George S. Ledyard
It's not my English that's the problem... it's that we are talking about something that has no vocabulary in English. So you have to work with someone who can do it and explain what he or she is doing. Then you will know what the description means. You have to feel it. You can't see it until you really know what you are looking for.
This is what I have to do with the folks at my dojo and at the seminars I teach. I have worked out a vocabulary which is body centered and quite specific. When I instruct I have to teach the students this new (for most of them) language. So when I say "touch the spine" I have to grab you and give you specific feedback about what that is and is not.
That's why this stuff really doesn't lend itself to mass transmission. There's a reason that O-Sensei basically taught either privately or in very small groups which allowed the students to get their hands on him multiple times every class. Even then, due to lack of systematic explanation, there was a huge variance in the extent to which they got it.
Stuff like this is why when I came across yiquan, or at least one interpretation of it, after floundering in my aikido practice, I found it so intensely refreshing. It was this very idea played out systematically - that whether you put it in English, Japanese, Chinese, or Greek - that so long as explanation in words precedes intuitive understanding of a concept, you are always going to set the stage for misunderstanding more often than not as the listener fishes in his own pre-existing experience for a best fit of what is being said. That even indicts the idea that if you feel someone else skilled at a concept (playing uke), that you are also expected to gain nage's personal intuition of the concept. I always wondered why I felt so unsure of what I had learned, why I kept discarding and relearning things all the time only to go in circles on many concepts; the longer I trained the shakier my foundational understanding seemed to be.
But the training was structured very differently with this, such that it tried to assume as little vocabulary as possible, verbal or even physical. Like when conveying the idea of relaxation, there is no one saying to "be relaxed" or having you see or feel someone who is relaxed and immediately trying to just reproduce that in yourself. It takes the tack of trying to directly induce the feeling or at least something similar to it in you directly, by a tool that can be objectively verified initially. Once the feeling is there, the tool is discarded in favor of deeping the initial sensation and expanding the situations in which it can be employed.
One example tool for introducing relaxation of the shoulder is to have the person extend one arm out straight to the side, bending over at the waist somewhat so the arm has a clear path to swing, then letting the arm suddenly relax and drop, swinging until it stops on its own. The person can then place their other hand on the shoulder muscles of that arm and feel any tensing or twitching of the muscles as the arm is swinging to a stop. The person can also feel someone else's shoulder as they do the exercise to see what the result feels like there and also have someone else feel their shoulder muscles to likewise externally verify. So long as both people using the tool are getting a similar external result, and one of the people is designated as having understood what relaxation is in the first place, then the idea of relaxation of the shoulder has been transmitted from that one person to the other. The student can then connect the intuitive feeling of his shoulder having been relaxed in using the tool to the general idea of relaxation throughout the body, deepening the feeling and expanding its scope. The tool is discarded afterwards, its only purpose being to transmit the idea of relaxation, and instead relaxation is practiced by just practicing the feeling itself so that the feeling can become pattern-less. That is where you have things like standing in postures, or slow movement, so the feeling can just be practiced in context.
I liked how for everything there were distinct exercises/tools for each stage - some are for transmitting a concept, others are for verifying the expression of the concept (sometimes the same as those for transmitting), and distinct others are for actually training the concept into the body. It didn't try to have all-in-one super exercises like kata or form sets where you were expected to do all of the above, understand, practice, and verify, all at the same time. You weren't trying to reproduce the teacher's skill, you were just trying to understand it for yourself, practice to improve, and then use the tools to verify that you were really improving. When that is systematically applied to all concepts in the training, the entire feel of training is just almost inexplicably different. It doesn't automatically make a better student, but it sure as hell feels empowering in making you feel like you have control over what you are learning.