Core Principles and the Taiji Classics

拳论逐句讲:十要、十三势,以及几句被读反的话

Taijiquan has a small body of short technical writings that practitioners treat as authoritative. Collectively they are the taiji classics, and they are unusual: dense, aphoristic, largely free of mysticism, and concerned almost entirely with how to organise a body. Much of what is best in the tradition is in them.

Core Principles and the Taiji Classics

This page covers what they are, where they came from, how to read them without being misled, and the specific principles that matter most — including two, double weighting and following the opponent, that are constantly quoted and frequently misunderstood.

What the corpus contains

The core is half a dozen short texts. There is a general treatise on the art, attributed to a figure named Wang Zongyue about whom nothing independent is known. There is a verse on the thirteen postures, and a prose commentary explaining how to apply them. There is a short song about partner work. And there are several essays by nineteenth-century authors in the circle around Wu Yuxiang.

Together they run to only a few thousand characters. This is a very small body of text to carry a whole tradition, which is part of why it has been commented on so heavily and interpreted so variously.

Where they came from

The classics were compiled and circulated in the nineteenth century by Wu Yuxiang and his nephew Li Yiyu, whose handwritten copies from around 1880 are the earliest datable versions.

Attributions within the corpus are unreliable. One text is credited to Zhang Sanfeng, which is certainly false. The Wang Zongyue treatise is said to have been discovered by Wu Yuxiang's brother in a salt shop, a story historians treat with reserve, as discussed on our Origins page. Some texts attributed to earlier figures were probably written by the compilers themselves.

Uncertain provenance is not the same as low value. These texts are technically excellent whoever wrote them, and their authority among practitioners rests on the fact that they describe things practitioners can verify, not on their attributions.

How to read them

This is the most useful thing on this page. The classics are not instructions. They are memoranda written by people who already had the skill, for people who already had the skill, as reminders and as tests of understanding.

This explains their characteristic difficulty. A line like the instruction to distinguish substantial from insubstantial is not opaque because it is profound; it is compressed because everyone in the intended readership knew what it referred to and needed only the reminder.

The practical consequence is that reading the classics will not teach you taijiquan, and expecting them to will produce frustration and elaborate misinterpretation. What they will do is give you language for things you have already felt, and tell you what to look for next. Most practitioners find that the same line means something entirely different after five more years of training, which is a sign the text is working as intended.

The corresponding warning: a source that explains a classical line in language even more obscure than the original has not understood it. Genuine explanations of these texts are almost always more concrete than the text, not less.

The ten essentials

The most useful single summary is a list of ten requirements dictated by Yang Chengfu in the early twentieth century and written down by a student. They are practical, checkable, and cover most of what a teacher will correct.

The head is held as if suspended from above, without tension in the neck. The chest is slightly contained and the back slightly extended, rather than thrown open. The shoulders sink and the elbows hang down. The waist is relaxed and functions as the axis of movement. Substantial and insubstantial are clearly distinguished, which is to say your weight is definitely on one leg or knowingly divided, never vaguely between.

Then: upper and lower body follow each other, so that hand, waist and foot arrive together. Inner and outer are unified, meaning intention and movement are not separate. The form is continuous, without breaks between postures. Intent is used rather than muscular force. And stillness is sought within movement, so that the breathing and the mind remain settled throughout.

Almost every correction you receive in your first years will be one of these ten, and it is worth knowing them well enough to check yourself.

Eight gates and five steps

The thirteen postures are not thirteen movements. They are eight ways of applying force and five directions of movement, and every technique in the art is built from combinations of them.

The eight are usually listed as warding off, rolling back, pressing, pushing, plucking, splitting, elbowing and shouldering. The first four are the fundamental pair-work energies practised in basic push hands; the second four are considered corner or supplementary methods. The English names are poor and each requires demonstration to convey.

The five are advancing, retreating, looking left, gazing right and central equilibrium. The last is not a direction; it is the condition of remaining balanced, and its inclusion in a list of steps is a deliberate point — you have not moved correctly unless your balance is intact at every instant.

The eight are correlated with the trigrams and the five with the phases. As we argue in the Chinese Metaphysics section, such correlations are usually retrospective numerological fitting rather than derivation, and this one is no exception. The eight energies are real and useful; the trigram assignment adds nothing to what you do.

Double weighting

This is the most important technical concept in the classics and the most misunderstood.

The treatise says that if you are heavy on both sides you become stuck, and that many years of practice can be wasted by failing to understand this. Practitioners commonly read it as meaning that your weight should not be evenly divided between your feet, and stances are adjusted accordingly.

That reading is too narrow. Double weighting is a condition of force, not of stance. It occurs whenever you meet an opponent's force directly with your own along the same line — when both of you are committed in opposition at the same point. At that moment the exchange becomes a contest of strength, which is precisely what the art is designed to avoid, and the more skilful position is lost.

The remedy is to make one side substantial and the other insubstantial, so that the opponent's force finds nothing to press against on one side while you act from the other. This can happen with the weight evenly distributed and can fail to happen with the weight entirely on one foot. Stance is a special case; the general principle is about not opposing.

Abandoning yourself and following the other

The instruction is to give up your own intention and follow the opponent's, and it is often taken to counsel passivity. It does not.

What it describes is a method. If you have a plan and your opponent does something else, you must abandon the plan or force it, and forcing it reintroduces exactly the direct opposition the art avoids. Following means letting the opponent's movement determine your response, so that you are never in the position of having to overcome them.

The classics are explicit that this is not surrender. Following is paired with sticking and with the idea that you yield when they are hard and adhere when you have the advantageous position. You follow in order to arrive somewhere better, and the moment the relationship turns, you are already there.

Practically, this is why sensitivity training matters more than technique memorisation. You cannot follow what you cannot detect.

Four ounces deflecting a thousand pounds

The best-known line in the corpus, and the most over-claimed. It is worth stating what it does and does not assert.

It does not assert that a weak person can beat a strong one, or that strength is irrelevant. Read in context, it appears in a passage about how a frail old man can prevail against several opponents, and the point being made is about the efficiency of skill, not the irrelevance of physical attributes.

What it describes is real and ordinary: a small force applied at the right moment, in the right direction, to a structure already committed and already unstable, produces a disproportionate effect. Anyone who has tipped over a loaded wheelbarrow at the wrong moment knows this. The skill is entirely in the timing and the angle, and acquiring it takes years.

All other things being equal, the stronger practitioner wins. The point of the line is that all other things are rarely equal, and skill can make the difference larger than strength can.

What the classics leave out

They say almost nothing about how to train. There are no exercises, no progressions, no timescales, no advice about conditioning. This is a substantial omission and it means the classics must be read alongside a curriculum, not instead of one.

They say very little about applications. The eight energies are named but not described in use, and the corpus is silent on most of what a fighting method must contain.

And they assume a teacher. Every line presupposes someone who can put a hand on you and show you what the words refer to. Reading them without that is like reading a manual for an instrument you have never touched.

Key points

  • The classics are a few thousand characters in total, compiled in the nineteenth century, with attributions that are unreliable and technical content that is excellent.
  • They are memoranda for people who already have the skill, not instructions; expecting them to teach the art produces frustration and elaborate misreading.
  • Yang Chengfu's ten essentials are the most practical summary and cover almost every correction a beginner will receive.
  • The thirteen postures are eight force applications and five directions, one of which — central equilibrium — is a condition rather than a direction.
  • Double weighting is about meeting force directly along the same line, not merely about even weight distribution between the feet.
  • Four ounces deflecting a thousand pounds describes efficient timing and angle against a committed structure; it does not claim that strength is irrelevant.