Saturday, October 27, 2018

On the Structure of Taekwondo Forms, Part 2

Intended Meaning

To summarize part 1, the sets in taekwondo forms are a mix of self-defense ideas taken from early 20th century Judo, karate manuals, and the creators' own ideas. This means that each set has an intended meaning behind it; and forms are not ambiguous collections of movements meant to be interpreted as the student desires. This is not to say you cannot apply sets in other ways -- instructors can and do -- but the forms still have intentional meaning.

A good application fits three criteria:

1) Expository Value: Does the application explain the movements in the form? What about footwork, stance, the secondary hand, or special instructions?

2) Practicality: Would this application work in a live context against a realistic attack? Does it adequately protect the user?

3) Continuity: How does this application fit in with the rest of the form?

Because of this criteria, the number of feasible applications is quite small. It's true that an individual movement, such as the low block, has many uses, but once you put that movement in a greater context its possible meanings shrink.

An unlikely application for the "mountain pushing posture"  in
Cheonkwon. A literal push does not explain the cat stance. A
front stance would serve this application better.
This also means that we should not have to add footwork, change stances, change open-handed movements to grabs, etc. to make an application practical. We should trust that the creators of the forms made them the way they are for a reason. This is also why it's important to study the official versions of the forms, as they tend to pick up changes as they get passed from school to school.

A much better application that has the bonus of explaining the
"single mountain block" appearing beforehand. The cat
stance is used as a trip. Turning around, as done in the form,
will drop the opponent onto their head. Thus this application
explains more of the form. Source: TomHillsVision

Rather than find multiple applications for a given set in a form, it is more useful to find that set's intended application and then that technique's applicability. Suppose you agree that the twin upward palms in Choong-Moo encodes a double leg takedown. Rather than try to explain how the motion could also be a double elbow lock, or a release manuever, it's more useful to train the double leg with a partner, discover new entries to the technique, how to counter it, what to do if it fails, etc.


For the record, I am not saying that I know what all the original applications are, that interpreting forms is easy, or that you cannot use alternate interpretations. What I am saying is that if we look for the historical meaning, then there is ultimately one correct interpretation.[1]

Changes to the ITF forms

Studying older versions of the forms can be useful in determining their meaning. The original 20 ITF forms underwent minor revisions between Choi's 1965 book and his 1983 Encyclopedia. Are the updated forms better or should we use the original versions?

Most of the changes do seem to have a purpose. The X-fist pressing block in Sam-Il was originally performed turning 90-degrees to the right; in the modern version you turn only 45-degrees. It's hard to believe this change was made for aesthetic reasons, but when you know the application then the change makes sense. The twin block set in Hwa-Rang was changed to take advantage of the new cross-arm setup. Tong-Il underwent significant revision between 1965 and 1983, seemingly to make it more challenging, but the 1983 version is nonetheless logical in design. To me, the revisions are evidence for rather than against the existence of applications.

The original (left) vs the modern
(right) bending ready stance A
However, some of the changes are a side effect of changes in hand techniques. The first five forms presented in Choi's 1959 book use the Shotokan karate low block setup. By the time the original 20 forms were finished in 1965, the wrist-to-wrist setup was being used. Between 1965 and 1983, the bending ready stance was changed. Originally one would bring both fists to the opposite hip -- as in karate and KTA taekwondo -- but this was changed such that both arms extend outward. While these changes do have applications (e.g. the wrist-to-wrist setup is brush-grab-strike), [2] application researchers should keep in mind what the original versions of the forms were.

Striking While Grappling

Reading through these two posts, you might be getting the impression that forms are mostly grappling techniques, but that isn't quite right. The dichotomy between grappling arts and striking arts is a modern one. Most historic arts taught both striking and grappling before they became specialized over time. Even Western boxing used to allow limited grappling.

The folks at Karate Culture have described karate as the art of "striking while being grappled", and we can say the same about taekwondo forms. Punches are still punches even if we pull the opponent with the other hand. Backfists, hammerfists, knifehands, elbows, knees, kicks, and throat strikes are found throughout the forms. Even the basic blocks may at times be used as strikes: the low block becomes a hammerfist to the groin; the inner-forearm block, a rolling backfist to the face. General Choi stated that of the three self-defense options against a grab -- striking, releasing, and joint breaks -- striking was the fastest and most likely to work. Hwa-Rang and Choong-Moo, despite the Judo influence, both begin with a prosaic block followed by a strike.

But because you are likely to be grabbed in a self-defense scenario, it is important to actually know how to grapple. Most sets in forms end in takedowns; sets that do not usually end in striking the vagus nerve or back of the head. Locks, while present in forms, are rarely to submit an opponent. They are instead used to break a joint or to position an opponent for either a takedown or a knockout strike.


Interpreting Forms

This post assume the reader has the basics down: that "blocks" are not always blocks, that you face only one opponent, and that as a rule both hands are utilized. Therefore, I wish to cover some non-literal instructions in forms that students should be aware of.

Slow motion movements denote special grappling maneuvers or high-resistance applications. The slow motion twin elbow thrust in Po-Eun is lifting your opponent off the ground, for example. However, there are plenty of normal-motion movements that represent grappling and some slow motion movements that may be applied as strikes, so do not feel constrained by this rule. [3]

Although stomping is sometimes used to attack the leg, more often it represents lifting your leg for a knee strike, rapidly dropping your weight, or both. A downward elbow strike is an overlooked application for the mountain block/W-block.

One-legged stances often represent sweeps. Right images are
from the Encyclopedia of Taekwon-Do.
Stances can have special uses. Front foot X-stances are typically used to trap the instep. Rear foot x-stances can represent leaning onto your opponent. Cat stances are often trips. One-legged stances can be knee strikes or sweeps. Even the front stance may be used to destabilize an opponent by pressing your front knee into their leg.

Jumps have varying purposes. The 360-degree jump in Choong-Moo represents lifting an opponent from a squatting position, and hence is not literal. But in other forms an upward jump may be used to complete a joint break or to get at the top of the opponent's head.

Finally, kicks in forms are usually aimed at the legs, even "middle" and "high" kicks. A low kick indicates that the opponent's leg is very close to you.

Ready Positions

Several forms have a special ready position. With the possible exception of Juche, these ready positions always have an application, but not always at the start of the form! The intended meaning is just as likely to be at the end of the form instead.

If the opponent is too far away to sweep, place their leg onto
your shoulder and push down on their knee with both
palms to throw. Source: EffectiveMartialArts
One example of this is in Cheonkwon. Suppose you cannot apply the mountain pushing posture as a trip and face push because the opponent backs away from you. If the opponent does this, then they have straightened their leg. We can hoist their ankle onto our shoulder and push down their knee with both palms, taking them to the floor. Because this is meant to be used if the previous application fails, it is an example of a chain sequence.

In contrast, I struggle to find a use of the overlapping hands at the start of Cheonkwon. I do not believe there is one.

Gooseneck lock.
An example from the ITF forms is the covered fist at the end of Won-Hyo. This may be utilized as a gooseneck wrist lock, another technique that appears in General Choi's books.

Ready positions occasionally appear in the middle of forms, such as in Ul-Ji and Taeguek 7, further supporting that they have applications.


Sometimes the applications in a form are built around a theme. In Hwa-Rang we control the opponent's arms. In Choong-Moo we attack the base. In Ul-Ji we attack the head. The nine (or ten) KTA black belt poomsae have thematic elements as well. Koryo relies on hard and fast striking. Keumgang emphasizes stability. Taebaek is based on the Heian kata. Themes are never universal, however, and I doubt that every form has one. While knowing the theme can be useful for discovering applications, it is easy to guess the theme wrong. The internal logic of the form is more important.

The Structure of Forms

Finally we reach the climax. Just what is the internal logic of a particular form?

Keep in mind that not all forms are designed using the same rules. This is especially true if you study a style which uses forms from many different creators and time periods, such as karate. The scheme I describe below just happens to work for many ITF forms, and has proven useful for analyzing the KTA black belt poomsae as well. In engineering we have a maxim that "all models are wrong, some are useful." I cannot say with 100% certainty that the below models of Hwa-Rang and Choong-Moo are true, but they have proven useful for analyzing both of the forms.

The general rule is this: in a taekwondo form, you only ever defend against one de novo attack from one opponent, at the start of the form. Every other set in the form is an alternative follow-up or a chain sequence. This means that you have the same opponent throughout the entire form. Chain sequences are to be used if the opponent counters you in some way, and thus the application depends on how exactly they counter you.

Hwa-Rang, in my view, is mostly alternative follow-ups. Suppose at the start of the form we parry an opponent's right hand punch, and then counter with our own right hand punch. If the opponent parries our punch, then we have several options, the first of which is following up with a left hand punch (step 3). In other words, almost every set in Hwa-Rang begins with the opponent parrying your right hand. An example of a pure alternative follow-up design, for a form with 10 sets, is shown below.
What is the benefit of such a scheme? Primarily this is just a way to organize and transmit ideas. However, by learning only one defense against a particular attack, you do not have the crippling problem of choice. You use the defense and then if it works, great, if it doesn't, then we have to ask questions. After the first set in Hwa-Rang there are multiple variables: Are you punching with your front hand or back hand? How close are you to your opponent? Did you succeed in grabbing and pulling the opponent's right arm? Did the opponent grab your arm? Each scenario provides different options. 

On the other end of the spectrum is a chain sequence, which better describes Choong-Moo. We assume the initial technique fails, and so we follow with a second one. Then if that fails we follow with a third technique, and so on. This process can continue through an entire form.

Notice that I still added a couple branch points in this chain sequence. This is because a technique often can fail in multiple ways, so occasionally the form will provide more than one follow-up. These make the form difficult to interpret, but provide meaningful information about how your opponent may counter your technique. The key again is to ask questions about how a defense might fail.

Most forms have a mix of the two structures, with the chains only going a few levels deep.

In all three example cases, the form is designed to help you survive a single encounter, not multiple encounters with separate opponents. This is what is really meant when we say forms simulate a fight. A good form predicts the opponent's most likely response(s) and tells you how to deal with that.

Examples of the starting attack include a haymaker, a straight punch, a lapel grab, a rear bear hug, a tackle, a football kick, or a shove. These attacks fall under "Habitual Acts of Violence", attacks you are likely to encounter in a self-defense scenario. The opponent may throw additional attacks as a counter to your techniques, but under this guideline you will never be attacked by a new opponent.

I suspect not all forms are built this way. Po-Eun might be an exception. But this is a good starting point for analyzing forms.


I believe the old method of keeping form applications secret is out of date. We live in the age of the internet where fighting knowledge is at one's fingertips. What's more, we do students a disservice when we teach them impractical applications for their forms. ITF taekwondo in particular bills itself as a self-defense art. So why not study the forms a little deeper? And if you want to study taekwondo as it was historically developed, isn't it worth studying the whole art, locks and throws included?


I received some comments about using the palm pushing block against a punch. This is one of two applications shown in volume 10 pages 141-3 of the Encyclopedia of Taekwon-Do. The other application is against a side kick, which is not an attack I expect to face in a self-defense situation.

Hwa-Rang application from Vol 10 of the Encyclopedia
The idea is that an untrained fighter is likely to over-commit to a punch, so you side-step the attack and push their upper arm or shoulder. If they do not over-commit then the push becomes merely a parry. Of course, no one throws a karate-style punch in reality, nor do they stand still waiting for you to punch. I would modify the above application such that the opponent is facing your front, and that you grab and pull the opponent's arm.

Another comment was that the ITF forms are not meant to be fighting templates. Who said this? Not General Choi. In his 1965 book he writes.
The "pattern" is thus a set sequence of movement of attack and defence in a logical order. Imaginary opponents are dealt with in sequence logically and systematically under the assumption of various situations
Other early taekwondo pioneers said similar. In Tan-Gun and To-San Jhoon-Rhee wrote
A student’s sparring or fighting style becomes his adaptation of the principles he has acquired from hyungs. The hyungs, then, are the student’s line between Tae Kwon Do training and actual fighting
Jhoon-Rhee also stresses the logical nature of the form. Taekwondo pioneers were clear that forms were meant to convey fighting principles. This is not to say that performing forms makes you a good fighter, but the entire point of forms is to record and transmit ideas which you may use in a fight.

Another comment was that forms are only for practicing movements. But you don't need a form to do that. We already do that in basics. My view is that in basics we practice movements with multiple possible applications, but in forms we see examples of how those movements may be used.


[1] An article that covers this in more detail is An Open Letter to Bunkai Researchers, although some of the points do not apply to taekwondo forms.

[2] Some assert that the only purpose of a bending ready stance is to set up a side kick. The problem with this logic is that not all side kicks are set up with the stance, and in Gae-Baek the stance is used without being followed by a kick. I read one taekwondoin state that the purpose of the stance was to "look intimidating". Needless to say, I disagree.

[3] A universal rule for slow motion movements has proved elusive. Another explanation I have read is that there is a detail besides power and speed needed to make the application work. But this too does not appear to be a universal rule.

1 comment: