Martial Arts Stack Exchange is a question and answer site for students and teachers of all martial arts. It only takes a minute to sign up.
Sign up to join this community
I've been looking for a semantic ontology defining the structure and varieties of attacks and defenses. Initially I'm looking for a very simple tree, something like:
Attack->
Punch->
Jab
Hook
Uppercut
Hammerfist
Backfist
Kick->
Front
Roundhouse
Side
Defend->
Block->
Move->
Retreat
Lateral Evasion
etc. etc.
Does anyone know of any attempt to systematize information about a Martial art or generically about combat in this way?
Thanks!
The BJJ community is big on mind maps, which are close but not an exact match in your search for ontologies. For instance, Aesopian has this one:
This is not surprising, since the entire concept that set BJJ apart from judo was the idea of an inexorable flowchart:
or
The original vision now has many new spin-offs, but the underlying philosophy of having a gameplan has had a tremendous impact on jiu-jitsu mentality, as well as mixed martial arts strategies. Very few dispute the truth of the positional hierarchy, which when boiled down simply says it's better to be on top than on bottom.
Rickson was arguably the best in known history at implementing this plan. Here are his flowcharts:
Mapping my BJJ journey has more (including Rickson's), but most are not very "complete", preferring instead to just explain the contents of a particular DVD. For example:
In my opinion, striking arts do not lend themselves particularly to mind maps or ontologies. There are way too many options--essentially it would just be a catalogue of techniques repeated several times over. The style of judo I'm familiar with doesn't use that kind of approach either, though I've heard that Russian and other Eastern bloc judo and wrestling coaches emphasize "chained" strategies that are similar. For instance, they'll attack with a particular technique, knowing that the opponent can only reasonably counter with A, B, or C, and for each of those they will have a fully fleshed out response, all the way down the chain to victory.
I don't know of an existing ontology, but we can create one. Everyone should feel free to edit this to make it more complete. I have organized it with the headings (big and small) as well as lists as end nodes. I expect it will become cumbersome very quickly. Feel free to rearrange. (Perhaps the lists would be better off paragraphed instead of bulleted, for instance.)
I've used the OP's ontology as a starting point, and have added judo's throwing syllabus (in English) as a basis for additional techniques.
Regarding the BJJ Ontology. I started putting one together a few months back. You can check it out through the WebProtoge project named BJJ - Brazilian Jiu Jitsu at http://webprotege.stanford.edu/
It's very similar to what Dave Liepmann has put together in his mindmap. The BJJ Ontology I started is described in the OWL (Web Ontology Language). I pulled most of the terms off the web by google searches. I am hoping to expand the ontology in the coming months.
If you want to systematize things, I suggest taking a look at a technique's 'tactical' value. Is it an attacking or defending technique or maybe a bit of both? For example, the concept of the Tactical Wheel from Western Fencing (and I think this is what Dave Liepmann was referring to with his comments about Russian and Eastern European coaches) has proved to be a useful tool in characterizing actions for sword arts. IMO, it's a good basic primer to tactics in general and, if you don't mind equating a kick with a punch or a sword thrust, it's can come in pretty handy. The Szabo tactical wheel (created by Lazlo Szabo) has basically four points, each successive point being a counter to the previous one:
Simple Attack (a single offensive action, such as a punch or a thrust)
Parry/Riposte (or Block/Counter, which is a defensive-offensive set of movements)
Compound Attack (an attack made evade the expected block, or nullify it, e.g., fake high, go low)
Counter Attack (a simple attack while the opponent is preparing)
There are also additional actions that can be inferred from this simple set of tactics. For instance, maybe you want to invoke your opponent's Parry/Riposte, so that you can counter riposte? Or maybe we can use their Compound Attack to set up a Parry/Riposte of our own? And so we go spiraling off down the halls of possibility.
I admit that it's simplistic, but since every martial artist needs some framework for setting up a strategy, it allows one to do so with very little information from your opponent. It also allows novices to start working with tactical concepts, even though they only possess a few techniques. Additionally, I would note that, as with all systems, things start to break down when you can no longer determine what is happening. As with anything, it only works when one can see clearly.
I have to admit I was stymied by the concept of "ontology", so, I had to look it up. And when I did, I realized it was a philosophical concept. After reading about it for a bit, I decided my answer was that yes, there has been an attempt to systemitize a martial art, but no, not a way for combat. Not exactly, anyway: In some styles, notably taekwondo, tangsoodo, karate, kungfu, etc, there exists the concept of forms (kata, poomsae, hyung, taolu). It is in forms that define the style. Of course, some styles have many forms, others changed their forms, others have only a single form, others have forms but seem to have no idea why...
That's your philosophical answer: the forms are the key to the style. The ancients knew this; it was the only means to record their style and pass it on. Some played games, like hiding techniques into representations of something else. For example, many techniques in Capoeira are hidden in dance. Many techniques in Karate and Taekwondo are defined by hitting yourself. Always, these techniques are representative of something else (who in their right mind hits themselves in self-defense??) While this will form a tangent into politics which has nothing to do with your question, the basic answer is yes, it has been done before.
However, others here seem to have concluded a different meaning to ontology, and so, I did more research and decided they're also quite right. It sort of follows that there is a canonical list of techniques within the style. I can say with assurance that two of the styles I practice - Kukkiwon Taekwondo, and ITF Taekwondo - the former has a textbook containing all of the techniques, while the latter has the famed Encyclopedia. The textbook and encyclopedia are rather incomplete, and form the basis for many who complain bitterly about the styles: both tomes contain techniques not seen in forms and yet not allowed in competition. Or they're missing techniques which are commonly employed in competition (eg, tornado roundhouse).
So I would say that, yes - there has been an ontology created for Kukkiwon and ITF Taekwondo, but it differs from your example in two distinct ways: first, the textbook and encyclopedia are complete as written, although not complete as practiced. Your example is, of course, not intended to be complete - you're asking for details to be filled in. The second way is that the techniques are categorized into a mind map or chart.
However, despite the books' shortcomings, I also see limitations and limited use of an ontology. What use could it have? Listing a technique that is never used, as it the case with the mentioned books; or naming or qualifying a technique, as you and others here have done. This method of categorizing techniques is what limits - not expands - our knowledge. Giving something a name makes that thing useful only for what its name implies. Some Chinese styles had it right, in my opinion: things get names from concepts having little or nothing to do with its purpose. Otherwise, we have things like "rising block", and suddenly, this thing is categorized as a defensive maneuver which can only be used for blocking. Never mind that it has striking, grappling, and parrying purposes: this thing MUST be a block, and ONLY a block, and NOTHING else. I don't agree with that mantra, which is why I'm careful not to categorize techniques, and refer to them by name only when I have to.
If you prefer to have an ontology, perhaps making the techniques more generically categorized: hand techniques, foot techniques, body position, etc would be more helpful. It would reduce the repetition of a technique appearing in more than one place.
Also, styles have techniques with no name at all. Aikido, for example, has henka-waza which are derivatives of named techniques, but executed in odd ways. Throws are a good example here: sometimes, a throw is achieved simply because the opponent has grabbed onto you, and you need not grab back at him to throw (this is accomplished via the open knifehand "blocks"); or, your opponent is in position for an easy throw but has not grabbed you, so you must (this is accomplished via the closed-fist "fighting" or "on guard" position or posture (the closed fist is the hint we are grabbing)). And yet the knifehand and closed fist have several distinct other uses as well: will you reflect this in the ontology? Will an ontology be allowed to have duplicates - two purposes for a single named technique?
Another example is the double-face punch, seen in more than a few TKD forms. Of course, these are not really face punch, yes? This would be highly inefficient (given some of us have established that a chamber is needed to add power). Hint: you are grabbing the opponent's ears. Or hair. So your ontology must reflect this: one technique for two distinct purposes. Do you list "double face punch" (which is not a punch at all), or do you list these two: "hair pull" and "ear pull"?
Next, what about context? The answers listed so far don't mention "stances", which I guess is good. But the ontologies I listed (the textbook and encyclopedia) DO mention stances, and here, this is a very useless item.
If you were going to mention stances, you might as well go ahead and list other techniques, like "head facing left", "head facing right", "turn the torso clockwise", etc.
While these are simply positions of the body, or transitions from one place to another, they arguably are not techniques any more so than a stance is also a body position or a transition from one place to another.
On the other hand, stances are one of those things we have which don't really clue us as to what the thing is doing by name alone. You know, like "cross stance". Is it offensive? Defensive? Both? What's its true purpose? Where does it go in your ontology? Some stances - like a crane stance - absolutely have offensive and defensive movements. For example, the crane stance (where the foot is not wrapped behind the knee) is representative of the person lying on his back with the opponent on top of him. The bent knee keeps the opponent from completely taking over, and affords the possibility that we can push off and roll the opponent off us. All that in a "crane stance", which isn't a stance; and if you're lucky, the outward knee goes to the opponent's groin or solar plexus. But it isn't a kick, because that would imply a transference of energy from defender to offender, whereas here, the offender is transferring his mass and energy onto the fixed and unmoving knee.
Oh, speaking of knees... So the other crane stance has us wrapping our foot behind the knee. Subtle difference - except that with the former, we're on the ground on our back, while the latter is used whilst standing. Not very subtle at all, yes? In this case, we're wrapping our front leg behind the opponent's, to effect a trip. This is one of those "hidden" strategies I mentioned before. How are you going to relate all this in an ontology? And that cross stance? That's usually a strike (using the knee) to a pressure point in the opponent's area of SP10, ST36, SP09, SP08 (or if you prefer, ou-li/kee-moon, shitsu-kwansetsu/dohk-bee, kokotsu/um-nung-chun; etc) This is something else that ought to be in an ontology, if you're striving for completeness and information.
What about chambers? Here's where we get into these philosophical arguments. Are they preparatory movements for a strike, or are they a technique in of themselves? How do you put it into the ontology? Should you? Where does it go?
What about breathing? And kihap/kiai? These are used for many purposes some say for power, pain absorption, harmonizing, aim, and maintaining stamina. It's a technique. Gotta put it in there. But where?
You see? Your list is 1-dimensional - a top-down list. If you categorize the techniques, it becomes 2 dimensional, but immediately becomes complicated because a single technique can traverse more than one category. Do you remember the movie Contact, with Jodi Foster? Do you remember how the aliens conveyed instructions to Earth for the knowledge to building the machine? They could have used a set of instructions listed top down. After all, that's how the machine got built to begin with. Instead, they made the instructions more efficient by using math, and by compressing it into a third dimension. It's the same here: your forms (poomsae, kata, taolu) are the most efficient means for representing your ontology.
With forms, you can have a 5-dimensional ontology: not just a top-down list of techniques (and no categorization) (and no names, except the part where you're being taught the form's movements), but movement, timing, and alternative interpretation are also included. All neatly represented - in context - which could never be represented in a top-down approach, whether it is a top-down list or a mind map or a chart of some sort.
Your written ontology might have 50 techniques in it. Your form may be defined as having 20 movements, but will be hundreds of techniques. And will include the "esoteric" stuff I mentioned: breathing, chambers, stances, etc. And the more you study your forms, the more techniques you will derive out of them. In this way, your written ontology is static and can be memorized - but no learning occurs. With forms, the movements are static (that is, they don't change), but the techniques drawn from them are nearly endless - and that's where much of the learning takes place.
Sources: Encyclopedia of Taekwondo; author: Choi, Hong Hi ISBN13: 978-1897307762 ISBN10: 1890307764
Kukkiwon Taekwondo Textbook (Korean-English); published: Osung ISBN13: 978-8973367504 ISBN10: 8973367501
If your style has forms, there's your ontology.
