanchi was ever executed with a greater "inward turn" than the "outside edge forward" variety. For example, I don't think Chibana's filmed performances reveals an inward turn of the magnitude greater than this.

So I'm afraid I don't see a "clear fit" between Chibana and what Motobu says of Itosu. I see a rather narrow, straight kiba dachi in Chibana's videos - not the wing chun "A" stance.

What I think has happened is that certain people have read Motobu's comments and either:

(a) misinterpreted them (leading to the rather odd, marked inward turn of the kind shown in the isshin ryu video); or

(b) used them to justify their own practices that have evolved independently of Motobu's comments.

As to the "sideways application", my point is actually that sideways stepping conflicts with the (excessively) inward-turned stance - not that the applications people might use "conflict" with such a stance.

Remember that "bu" in Chinese actually means "step" not "stance". This should give you an idea that "stances" were never intended to be "static"; they were snapshots of movement. To me, there are no "stances" but rather "steps".

Naihanchi steps exclusively to the side. Here, the shape of your feet etc. is not appropriate to such steps.

So the fact that applications can be applied once you've finished stepping is neither here nor there for me; things don't happen so statically in real fighting.

Apart from this, I don't see too many bunkai where people actually adopt an overly "inward" stance. Our school is one of the few that uses inward stances like sanchin in a dynamic context (ie. keeping feet "inwards" during stepping even in bunkai). But it is important to note that the stepping is never straight sideways like in naihanchi; the closest you get is a step and pivot as per this gif. And even there, you'll notice that I'm only turning the outside edge to face the uke - my feet aren't turned in any more drastically than this.

Thanks again for your comments!
David said…
Dan, Motobu's account is a historical record. His account states that Itosu's stance turned the feet in beyond where he (Motobu) thought was reasonable so more than that straight outer edge you are referring to.

Itosu has been called a student of Matsumura, but there is more compelling evidence that he was a student of Gusukuma. Another lineage, potentially another way of doing things. That it never really caught on (the only thing you can deduce from photo evidence) is neither here nor there. It may be that the straight foot is the far superior method, my point is simply that it was done and done by someone of prominence. The point about the applications is simply that if (as I have found) many of the applications for this kata can benefit from the A stance, then there is good reason to perform the kata using it.

The kata is just one aspect of training and as such does not in itself need to be super efficient in every respect. You sacrifice some areas less important to the final product (the fight itself) in order to accentuate other points that are more important.

Lastly it should be noted that I do not practice this kata with the A stance.
Dan Djurdjevic said…
We have no evidence of what Motobu meant by inward turned feet. Presumably he meant more than was evident in Motobu's kata (inside edge straight - giving a slight outward bias). Outside edge straight might be all he meant given that this gives a slight inward bias. I adhere to this view since all early photo and film evidence of Itosu's students (eg Chibana) is consistent. To my knowledge not one bit of that evidence points to a greater inward turn, never mind something as radical as the A stance.
Dan Djurdjevic said…
As to the applications benefiting from "A" stance, I have found many forms featuring such a stance. However it is worth noting that before any sideways cross-step (assuming there is one at all - and sideways stepping in such forms is very rare), the "A" stance is always adjusted (by pivoting) into another stance entirely (eg. usually something like "renoji dachi" - think a cat stance foot position, but both feet on the ground and largely even-weighted).

A good example would be wing chun's muk yan jong (wooden dummy) form: you adjust from "A" stance into a "renoji dachi" before making sideways cross-steps, which typically end in another renoji dachi. At some point when facing directly forward, you pivot from renoji back into "A" stance.

But otherwise, sideways cross-steps are never, ever done from "A" stance directly.

I have never heard of this sort of "adjustment" or "pivoting" into another stance being done before the steps in naihanchi. Nor does the stepping imply such adjustment/pivoting in my view. And yet I think this would have to be the case if one were to step efficiently side to side in "A" stance.

So, with respect to those who do the "A" stance, I think it is biomechanically wrong. And I don't see any reason to assume that Itosu and others of his era would have used such an odd construct. Yes, it is possible. But in that event it is surprising that I can't find a single old photo of such a stance being executed - by anyone, including Chibana. I might be wrong - if you can find such a photo, I'll be obliged to change my view! You would think there would be at least some pictures of elderly Itosu students performing in, say, the 60s...

In the meantime, I think it is telling that Motobu's stance was clearly biased outward (as you can see from this photo I suspect that even the slightest inward bias (ie. outside edge forward) would have been an anathema to him! So when he speaks of "inward" stance I don't have a problem imagining something very different from the "A" stance - and far more consistent with the available photographic record.

In any event, one rarely sees a whole lineage of movement "disappear", only to "reappear" half a century later. When it does, someone has "rediscovered" something. It hasn't been "passed down". To me, the absence of photographic record of "A" stance naihanchi until recent decades indicates a "rediscovery".

I think that before this "rediscovery", naihanchi was as it has always been: Some folks keep their outside edge straight, some their inside edge straight. Some are in-between.

Some even go further to do shiko, some do A stance. But the latter 2 are, in my view, either "copy errors", some sort of literal interpretation of what people have read, or a modern concept loosely justified by references like Motobu's.

But I am more than prepared to move from my theory with even a little photographic evidence of an "A" stance naihanchi variant - say from an old Itosu student performing in the 60s! Is there such a thing? It would be exciting to find it!
ojisan said…
Dan and David,
While Funakoshi claims Azato as his primary teacher, he states that he studied naihanchi under Itosu for almost ten years. Given that, I think his foot placement would tend to reflect what he was taught. (albeit he changed a lot of his forms)
ojisan said…
We have photo and film evidence of naihanchi from someone who claimed to have studied that form under Itosu for almost ten years: Gichen Funakoshi. Now I grant you that he (Funakoshi) changed a lot of the stances in his other forms, but the 1924 film footage suggests that he had not yet tinkered with "Tekki".
Dan Djurdjevic said…
Thanks ojisan.

While I'm not sure that the footage is actually from 1924 (I suspect it is at least a decade older) I would agree that it is still unlike modern shotokan (longer, deeper stances, etc.).

To my mind, the concept that Funakoshi should be dismissed as an authority on Itosu's karate is not valid. He, like the other masters, holds part of the puzzle.

My point about Azato was not to detract from this, but more to explain how "unique" things (present in the early footage) like the haiwan nagashi uke might have come to be in his kata.

Otherwise, I would be very surprised if any student of Itosu had the radically turned-in "naihanchi dachi" we see today. I think Funakoshi and Motobu and others are remarkably consistent in these things. There is more consistency than there is difference imho.

Thanks for reading and for your comments.
TimoS said…
Interesting article, but one issue though: Kyan was not a student of Itosu. Kyan learned his karate from his father, Matsumura, Matsumora, Maeda, Oyadomari Kokan, Chatan Yara and someone whose name I can't remember but taught him kata Tokumine no kun. All but one of Kyan's students (Nagamine) say he didn't study with Itosu and so did e.g. Chibana. When you look at the kata in Seibukan and compare them to the Itosu lineage, you see that they're not the same. Also, Kyan did not teach the Pinan kata nor did he teach the dai and sho versions of kata, unlike Itosu. If Kyan had been Itosu's student, I think he would've passed on the Pinan kata, his karate would look more like those in Itosu lineage and/or he would've taught the sho and dai versions of kata, but since none of those match, it is more likely that Itosu did not teach Kyan. I think the misunderstanding stems from Shoshin Nagamine's book, because as far as I'm aware, he's the only one who made the connection.
Dan Djurdjevic said…
Thanks for the clarification Timo! Yes, was going on Nagamine's book, and not being of any lineage connected with Kyan, I did not know he wasn't a student of Itosu's. It makes perfect sense, as you say: he would have passed on Itosu's pinan series.

Thanks again!
wasaba said…
Hello. Great dissection on Naihanchi. Looks like I'm late to the discussion! As an IR practitioner, I'd just like to add a few thoughts. To establish a background, my instructor was taught personally by Tatsuo Shimabuku and received his Shodan from him. Re: the stance. We do practice the pigeon-toed stance. Admittedly, in films TS doesn't use it; in fact, many of his stances differ from what he taught -- at one point when he was teaching he basically implied -- do as I say, not what I do. After decades of practice, I've come to believe that our stance is an exaggeration, but not entirely incorrect. We view Naihanchi as a complement to Sanchin, so you can imagine some similarities in physical structure. From a LFF Sanchin stance, if you turn your hips (torso) about 15 degrees to the right, you assume the Naihanchi stance (on that angle) that we practice. Additionally, the toes-in accompanied with a deep knee bend (difficult) can offer protection to the groin by capturing a kick with the upper thighs. Also, we do the palm up hook block and we do the bent wrist strike - which we practice as a 'bridge of nose' strike. In closing, I might add that TS deviated from the norms of accepted Okinawan practices when he created Isshinryu. While I appreciate the historical practices, I long ago accepted his interpretations without criticizing others. Thanks for your diligent research.