Firstly, hold both arms outstretched in front of you. Relax as much as possible and stay like that for a minute or so. So how "relaxed" do you feel now? Even the most efficient use of the body cannot change the fact that you are having to generate a number of large forces to resist the pull of gravity. If you want to keep everything straight, you can only relax the muscles up to a point- a point where they still have a substantial amount to do. The intention of staying there prevents full relaxation of countless muscles- because this intention is stronger than any conflicting intentions to relax. If you are not willing to have your arms fall down, obviously you cannot release the specific muscles that are stopping them from doing so- without achieving the impossible.
Now shake your hands and arms into release and let them flop to your sides. Although still in a one-ended support system (via the attachment in the shoulder) they have now been brought into a natural balance with gravity, so when you "relax" you can relax into the truest and most literal release of effort. You can hang your arms as freely as you like. Once they are in balance with gravity, relaxation cannot cause any unwanted movement. We all tend to retain at least some muscular activity even in this state. For a healthy person they should be pretty miniscule efforts compared to those that are required to maintain an outstretched arm- although there is much value in learning to perceive and remove these habitual traces of tension too.
When in a one-ended support system, any outstretched position can only be stabilised against gravity by constant muscular contraction around EVERY successive joint (note that this is not a case of "spreading out" the work- ie. in one-sided support no amount of effort in any particular joint can contribute a thing towards stabilising any other joint). When extended, the arms could be held with the least effort physically possible, but every successive joint must still be balanced by successive efforts in localised muscles- so it is tiring. It's like trying to hold a long sword out horizontally, without any help at the other end. Note that this is not a poetic metaphor but a very close comparison. It takes vastly more of an effort to support a horizontally extended structure from just one end (even without the internal joints to balance). Now imagine resting the sword upon something at the other end. The force required from you to keep it balanced will be drastically reduced. This is the key difference between a one-ended support system and a two ended support system.
Now let's think of the playing position and how to achieve a state of two sided balance. We already established that a released arm can slip off the keys. But how much force do the fingers need to provide to create a point of support and at which angle? Well, head down to your cellar and grab the freshest and most intact corpse that is lying about. Tie it to a chair, seated upright. Grab a couple of the fingers and pull them towards you. Can you align the corpse's arm into what would be a suitable playing position ? It really ought to be very easy to do so.
You need to consider this when lining up your own stool- to be sure there is enough angle to ensure that release of the upper arm is providing a big enough component of force that will pull backwards at the elbow. Seeing as your subject is lifeless in the experiment, in this case it is abundantly clear that there is no requirement of any internal muscular activity in sustaining this balanced alignment. The only actively generated force is being provided by your pull. With a living subject, the equivalent balancing force will be provided by a small ongoing grip that occurs internally, within the finger. Also, remember that, unlike the corpse, a living person has the option to employ some more active muscular support at the shoulder- meaning that the size of this balancing force can be made even smaller still!
For those of you who are fresh out of cadavers, simply get hold of a bottle of chloroform. An unconscious subject will do just as well. However, if you're fearful of acquiring a criminal record, just get a friend to shake their arms as loose as possible and then pull their hand towards you (starting from a position where the whole arm hangs straight down). The problem is that most people will be inclined to use their muscles to start lifting. However, if you shake their arm around, to encourage release (whenever you feel them helping) hopefully you will be able to get a good feel for just how small the balancing force really is- compared to what the traditional arm-weight school might have you believe about how "heavy" the released arm is supposed to be.
From this experiment, it should become evident that the downward forces that stem from universal release are not large.While an arm may be a good few kilogrammes in mass, the idea that even a wholly released arm automatically creates some gigantic force against a piano is a pure myth- as this should illustrate to you. The force that results from full release is fairly small (except when momentum is accumulated by dropping from a great height) and largely acts horizontally upon the hand. Even if you were pulling on the end of a big heavy steel chain (tied to a fixed point at the other end) so as to keep it taut, the force applied would still be primarily horizontal. When supporting your friend's arm, you should easily feel that your action is almost entirely towards yourself- and is not based on holding the arm up. A released arm will primarily be pulling away from you, not downwards.
Try getting your friend to pull your own hand out in the same way, and see if you can get a feel for releasing absolutely everything. The average arm should not feel particularly "heavy" at all, but simply enormously free and released. However don't go out of your way to force it to be "light" either! Remember it's very important to avoid any preconceptions about what is supposed to result- as they translate into active yet unnoticed muscular efforts all too easily. Bacon's old truism about how experiments tend to "prove" what they are supposed to is all too true here- so remember simply to strive for release and OBSERVE. Keep shaking the arm loose and then let it truly "hang" between the shoulder and the point of support at the finger. If you can get the whole arm to go as limp as that of a corpse (even in your hand for now), you are getting a feel for the most truly neutral yet balanced starting point- the purest "suspended chain" state that the arm is capable of. Alignment can be difficult to attain when viewed from the exterior, but finding alignment really is spectacularly simple when you view it this way. When the arm hangs between shoulder and hand the alignment can literally create itself. In this experiment you cannot do a thing to make this state occur You simply have to let it occur in response to an external force. When I get onto the finger side I'll show how, in regular legato playing, alignment can still be made to create itself to a very large extent- even when the hand and shoulder are becoming active rather than behaving this passively.
Note that when you pull a loose rope or chain from both ends, it straightens itself- without any remotely precise activity. However, if you push on a loose rope or chain, it can go all over the place with very little predictability. This same principle applies here- remember that the arm is literally a chain of joints. This is no vague metaphor. Actions that are based on pulling a chain of joints create order and lead to predictable and easily repeated results, whereas actions that are based on pushing create greater complexity and make the results vastly less predictable (unless links in the chain are fixed into position). Of course, sometimes great pianists will break with standard alignment and use variants for good reason. However, the baseline for regular alignment is spectacularly simple when based upon allowing a released arm to hang itself into place between shoulder and finger. This is what I mean when I say that great pianists don't just make it look easy. A pianist who tries to create alignment from the outside alone is simply attempting something that is far harder and far more complex than what an experienced virtuoso really does (especially if they are trying to push with the arm, throughout passages of regular finger work!). A true virtuoso feels a few very simple balancing actions on the inside and the majority of alignment can ensue as if by magic- because that is the state in which gravity balances the middle of the chain. I should stress that I am not ruling out some fine tuning with the muscles beyond this- but it's far easier to release into the starting point of a neutrally suspended arm than to try to start by trying to manually crank every individual area into a very specific place.
In part two I will go into some exercises that are based around starting to employ some slight balancing forces at the shoulder, without resort to stiffening. In practice, the shoulder needs to be able to release/support as much or as little of the arm's weight as you desire, in order for the fingers to perform a variety of different actions, while maintaining balance. However, if you go about supporting the weight in the wrong way, the fingers end up with nothing to pull against and the whole arm starts having to stiffen to stabilise their actions (and likely having to push through them to compensate). This is also where the third axis of sideways motions comes into it. It is certainly not desirable to let the shoulders slump inwards against the body, so it's also important to learn how to take support at the shoulder- provided that you are still releasing enough to keep the chain of the arm hanging taut, rather than being notably held up around the elbow and wrist! Anyway, more exercises (to start on the active efforts that need to be added to this initial state of total release) will come shortly.