Why Is Fengchi Called 'Wind Pool'? Acupoint Names May Be Compressed Codes Rather Than Poetry

The name Fengchi—“Wind Pool” is easy to treat as ancient literary imagery: wind arrives here and gathers like water in a pool, hence the name.

But when more than three hundred acupoint names are placed together, the ancients can be seen repeatedly using a limited vocabulary of roots:

Gate, pass, door, surge, meeting, crossing, communication;

pool, sea, spring, stream, valley, abyss;

palace, storehouse, hall, dwelling;

Heaven, Earth, upper, lower;

spirit, soul, numinous, Yang, Fire …

These characters may not have been chosen merely for memorability. They may compress a description of what role this location plays in the overall system.

The character “Wind” is the most striking test case

In the current data, six acupoints contain the character “Wind” (feng):

• Fengmen
• Fengshi
• Fengchi
• Fengfu
• Bingfeng
• Yifeng

In the existing Six-Qi vectors, all six contain Wind in either the primary or secondary label. Five have Wind as the first primary label; only Fengfu has Cold primary and Wind secondary.

This consistency is striking.

But at present it can only be called internal consistency, not independent proof. The reason is that in the previous version of the Six-Qi score, characters in acupoint names themselves contributed part of the signal.

A more convincing experiment must first hide the names, recompute Six-Qi relations from phase, harmonic, geometry, and symptoms, and reveal the names only afterward.

If the Wind-named points still cluster toward Wind, the naming hypothesis will have passed a true blind test.

“Pool, sea, spring, stream, valley, abyss” look more like medium and convergence

Water-landscape roots produce an even more interesting result.

Combining acupoints containing pool, sea, spring, stream, valley, and abyss gives 36 points. Their average Dampness score is about 0.384, compared with about 0.189 for other points—nearly twice as high.

Among them:

24 / 36 = 66.7%

have Dampness as the primary Six-Qi label.

Geometrically, 16 of the 36—about 44.4%—fall into bridge roles.

The most stable signal of these water-landscape roots is therefore not “they always lie at one exact angle,” but rather:

These positions are more likely to carry system characteristics of convergence, medium, low potential, dampness, or transfer.

In other words, the name may first describe a physical character, not a coordinate.

“Palace, storehouse, hall, dwelling” look functional rather than topographical

Another group of roots behaves differently: palace, storehouse, hall, and dwelling.

These terms imply containment, residence, and center.

There are 13 relevant points in the data. Zang-fu or spirit-type symptoms make up about 27.8% of their indications, compared with about 14.3% among other points—roughly 1.95 times as high.

This suggests another naming mode:

Some acupoint names may not describe “a valley” or “a pool,” but rather say that “this location behaves like a containing or central functional space.”

If this relation replicates in more independent datasets, acupoint naming begins to look less like a mnemonic and more like an ancient language of functional classification.

Not every attractive interpretation works

The most important part of this research is also recording failures.

For example, points whose names imply an upper axis through characters such as “Heaven,” “upper,” “head,” or “skull” do show slightly more symptoms of the head, face, eyes, ears, nose, neck, and throat—but only from about 12.0% to 13.3%, a ratio of just 1.11.

That is not enough to support a hard rule such as “if the name contains ‘upper,’ it must treat the upper body.”

Likewise, gate/boundary roots intuitively suggest nodes, but their actual geometric roles are not overwhelmingly node-dominated; phase-shifter roles are often common.

This reminds us that ancient acupoint names are not a physical textbook to be translated character by character.

The same character may describe qi transformation, terrain, function, motion, or anatomical position. Statistics, rather than association, must distinguish among these possibilities.

How much can a name predict?

We also performed a stricter test: give the model only the Chinese characters in the acupoint name, withholding phase, channel, and indications, and ask it to predict the Six Qi or geometric role.

Five-fold cross-validation produced:

• primary Six Qi: about 36.3%, above the majority-class baseline of 29.6%;
• best geometric role: about 41.3%, almost identical to the majority-class baseline of 40.2%.

This result is crucial.

It suggests that acupoint names really do carry some qi-transformational information, but are almost insufficient by themselves to reconstruct precise geometric roles.

The most reasonable interpretation is therefore not “the acupoint name is a phase password,” but:

An acupoint name is more like a compressed label that writes the qi-transformational property, flow character, or functional role the ancients most wanted to remember into two or three characters.

Re-understanding the name Fengchi

“Fengchi” can therefore be separated into two layers:

Wind hints at a qi-transformational character and direction of motion.

Pool hints at convergence, low potential, medium, or transfer.

We cannot force the data to fit the name merely because the name sounds meaningful. But we now have an operational research path:

First hide the name and derive acupoint properties from geometry, phase, symptoms, and qi transformation. Then reveal the name and ask whether the ancient characters hit those properties more often than chance.

If this blind validation succeeds across hundreds of points, acupoint names will change from “ancient poetry” into a very unusual historical dataset: compressed annotations left by the ancients on a channel control system.

That is one of the most promising directions for this project to continue pursuing.