Happy to oblige. First, I will make what I perceive to be your case (mechanically) to distinguish it.
In order for your "grounding" to occur a load path is established between the point of input and the ground. It is basially creating arching action through your structure. You have specifically illustrated this on your website and posted it here before. On that I have never disagreed with you. It is entirely mechanically correct to do that. I just have cause to know that it does not capture the dynamic of kokyu.
Much of kokyu lies not in establishing that path -- but in preventing it from being established in the first place. Once you "fix" me to the ground with a load path -- it is just a question of how much horizontal thrust is necessary to topple me. That may be effective in a brutal sort of way, but it is certainly not
kokyu. If do not allow you to "fix" me in that condition, then you have a problem.
First, some simplifying assumptions. Disregard uke's elbow. Disregard uke's hips and knees. Assume that they are rigid and not bendable. You'll see why in a moment.
Draw two stick figures each made of two lines one line with the node at the ground and a node at the shoulder with the second line from that to a node at the wrist. Connect them at the wrist. You now have a figure that looks roughly like an "M." From the hand grab, uke pushes me. Pick oneside of the "M" -- label him the attacker. Draw an arrow along the line of his "arm" pointing at the wrist connection. That is his push. Note: I eprcieve a linear push, but remember that he generates it with a rotation of his shoulder (or from his support at the ground, or both).
One way to maintain my stability is to resist directly, that is, put another arrow head-to-head against his arrow of force -- maybe even a larger arrow so as to topple him backward. It might be applied by causing my arm to rotate from the shoulder into and against his push. Problem with this is the large torque involved on my shoulder joint. (Remember we are disregarding the elbow for this simplified construction.)
It might also be applied by fixing my shoulder rigidly and moving my whole body inward at the support node (and upward if we let Mike have his linear springy legs) That also creates a component of force back up uke's arm against his arrow. This is less direct, and some force is not opposed directly, but there is still force on force resistance.
I interpret Mike's "bounce jin" to be in this mode of the mechanics. But it is resistive. Although the forces on the shoulder are lower and more distributed than trying to lever the shoulder to rotate the wrist upward, they are, in fact, still quite high in order to maintaining rigidity. Thus, if I were to resist in this way -- the equivalent of arch action -- I would ground that force to resist the horizontal thrust backwards, precisely as Mike describes.
I have to make my body essentially rigid, as I indicated. No bendy bit (hinge) at the shoulder. Then there would be three hinges, the grab and the two respective points of connection at the ground. Potentially stable. Hinges can open in a variety of configurations, but typically, if four hinges form, an arch fails. This paper in Figure 1 (p.2) shows a four hinge arch failure. http://www.dur.ac.uk/charles.augarde/pubs/c10.pdf
. Problem is, we are full of bendy bits (hinges) to exploit, but the shoulder is the only one I need in order to show this in simple exercise how to destabilize structure without the previously mentioned resistance. You will note that in the "M" simplified diagram I described there are five (5) hinges potentially already operating.
There is another way. Instead of putting your arrow head to head, place that arrow perpendicular to his arrow going down the "arm." He can't stop me because it is perpendicular to his force. I therfore don;lt have to be rigid just be easily able to move the connection around. This causes his arm to tend to rotate from the shoulder, back and downward. If I move my support inward the same amount as I extend his arm perpendicularly, then my condition of vertical support has not changed but his has changed radically. What I have done is to apply a moment to his shoulder, using his outstretched arm to do it.
The problem for him is that he is fixing his shoulder to apply the equivalent moment in the way that allows it to press down and forward at the wrist, which is the same direction of rotation
that I am applying by taking his arm down and back toward him. The result is that he cannot stop his shoulder from popping out and turning into my need hinge, since he is setting it to go the way I am taking it. This means all his force that was going down and forward in the push has been instantly transferred to his shoulder going up and forward, literally lifting his own center with his own push rotation that I have just relocated to his shoulder with that famous "pop" of the center rising.
I have thus "floated" him by manipulating the rotational moments of his body, without any component of resistance.
Now in the more conventional seated kokyu tanden ho, the rotation of my wrist is accomplishing that same dynamic at a smaller scale. His grab at my wrist defines a tangential radius. If I rotate my wrist around that radial connection it is the same as the perpendicular tangential force applied at the wrist of the "M" diagram. The combination of wrist rotation (tenkan) and extension (irimi) results in the same reversal of his applied moment as shown in the more simplified model.
The seated kokyu tanden ho exercise does not remove the support hinges entirely, since you can still lift a hip, but it severly restricts them, forcing you to address the moment manipulaiton at the point of connnection rather than just by moving in or down with the whole body rigidly, and missing the necessary alteration of his structure.
Once we move from the simplified model to the fully articulated body, I have even more options for hinges, adn they are fully three dimensional hinges to work with. The 2D arch thrust line becomes a 3D spiral line. The possibilites become far more complex, but the mechanics of ultimate hinged collapse are the same.
Basically, form primarily controls stability (as opposed to strength.) The substance does not really matter to stability (assuming a minimal material strength.) Any place where the thrust line of the applied load moves to the outside line of the arch (above or below) inside or outside, a hinge forms -- hinging at the point where the thrust line intersects the edge of the structure the thrust line, and opening (in tension) at the perpendicular point opposite that. If I locate that thrust line (ikkyo line) at a potential hinge (joint) that is already trying to rotate the way I want it to go, then it instantly enters a progressive structural collapse and I just have to follow it through (irimi), so that he has no room to recover.
I start with two "gimme" hinges at our respective points of support, I get the third by entering (irimi) his attack a,d Connecting ( musubi) with it, and the fourth by displacing (tenkan) his attacking rotational moment using tangential and perpendicular manipulations. The sum of this in intuitive terms is applied kokyu. I can with more experience control where that thrust line (ikkyo line) is felt in order to form one more hinge somewhere in our joint structure (sometimes I volunteer to be the extra hinge -- think kaitennage or koshinage) altering the existing moments by perpendicular/tangential forces (irimi/tenkan) applied at the point of connection (musubi), in the manner described.
For those that want to dig deeper :
Here is a description of a short program that allows you to visualize the load conditions that lead to hinge formation. http://www.brantacan.co.uk/archsim.htm
and the link to the simple DOS program that allows you to play with it. http://www.brantacan.co.uk/ARCHSIM.EXE
. You can play with the shape of the necessary arch funicular shape (i.e. -- to match the thrust line for a given load under different load conditions) here: http://acg.media.mit.edu/people/simong/statics/
(click on the "hanging cable/arch," and be sure to turn off your pop-upblocker.) If you take that funicular load map and place it on a actual structure you can tell where potential hinges will open.
Both are perfectly safe.
You cannot do aikido with them, by any means, but you will gain a better idea of the concepts that are operating in in these mechanics.