Collapsing fingers- a simple illustration of why they occur, why they are harmful and how anyone can quickly learn to prevent them (via the action of extension)

(Nb. I wanted to get this issue dealt with in order to avoid the need for any digressions about how to avoid the problems of collapsing joints, within my next post in the series about core piano technique. For anyone who sometimes suffers collapsed joints, this is an essential precursor to the upcoming post. However, I am deliberately ensuring that it can equally be understood as a standalone post about this important issue, without any need for external context) 

Collapsing fingers are a classic pianist's problem. Particularly with the second finger, it's extremely common to see the joint nearest the fingertip abruptly give way and double back on itself, to form this type of shape:

(many thanks to Azim Akberali for kindly producing all diagrams)

All too commonly, this is put down as being an issue of strength (particularly in young children)- a defeatist attitude that suggests that the answer is simply to ignore it and merely hope that matters will change of their own accord, given time. However, following some recent experiments with a number of students, I've become quite certain that the real issue lies in the coordination of muscular activity and that it's almost entirely about the basic manner in which movement is conceived. Before I go into any further detail about hows or whys, try this potentially very revealing exercise (please read the descriptions before following the video examples):

Firstly, ground a key gently with a lengthened thumb, in order to help the arm gain a touch of support (without either feeling in any way clamped or forced down against the point of contact). Just once or twice, draw the fingertip lightly back towards your palm in mid air (these are the most sensitive joints of all so I have to stress the importance of being gentle- rather than truly "gripping" against perceived resistance). Here, your joints are closing inward. Then attempt to do the same, when sliding the fingertip lightly and easily across the surface of a depressed key, in towards yourself. Do this with absolutely no sense of digging down through the finger with arm pressure. If you feel the key starts blocking your path, lighten up your arm and make room to continue more gently, without getting squashed down!. Note that the knuckles are given room to drift up and become the highest point- which is important for avoiding any sense of burden. Definitely don't do this with the knuckles squashed down!

So, did the tip double back on itself or did it successfully curl in on itself as it did in thin air? Ideally, line up a mirror beside the piano so you can observe yourself properly. Recently, I saw no less than four students who found this task almost impossible, even when being extremely gentle indeed. Instead of curling inward, on each occasion that final joint would abruptly give way and fold considerably back on itself. Logically, this might seem to suggest muscular weakness- given that they were actively attempting to close the finger tip, yet were incapable of doing so? Well, it might seem that way, but now try example two. 

Here, the finger starts curled and the same two joints perform the opposite action of opening out forwards. Again do it in thin air just the once or twice and then experiment with how it feels to gently slide along the surface of a depressed key- initially starting nail down (but again with a sense of a high knuckle), in order to feel the whole range of movement. Of all these four students (who doubled back severely when they tried to grip their finger tips inwards), not a single one experienced that same sudden collapse! They could easily go from a curled position all the way to a very open one (where just a slight curve of the fingertip was still present) without any sudden jerks or any loss of stability. Within a very wide range of angles, there was simply no danger of collapse at all, any more- as long as the student was allowing the finger to lengthen out freely and easily, without the arm jamming it down hard against the point of contact with the key. Now, I don't want to say this will definitely apply to EVERY case of collapse. However, I'd stress that each one of the students who I speak of here had highly flexible joints that had formerly been so "weak" as to instantly double back into positions so extreme that I'd literally have to break my own bones to even begin to recreate the shape. As soon as they replaced the concept of gripping with that of lengthening, their hands started to find positions that looked as effortlessly "strong" as any hand.

 

This could easily seem plain bizarre- and (in spite of having previously thought about these issues in depth) I was actually rather surprised to see these formerly "weak" fingers consistently becoming so capable at once. Why on earth would it be that trying to perform an action that usually closes each individual joint of the finger would uncontrollably collapse the tip in the very opposite direction? Why would the action that lengthens the finger (ie which moves the fingertip away from a rounded position and in the direction of the position seen in collapse) be readily possible without risk of triggering that extra step into collapse itself? The whole thing seems to be completely the opposite of what common sense might dictate, but actually there's a very simple mechanical explanation (simple enough that it can be given purely in common sense terms and without any need for mechanical jargon). 

Obviously, when we perform the curling action in thin air, the finger closes. The problem is that it's not so simple when we're interacting with a piano key (and the more the arm is pressing down through the finger, the less simple it becomes). Consider, firstly, that the joint which acts to close in the finger tip cannot typically be moved separately from the action that causes closure at the finger's middle joint (except by a tiny percentage of people).

The impossible movement of the tip seen there could only be effective at preventing collapse IF it were possible to execute it without the finger's middle joint also getting involved. As you'll almost certainly see if you attempt what is seen in the diagram, however, the middle joint of your finger will begin to curl up too. Consider now what happens if the arm presses down while this is going on. The fingertip is now effectively fixed against a certain point on the key. That means that if the mid joint starts closing up, it is dragging the final joint backwards and towards collapse. If we depict the finger tip and its middle segment as isolated levers it should be completely clear, from these before and after pictures, that any pull from the upper lever is dragging the joint towards our classic collapsed finger position.

Let's translate that back to the fingers themselves  in these diagrams:

and it should now be abundantly clear why the action that might usually close up the finger (away from the piano) will instead be trying to drag the fingertip into collapse. The middle joint and fingertip joints will function just the same as those two levers had. Once the middle joint has started pulling you towards collapse, the more you try to grip from the weak joint that controls the fingertip (in a bid to keep it rounded) the more you ALSO engage the stronger middle joint that is destabilising it. This creates a vicious circle where the harder you should to seek to oppose the collapse, the more actively you contribute to causing it- by working directly against yourself.

Consider now the lengthening action, seen here in a before and after:

(Note that this action still involves the act of closure or 'grip' at the knuckle joint. The difference in what I am speaking about exists exclusively in whether the middle and fingertip joints complement that by opening or by closing. In standard human grip, all joints close in. In this action, the two lesser finger joints instead open out, while the knuckle alone goes into a more closed position. If that sounds complex to you, simply imagine the finger lengthening to match the blue line, or see this post on the finger actions for more information on how to develop a "feel" for this.)

When you are familiar not merely with the hand's most standard gripping action, but also with this complementary lengthening action, it quickly becomes possible to learn to balance on a depressed key without fighting against yourself. By alternating between the action that slips towards you and that which slips forwards, you start to discover a direction of action that exists somewhere between the two. The line of force is actually a touch forward (rather than perfectly straight down), yet not so much as to make you slip or to cause any feeling of instability. Sometimes it's as simple as getting the student to experience both forwards and backwards slides first and then simply allowing them to discover what feels right, without too much rationalising. However, if a student's habits are strongly based around desire to move keys by pulling the fingertips backwards, it may take some time before this rather different conception of movement becomes fully habitual. If so, I advise spending some trying to consciously perceive the slight sense of aiming finger activity forwards, even whilst keeping a key depressed. The feeling is not that you're not actually going to slip, but that you're ALMOST doing enough to cause the finger to start to sliding slowly but steadily forwards. Given time, this slight forward action will start to feel normal and instinctive and the student will stop accidentally retracting the tip in a way that causes collapse. In the end it may still feel like the finger is acting forwards or it may simply feel come to feel "right", in a more abstract and non-analytical sense. Whatever the case, it's worth coming back to explore both directions of the exercise, from time to time (and for each finger- not just the 2nd, that I used to illustrate). Willingness to slide smoothly but freely in either direction gives a wide sphere of experience to your senses and allows various muscles to let go and stop trying to force stability. When you go back to a regular balance point, it often turns out that you can be more than stable enough without actually bringing back those efforts. 

Addendum

What I've written so far will hopefully have both given a rather comprehensive summary of the objective background to collapse and provided a means of conquering it that is both simple and direct. However, I'd also like to follow up with a few additional points regarding some of the issues that have arisen here. One part of this is extremely significant for pianists with small hands. But, firstly, even if you don't have any collapse issues, consider this- are you certain that your middle joint isn't placing a burden on your last joint, by needlessly pulling back at it while you play? Just because a finger doesn't collapse, it doesn't mean that an inner battle isn't going on between conflicting actions of the two joints- if you happen to employ the type of gripping action described.

With that in mind, the exercise is equally worthwhile for those who never experience collapse but who may be hooking themselves too forcefully into the key. If the finger is typically gripping back in towards you during regular playing, the arm tends to have to press down for the sake of stability. This can really burden the tendons and I strongly suspect that a conception of movement that is exclusively based on trying to grip inwards at the tip can contribute directly towards tendinitis.When you add a capacity to lengthen these weaker joints to the more common desire to grip from them, there's vastly less physical effort. Many people speak of curled fingers as being a cause of tendinitis, but I think this is a very misleading simplification. Curled fingers that are acting to curl up further will become burdened. HOWEVER, curled fingers that are gently lengthening in the last two joints (ie uncurling outwards) need experience no such burden. You have to consider both the position and the type of activity that is in progress, to get to the bottom of these issues. Neither gives the whole picture, without consideration of the other.

Consider what happens if you reach this type of open position (as is commonly seen by the likes of Rubinstein, Richter, Gilels and many other great artists- none of whom ever experienced any physical problems that I ever heard about)

Any use of inward grip from those two secondary joints will cause the fingertip to be dragged virtually horizontally, as shown by the red arrow. Unless the arm is forcefully digging down to create friction, that will cause phenomenal instability when the finger moves and compensatory tensions. From such a position, it is far healthier to appreciate the action that lengthens out the very same joints. This completely eliminates any need to hook in via strong arm pressure. Remember that both actions involve the stronger closing action of the knuckle itself. It's just that one also involves a significant burden on the weaker joints, whereas the one that involves lengthening them out makes for a vastly lighter workload (plus a more productive line of motion).

One point of interest to consider is that many pianists and teachers still actively preach the value of drawing the fingertip in towards the palm. While I'm on this topic, I'd like to illustrate how I'm personally 100% convinced that the efficacy of such advice actually depends on the student taking this as a springboard to something altogether different. This paper speaks about fingertip grip in relation to a pianist's health.

 http://www.legacyweb.rcm.ac.uk/cache/fl0026842.pdf

Also, there is an Irish pianist called Roy Holmes who used to have various Youtube videos that claimed that all finger technique should involve aiming to draw the fingertips back towards the palm (I don't know if the fact that they have been withdrawn might suggest that he's since had a rethink). Firstly, if taken precisely as detailed, this approach would be quite functionally impossible for anyone whose fingers collapse easily. Rather than being healthy, it would frequently encourage the joints to slip out into their extremes of motion, via jolting collapses. Some hands just will not suit the approach, due to their individual characteristics. Even for someone whose joints do not collapse, it could still be hard work, for the reasons already detailed. So how could such a thing be promoted in the name of "health"? 

Consider this- if you pull back into a position that is still curved, what happens next? Do you really keep gripping from the fingertip while keeping that key down? If you did, you'd have to hook in for stability against slippage, as illustrated by the last diagram (and the line of attack would also tend to drag the wrist forward and trap it into an uncomfortably bunched up position). I don't believe this could possibly be wise as a baseline for technique. BUT- consider this: what is the most natural thing to do once you've found a curled position? It's to switch to the lengthening activity, for ease of balance. Perverse as it may sound, I believe that approaches which are founded on preaching the importance of gripping from the fingertips work specifically because they create a position where you may start to feel that it's unnatural to grip any longer (ie. the advice actually describes an action that is potentially dubious, yet which prompts you to move into a position from which you may instinctively begin to a get a feel for the totally different action that suits the situation!).

If you can do it without collapse, try sliding your fingers along the keys like in exercise 1 again. After having had the experience of working at both drawing the tips backward and lengthening out forwards, can you feel an instinct that makes you want switch over to the lengthening activity, for balance- in the instant that you decide to stop sliding? If not, try it even more gently against a table top. If you don't perceive it on the piano, it should at least be possible to detect this change on the table top. At this point, either you could hook in by pressing the arm and work pointlessly hard or you can stop gripping from those joints and instead keep the balance via a much more subtle lengthening action (that keeps the key down and preserves the shape of the finger). This is perhaps what some people mean by "and then relax"- which might be more accurately translated to- "and then switch to a different style of finger activity, in order to find easy balance" That may sound wordy by comparison, but not every one will get this on instinct, merely as a result of indirect (or arguably even "contradictory") instructions. I believe that any deeper benefit of the "carezzando" idea is dependent on whether fingertip stroking triggers a position in which literally the opposite activity has to kick in, as part of a major muscular reorganisation. If you're equally used to finger lengthening actions and willing to go on instinct, you may well find a simple and easy balance without thought. However, if you get lost in an artificial premise that your fingertips are "supposed to" grip and devote all attention to doing so, you may very well place an unpleasant and unhealthy burden on your fingers- especially if you attempt to execute rapid-fire semiquavers with these highly indirect scraping actions. The more successfully you should literally adhere to the advice, the more uncomfortable you would likely find yourself- unless you should begin to perceive a change-over, by which the activity switches to the alternative manner of finger activity.

Although I am totally convinced that actions that involve extension are literally essential to the possibility of advanced pianism (and suspect that you could theoretically flourish with virtually no deployment of active fingertip grip), by looking at it from both angles you can learn a great deal. As I said, allowing slides teaches you how to let go of some unwanted efforts. In that respect, the "carrezando" idea can likely have benefits- (as long as you genuinely do it lightly and without "scraping" hard across the keys or "hooking" forcefully into them). What I am not particularly keen on, however, is the danger of implanting a person's mind with the misapprehension that fingertip grip might be one of THE standard ingredients behind pianistic actions, without drawing any equal attention to the important alternative. Unless a person is lucky enough to stumble on a "feel" for the fact that balance is vastly easier via an opposite activity, they may get themselves into all of the problems that are often associated with "curled" fingers- by using both a curled position and further curling activity. Grip is second nature to most humans already- which is probably one of the reasons why children so frequently collapse their fingers by gripping from the fingertip. In most of life, that's our default hand-action. What we are less inclined to have any immediate feel for is the vital role which lengthening actions can play.

On the subject of health, I want to illustrate one further issue, that demonstrates a powerful reason to appreciate why the lengthening action is extraordinarily important for small hands (and which further illustrates why it's a very bad idea to approach pianism solely in terms of the gripping actions, without reference to opening of joints). Lengthen your fingers out as shown before and then practise gripping lightly inward in all joints. Watch your knuckles this time. Do you see the fingers closing in towards their neighbouring fingers when you grip inward in all joints? Do you see how the space opens when you combine the slight lengthening of joints with a knuckle grasp? Here's a demonstration of both (followed by a demonstration of the fingers purely being drawn inwards and outwards, independently of other motions. This is a great exercise for freeing up any long term tensions that may be lurking there).

If you fancy trying to keep that space between adjacent fingers open while gripping, do you feel the sheer level of conflict inside your hand? Muscles are fighting to contradict the action of your very own muscles. All it takes it to add some degree of lengthening in the lesser joints, and you can eradicate this senseless fight (which is exactly what is going on when most pianists strain to reach big chords).

Ask yourself, even if you have a big hand, do you really want to be limited to actions in which you are actively making your hand smaller with every movement you make? Could that honestly be healthy, if taken as a regular pianistic action to default to? Or does it make overwhelmingly more sense to depend chiefly on an action that actively opens your hand out and frees you while you use it? I don't think it's any coincidence that my reach between 2 and 5 has increased from an awkward 7th to an easy octave since I willfully introduced lengthening actions to my technique.  I'm certainly not looking to try to ban actions involving fingertip grip from pianism outright. However, given the compelling evidence in favour of actions that instead utilise extension (which stacks up way beyond the particular collapse issue that this post started out on), this alternative strikes me as the single most worthy action of any studying pianist's conscious attention. I feel certain that those pianists with smaller hands who flourish (unlike the unfortunate many whose fight to reach chords is all too easy to detect in their sound) do so specifically because their technique regularly involves the lengthening action. In my opinion, pianists whose attention is drawn exclusively to the actions of closure are more likely to be hampered by that advice than to stumble on the alternative- and overwhelmingly more likely to suffer injury via the internal battles. When pianism is approached in terms of both qualities of movement, students are vastly more likely to find their own feel for what works.

Achieving effortless balance within piano technique- connecting your whole arm to the keyboard, without burdening the hand (part i)

(Since first publishing, videos have now been added, plus a couple of additional steps for achieving freedom in the arm)

In what will now be a two-part second section to my core posts (I recommend reading this first section on the fingers first), I'll be dealing with various issues- including how to feel the internal workings that determine whether a particular hand position will be effective (involving reference to the different positives of both "firm" and "relaxed" fingers). In particular, I'll be looking at how to achieve simple and low effort balance (of both finger and arm) after sound has occurred- while a key remains depressed at the keybed. However, in part two I'll also be spending a little time on a means of feeling how to ready a finger for easy key movement- setting the scene for the final core section on actually sounding the keys efficiently and with predictable results in sound. As with my last post, I'm going to keep the scientific background to a bare minimum and make this almost entirely practically oriented. However, this first half will end with optional "behind the scenes" information that further illustrates why I am asking you to perform these exercises and shows what they can objectively help you to achieve. If you have a naturally enquiring or skeptical mind, the addendum should make sense of various questions/potential objections that might arise for you along the way (in fact, if you're mid-way through and find yourself thinking "Why the hell do you want me to put my hand in that ridiculous position? That's not how I was taught to play the piano!", you may want to START with the addendum- where I'll both show why this is the easiest introduction to judging quality of balance in ANY position, but also prove via videos that great artists regularly used exactly this style of position in a literal fashion!). The proof is in the pudding and I hope the exercise alone is adequate to get a feel for what efficient balance is like- but I would still strongly urge you to consider reading the background issues in the addendum, for the fullest awareness of what is going on. It will explain many specific reasons for instructions that might otherwise seem a little counterintuitive- and I assure you the background reasoning is not at all complex or advanced!

Anyway, I'll start with some brief practical background issues right now- about why this particular issue is so important to piano technique (and why even highly advanced pianists can benefit from going back to these simple foundations). It will doubtless seem odd to start putting things into practice with reference to what goes on AFTER a key has been depressed. "Shouldn't you start with moving the key?" you might ask. However, the state of your hand (and indeed of your whole arm) before you start depressing a key is of paramount importance. You may not be able to influence the sound of an ongoing note after you have struck it, but what you definitely can do is to ensure that you are optimally poised to control the note that is to follow. This is extremely important if you want to develop the ability to precisely control the intensity of notes that come in rapid succession. The quality of balance that ensues after a note has been played should be regarded as the FIRST element that contributes to an ability to exercise control over the sound of the NEXT note- not as a mere footnote to a sound that has already occurred!!! Tensions are frequently a matter of fighting to remain stable due  to not having found clear balance on the last key. Before you can expect to evolve into good poise merely as an automatic result of depressing a key, you first have to be intimately acquainted with that poised state.

Would an athlete be happy starting a 100m sprint having slumped casually against the floor- in the vain hope that they might spring to life in the instant that they hear the starter's gun? No. If they are not properly prepared, they cannot suddenly recover from a disadvantaged position. Even Usain Bolt couldn't win if  he began from a posture where he was teetering on one leg whilst swirling his arms around (like a cartoon character trying not to fall off a cliff). It would be no better if he only avoided toppling like a drunkard by tensing up every muscle in his body. He needs to be primed for action against the starter blocks, in a quality of balance that cannot be effectively summarised by such overly simplistic terms as "tense" or "relaxed". A good pianistic hand typically needs to be priming itself in much the same way. To exploit your pianistic "starter blocks" properly,  you need to be looking both forwards AND backwards- not waiting for the exact moment where sound is due, before acting. In most standard playing, we have a situation where the quality of poise against whichever notes were struck last is an inseparable part of the readiness for the notes which are to be sounded next (which will be demonstrated with specifics, in future). In that respect, I'm actually working in the natural order of events by covering basic balance issues first (in reference to one of the basic finger actions that has been detailed in the first core post).

Anyway, in this first half, I primarily want to tie together the concept of the "floating arm" (ie. where weight is supported at the shoulder end of the chain) with "arm weight" ideals (where the weight rests down on the keyboard via the hand). Those who advocate the floating arm (notably Alan Fraser) are trying to ensure that the fingers are free to move without being burdened by a heavy load from above. Those who advocate resting the weight of the arm on the fingers are trying to ensure that the arm can feel a sense of being supported- rather than become trapped into a position where it is being stiffly held up against gravity.

The thing is, you don't have to choose between the two positives, or get locked into a one-sided view. With the following exercises, I want to show how easily you can reap the fullest benefits from both models at once. After playing a key, the arm should be able to feel as much comfort as when you rest it casually on a tabletop. HOWEVER- what this doesn't mean is that the hand should therefore feel as if it's bearing the full weight of your arm, against the keybed! As far as the hand is concerned, it can still feel as effortless as if the arm is just floating behind it, without any sense of the hand being forced to prop a significant load up against gravity. If you think that sounds like a deranged paradox (rather than anything that is at all realistic)  let's get into the exercises, for a rapid illustration. I strongly advise lining up a mirror by your piano for observation purposes.

Exercises

(Note that many of these involve arm movement, whereas others are primarily about settling into a more stationary position. Be careful not to overdo the stationary positions. There's a way of involving miniscule traces of ongoing arm movements (that are virtually invisible to the eye) even when "still", that helps safeguard against tightening. It's good to gradually reduce the movements of the arm and then still imagine you are almost doing them still, as you settle into balance. I'll detail more about these issues in the next post but, for now, notice which exercises bring most freedom to you, while keeping the hand and arm actively connected up. If you feel something tightening rather than loosening at any time, switch to whichever exercise loosens you most, via arm motion. If that doesn't work, switch to the other arm or just have a break and come back later. If you fight on through any tiredness, you're literally achieving the very opposite of what is intended from these!).

Start by extending out a thumb and finger as shown (there's no need to squeeze or grip inwards- just lengthen out!) and hover a cm or so above the keys (example 1)

Try to stay like that for 10 seconds or so (nb. I gave a shorter illustration, as there's not much of a spectacle on display when it comes to viewing this one). Before long, you will almost certainly start to feel tired in many areas of your arm. As explained in this post, one-ended balance requires an assortment of supporting muscular efforts around EVERY joint- or gravity would create imbalance and therefore movement. Your shoulder has to bear the arm's whole weight, without action from any other joint being able to reduce the load. Further joints simply have their own additional workloads to perform. Unless you experience severe discomfort, try it for another 20 seconds or so- but, this time, can you distinguish between the muscular activities that are genuinely 100% necessary for balance and those that serve no function at all? Try to relax anything that is not needed- so just a fraction more release would be enough to allow gravity to cause motion. Try to acquire real awareness of the activities that are truly essential- especially in the shoulder, elbow and wrist.

Now (after shaking your arm loose, perhaps) rest your thumb and second finger together on a single depressed key and compare (example 2). Don't push down, but try resting a released arm passively and limply on top of the extended thumb and finger (in the unlikely even that this causes any pain then please don't force anything!). Hang back though- so anything that might be bunched forward gets lengthened out by release of efforts. Can you feel the specific differences in effort? The benefits of the arm weight approach should already be clear. Because you now are in good contact against the keyboard, the burden has been taken off various muscles in the arm that previously had no choice but to engage. There is now support from two ends. While this does not guarantee that any superfluous efforts will all be released, it is at least possible to remove the overwhelming majority of prior activities. Previously they had been required, not to drop. Now it is possible to strip down to a low level of muscular activity without falling down. This mere possibility tends to inspire plenty of muscles to start letting go- but you should take a little time to look for any other needless efforts that your instincts might have missed.

Okay, that's the obvious plus side of arm weight ideals. So the arm weight folks are obviously correct and those who want to bear the arm's weight at the shoulder are obviously in the wrong, yes? Well, if I wanted to produce propaganda for extremist arm weight ideals I could stop here and leave a powerful one-sided impression (I hope that traditional arm weight advocates will not get their hands on this example and abuse it out of context!). There could hardly be a more persuasive argument for the benefits of resting arm-weight on the keys and not bearing it at the other end, right? Well, in a way. However, this is absolutely not the whole story! It's very easy for what might seem to be airtight logic to be founded upon incomplete data- rendering the "obvious" conclusion invalid. What happens if we reduce the amount of weight that rests on the thumb and finger now? Since when was arm weight limited to a simple on or off switch? Does anyone seriously think the body is that coarse and simplistic in how it functions? As long as you were only resting and not pressing down, you should have felt your hand was far from in danger of injuring itself from the experience. However, what if you could find a way of using even less effort and a greater sense of freedom still?

So, this time aim for a rather smaller pressure against the key bed, while still keeping clear contact with it- ie don't release enough to allow the key to start rising or even wobbling). Firstly, rock your arm lightly from side to side, sometimes fairly quick and sometimes slowly (example 3).

 Look for a sense of release- but don't release that mild lengthening activity in the fingers or let them droop! Now try going out to the side (and mildly forwards) with the elbow in a big circular path. Then let it slowly come in all the way back to your body. Note the sense of length in the wrist (example 4). Keep changing direction, but slowly reduce the size of the movement- until you feel you have found the place where your arm feels just right. It should neither be slumped in against the side, nor significantly "held" up (with the elbow still poking massively out) via large efforts in the shoulder. You have to find the position that is right for you, but it should be more open in the armpit, compared to when you truly relax the arm all the way back in to your side (can you feel how you end up working harder if when you let it slump against your side as dead weight?). When you get it right, the arm hangs itself between the shoulder and finger. Everything  in between should feel as if it's supported automatically- without even a trace of tightening to stop gravity collapsing it. It's important to do this very slowly- rather than in a hurry. I just demonstrated roughly what it might look like to give you the idea- but you need to try to perceive yourself with acute awareness, in order to strip away pointless efforts without losing the useful activities (which will need to remain present in both hand and upper arm). Spend plenty of time searching, but if you start to get tired then switch arm or take a break. This post features a very similar exercise (given in more detail, away from the keyboard). It's not compulsory, necessarily, but it offers a good preparation for sensitivity to the weight-regulation mechanism. However, if you find your upper arm getting tired or feeling stiff, you should probably take the more detailed version of the exercise as a compulsory preparatory exercise, to improve the basic action.

Anyway, how much lighter are you now? Can you feel the springs of the key pushing back up and trying to collapse your finger and thumb? Can you feel that the action of extending them out does just enough to balance a small but clear pressure that is trying to return the key? If not, there's way more force than you need and you're probably just plain tense at the moment, even. This is the worst kind of "firmness"- which is probably more accurately viewed as general "stiffness" that exists in balance with nothing in particular. Aim to feel the key pushing back at you with total clarity of awareness. Balance is what happens when the load on one side of a see-saw is sensitively matched to that which can be observed acting on the other side. If you encase the pivot in concrete (immobilising the seesaw outright) that is not an act of creating "balance"! Neither is "firmness" against a force that you cannot even perceive. Ironically, when people don't realise that the fingers need to maintain activity in keeping a key depressed, it's usually because muscles work so coarsely and excessively that they have lost awareness of these big efforts- not because the level of activity being performed is insignificantly small! Tightness readily interferes with the brains ability to register what the muscles are actually doing (especially if you've done it so much that a long-term habit has started to evolve).

As a teacher, I can tell you that it's amazing how stiff a student can sometimes be while believing that their hand or arm is "relaxed". That's not their fault though- everyone has different blind spots in their self-perception. The problem with an impression of comfort is that it is relative to what you typically experience and not necessarily a reliable gauge of what you are really doing. However, the better you can perceive what purpose the action of fingers should be serving (ie precisely cancelling out the force that is trying to send the key back up/collapse the knuckles) the easier it is to strip the muscular activities down to those which are actually useful to the task- rather than counterproductive. If you can feel  what you need to match your actions to, you will get a lot better at saving effort. Sometimes let yourself get collapsed back a little by the key (in order to get an internal measure how strong the force really is) and then see exactly how much activity you really need to regain quality of contact and ground the key again (example 5).

Don't hurry this (as if you're simply ticking off boxes and then casually pressing on to the next instruction). Keep looking to perceive the details more and more clearly (but be very careful not to clench the arm on this one- can you feel it going from something like that initial state in example 1, back to the renewed looseness that can be found after regained contact?). Other times, rest down a touch extra and then gently allow the knuckle to droop down, before lengthening back up and away (example 6). Depending on how weighted or light your arm is, relaxing the fingers could either allow the key to go up or it could allow the knuckle to drop down. Which is most normal when you relax? Deliberately try the other one too, regardless of which side you are on now- so you're not getting lost in a single extreme. If you're sensitive to how you balance your arm, you should be able to experience either. Anyway, can you yet perceive the sheer extent of the role that the extension action must play in keeping the key down, when the level of weight reduces? By seeking to perceive this with sensitivity, rather than employing aimless strength of muscle contraction, you are learning precise coordination that keeps the arm at ease (rather than held up as in the first example). In a sense the fingers are quite "relaxed" (in the sense that that they are very comfortable indeed), but in another respect they contact clearly enough to be "firm"- just not as severely as when more weight was bearing down.

(Ultimately, descriptive words are a poor summary, anyway- compared to what you can learn by starting to perceive the FEELING of the exact interaction between the force that a depressed key sends up at your finger and the particular intensity of lengthening that most readily balances this out. You cannot convey that abstract feeling through vague subjective language in a particularly meaningful way- as it will typically mean something totally different to another person. However, what words can do is show a person how to zone in on balances of simple activities and find their own internal feelings of what it is like to explore them with a sense of purpose.)   

Also, don't be tempted to pinch inwards! This can destroy your sensitivity, by forcing muscles to work in direct opposition to each other. Focus the activity into the twin goals of keeping the key comfortably down (without excess force) and helping your arm to be at ease. There's no need to fight against yourself! Rest the thumb and finger lightly together but always strive to perceive that the lengthening action is your specific means of sensitively balancing the key. The fingers should be mildly firmed by the interaction with the piano (and definitely not full of slack) but they should be in no way stiffened or clenched together- and certainly not compressed by the arm. If you catch yourself pinching together, can you feel how mere lengthening out can replace that need to actively grip inwards? A good test is to sometimes start together but then to open out slightly (example 7) between finger and thumb

(nb. if you are prone to collapsing joints then hopefully keeping it to literally a millimetre or two of space should safeguard against that. If not, just keep the 2nd very gently propped up by the thumb without tightness, to make collapse impossible. I'll deal properly with reversible joint issues in part ii).

Are they still as stable as they were when touching? If you are doing the extending action well, you should feel as secure either way. Although squeezing in can make it very easy to experience extreme stability, learning to exploit extension effectively will ultimately offer all the same benefits for less physical effort. Try comparing a deliberate bunching together with the lengthening. Can you perceive the slight tightening in the wrist with pinching, compared to the true freedom that can be found by matching the lengthening activity to the force that the key sends up at your finger?

Can you feel a clear connection between your fingers and the piano now, but with drastically less stress on the hand? Does your arm still hang loosely back from that contact, though? If not, can you find a way to make it do so? Did you relax the fingers a little too much now and lose the quality of contact? Be sure that they are actively lengthening, to push the knuckle right up. If it droops even a little, the contact may not be good enough to hang freely back from. Pushing gently on the knuckle with the other hand will confirm whether you are actively opening it by lengthening out (example 8). Remember though, this is to test that the knuckle is being pushed UP and away! It is not to encourage extra downward pressure from the hand that is on the keys, but rather the very opposite! If you pass the test (by not collapsing) that's good in one respect, but maybe you are actually too firm still? Can you now make the pressure against the knuckle from your other hand smaller still- yet feel as if you're only just lengthening enough to continue to support it? Sometimes almost be on the verge of giving way- so you pass the test, but only keep the knuckle up with the very least effort possible. There's no single "correct" balance. There's actually a good argument for sometimes very marginally doing a little extra activation of the fingers on purpose, but be aware that there's usually much more sensitivity if you're still rather close to the exact balance point, rather than making yourself totally solid. Even the test described here is better viewed as a test of coordination and sensitivity- rather than as a challenge as to whether you can grit your teeth and survive an onslaught of brute force. When done insensitively, pressing down on your hand can make you stiffer and less perceptive. If you've gone about it the right way, you should go onto feel an improved quality of connection to the keyboard with LESS effort- not more! As a counterpart to this exercise, sometimes place a finger underneath the knuckle and check if there's room to raise it up further without losing contact with the keys (example 9). If there is, you were drooping and were not extending properly! In this case, you definitely underdid the lengthening action. Fix the position with your "teacher" hand. But did you lose quality of contact between the fingers and the keys now? Did you lighten too much and leave your arm "held" rather than hanging? Don't worry if you do slightly miss the right balance at first. Over time, alternating between these two exercises is a powerful way of converging on a level of activity that is neither too little nor too much, but matched to requirement.

Is the arm yet feeling truly lengthened or is anything bunched forward that need not be? Try slowly going forwards- so the wrist goes up and above. Here you have to be especially careful to maintain length in the thumb and finger! Any of these added arm movements can either be coupled with a mild pressure against the knuckle from the other hand, or the slight upward assistance (as in examples 8 and 9) The latter is a very good idea here, as it's very easy indeed to droop and lose connection! Go slowly and try to perceive the adaptations required to keep the sense of length in the fingers as you enter different positions with the arm. Once forward, the whole arm eases back (starting from the shoulder end) to pull the wrist straight within a longer line- but it should always stay free and movable (example 10).

Also don't pull back so hard that you'd need to try to "hook" down into the piano key, to stop the point of contact sliding back. Any pull is merely to create length in the arm. To make sure you are not digging in, sometimes deliberately slide the contact against the key forwards and backwards (example 11). There should be clear contact but minimal sense of friction- definitely no bumping or grinding as you travel, but steady and simple motion. Again, watch that the knuckles don't start to droop but feel you do just enough to be lengthened in the thumb and finger. The more you go slowly back and forth, the easier it should be to get the arm to hang loosely between the key and the shoulder- without any areas bunched up (except the knuckle) or held in place by tensions around joints.

So, after all these variants, do you actually have to rest notable weight on the piano to get a good level of all round comfort? Hopefully, you should already be starting to see that the effort involved in a "floating" arm is a world apart from the major efforts associated with the arm that had been literally supported in space from the shoulder end alone. It should be far closer to the ease of our "resting" arm, even though we're not going out of the way to rest weight down any more. That is- as long as we have found a top quality connection to the keys (that can keep the knuckles elevated) and a sense of length and freedom in the arm. Miss either of these and you will struggle to achieve even a fraction of the comfort that is potentially available to you. Get it spot on, and you'll come to find that far more ease is afforded to the forearm and wrist, in particular, when you are not actively trying to rest down upon the hand. See the major difference between these two diagrams (the first showing an arm that is held up in a forward position and the second showing a well aligned arm- after the bunched up wrist has been pulled back into length):

(All diagrams generously provided by Azim Akberali)

Check these issues regularly in the mirror to be sure. Try getting it wrong now by holding the wrist a little higher and forwards and staying there for a while. Can you feel how the arm no longer balances with the same ease? Slowly release and feel how the upper arm will drag it back (in the general path shown by the arrow on the diagrams). Is it now even clearer how much effort you are saving when the wrist is lengthened out, compared to when it is a little bunched up? If you're stuck forward and over then no matter how well the fingers connect, the arm will be working harder to keep you up- until you actually let gravity drag it back into alignment (after which you can strive to lighten without falling into the trap of holding anything forwards, this time). It's no use connecting the hand to the keys if the arm is not taking the opportunity to lengthen back from the contact that you create.

Sometimes let the fingers relax as much as possible so they droop down into collapse. Can you feel how the arm has to work harder to stop its weight from squashing the whole palm down into a cluster (example 12)? I didn't illustrate it on the videos, but try a high wrist and drooping knuckles all at the same time. Can you perceive quite how much extra effort is required in general if you fail to focus enough intent into employing a simpler yet more productive connection? Now look to return to the low effort state- where the lengthening actions of finger prop the knuckles up, and where the arm hangs back without bearing down heavily on the hand. Are you getting a clear feel for the radically different possibilities involved in different styles of setup?

Via a combination of these exercises (especially those that involve starting out with movements in the arm), can you keep reducing the pressure until your thumb and finger feel as if there is zero arm weight bearing down though it at all (so it's purely that feel of lengthening the thumb and finger that makes for enough pressure to keep the key safely down without risk of wobbles)? Can you somehow have your arm feel as easy as when you had been resting down, even now? You're probably not 100% there on this extreme version quite yet, but getting gradually closer towards this seeming paradox is one of the big secrets to low effort technique in general. Compare again now to when you had been contacting nothing but thin air. Just rather light connection between the actively extending digits and a depressed key can make all the difference to the effort required, even if you're not trying to rest weight on them at all! In fact, try now with so little weight acting that the key doesn't even begin to go down (example 13)!

Can you feel how even resting on the mere surface of the keys (but still trying to push your knuckle up and away- to firm the fingers slightly and get a sense of connection) can offer significantly more ease and support than resting on nothing? This is VERY tricky so don't overdo it if you feel the arm starting to get tired, as it would in example 1. Come back to this part regularly as a test of your progress, but don't do more than a little with it in each session! The difference in effort between this minimal support for the arm (to get an idea of how little weight rests here, note that the pressure from a mere 50g mass would send an average key down!) and no support at all may be startlingly vivid to you or it may as yet be relatively subtle or even an elusive mystery for now. If you work on the previous exercises, over time you should build up to the point of being able to feel a truly significant difference in the arm's comfort, depending on whether you are supporting yourself in mid air or resting on the mere surface of a key that is not depressed (nb on this one, above all, I cannot overstate the necessity of the arm hanging back- or you literally cannot even begin to capitalise on that which is possible!). Can you even start squashed down, and be sensitive enough to feel the fingers push you away from the key to raise the knuckle WITHOUT moving it (example 14). Sometimes let the key move a touch, to be sure you are feeling how much resting on it is possible. However, the lighting in my video perhaps makes it look like the key was perhaps moving a good deal more than it had been. If any key movement occurs, find a way to keep it slight and definitely don't ground the key. If you can start to feel firmed against the keys (which is absolutely not the same as firming against nothing in mid-air!!!!), this is the good kind of "firm". There's so little force involved that it would be impossible to extend out too much- without the key going down, to tell you about it. In that respect, there's no way of overdoing it without instant indication of the excess. At this point, when you are back on a depressed key, you'll barely have to feel any weight resting on the note at all, to get extreme comfort (and the feel for the springs trying to raise the key up ought to be spectacularly vivid to you). This stage could take a week or even a month of sensitive work so, for now, just continue to observe the differences in various states and see how much improvement begins to occur. Try coming back to these exercises for just a few minutes everyday.

If this exercise still seems abstract to you, I should stress that it relates to just about EVERYTHING in piano playing! Even the shortest staccato demands sensitivity as to how the springs throw the finger back up. Once you can perceive both the resistance and an interaction with it that is based on finger lengthening, you simply need to figure out how to integrate this basic state into regular playing- and start to get it ingrained as a kind of automatic reflex, that you can choose either to perceive or leave to instinct. If you can learn to reliably achieve this quality of whole-arm balance after any note or chord, the level of control (not to mention comfort) afforded to you will improve beyond recognition.

If anyone struggled to perceive the differences in effort clearly, I'll reiterate once more that the basic key to comfort in the arm is to feel an adequate act of extending out the thumb and finger into clear contact and to hang or even lightly pull the arm horizontally back from there, in order to settle into a sense of being lengthened. The secret is not to go out of your way to try to rest considerable weight straight down through the thumb, or to feel the arm sagging down! In a finished product it's the lengthening activity in the fingers that secures a quality point of support and makes functional release possible. If that occurs properly, any pressure from the arm is merely an optional extra. There's an argument for resting down more in the earlier stages, but once you've definitely made contact, the rest is all about freeing yourself by raising the knuckles up and away- not continuing to squash down. The arm only needs to hang back from a well connected hand without getting held up forwards.  To find comfort it does not need to be so inconsiderate towards the hand as to place significant levels of weight through it!

Resting down as a rule with full relaxed weight (as some arm-weight extremists preach) imposes nothing but limits, if taken literally. It's an option, not a permanent requirement (nor even the "normal" way to be). It's not necessary to rob your fingers of the scope to move as freely and easily as Art Tatum's are seen to be doing here.

Addendum

It took me a long time before I came to truly understood why Alan Fraser's books are regularly based upon getting into such positions as those seen here- and it was only when I discovered the lengthening action of fingers that I started to get a deeper insight. If you're looking at the examples and thinking that the position looks absurd, firstly consider that this is primarily a chance to experience the LEAST effort possible to connect the hand effectively.  I'm confident that this is an objective truth and will try to give a practical demonstration of why. After experiencing this particular low-effort balance, you have something to compare other positions to- and can decide for yourself if you still prefer an alternative. However, regardless of what traditional advice says, I assure you that such positions are not unusual at all, among the greatest pianists. Thanks to Youtube, this is available for all to see. Here are just a handful of examples.

 (this first one of Clifford Curzon cannot be embedded)
https://youtu.be/uzljkzRdrLA?t=3m30s

But what of the curved fingers that most  traditional methods preach? Where does a near-vertical and fully lengthened  out 5th finger (something that is actually VERY common among great artists) come into that? Well, let's clarify this in reference to Alan Fraser's "standing up" analogy (which essentially implies the same lengthening activity that all of this post has been grounded upon). I want to go a little deeper into the reasons why this type of near vertical balance is objectively the lowest effort position that can be experienced. Consider what happens when you are literally standing up out of a squat. Quite simply, you are lengthening your legs away from the point of contact with the floor. If you relax your knees once standing, your legs will give way and you will fall back down. There is a necessary level of muscle activity to keep balance once stood up- but the smaller the bend in the knee becomes, the smaller that ongoing activity needs to be. Try slowly squatting down. Can you feel how much harder the muscles work if you stop here? Push yourself back up. Here is an illustration of before and after:

Obviously the first position requires more effort to keep balance. Once the leg is close to vertical (nb. leave a slight remaining bend, to avoid straining the limits of the knee joint), the muscular effort involved in balance has become drastically smaller.Of course, the more you push your body up, by lengthening your legs, the more work you are doing against gravity- to lift your body up. HOWEVER, getting that work done creates a position in which your efforts are virtually done and dusted (because the joints are now almost perfectly in the line of force, rather than at a big angle to it). In effect you "earn" an easier balance for yourself, from having done the preliminary work of opening out at the knee (ie. standing up).

It's much the same with upper body posture too. Relax your posture too much and your neck will slump forward and down with gravity. Many muscles now have to work hard to stop gravity pulling your head all the way down. Lazy posture creates more need for larger intensity of muscle contraction. However, lengthen your spine out and you end up needing less effort to keep in balance. In all of these cases doing USEFUL work by moving up against gravity leaves it easier to continue to balance, from that moment onwards.. The finger works in very much the same way. In this exercise, the initial work of creating an extended finger and thumb will permit the least physical effort possible from the hand, in order to both ground a key and help out with the arm's balance. As long as you are not either pressing from the arm or resting down too heavily, this is the objectively optimal way of keeping the effort down without compromising stability. People usually tend to assume that such a position would involve working harder (and that allowing the knuckle to relax down  would reduce the overall effort, once the key is depressed) but it's not so. Keeping it elevated is what makes for optimal efficiency. Once you've done the small level of work to get the knuckle up, you have a very easy balance to maintain- just the same as when staying standing up rather than squatting down and then trying to stay like that. It's possible to balance well without being fully extended, but it's harder to get it perfect and be sure there's no general stiffness. The easiest thing both in a squat and at a piano is to finish the job and get close to upright. The finger need not be altogether vertical, as with a standing human, but it does help to be between an angle of 45 degrees and vertical, if you want to feel a low effort balance.When standing up out a squat, you create ease by going from initially bent legs to straightened legs via the action of straightening- not via an action of bending your legs further! The bend was only there as a starting point. What happens if we compare this issue to the equivalent issue in a finger?

I believe that the irony is that "curve your fingers" advice is most useful when the pianist feels that the activity is the very opposite ie.UNcurling the fingers. Pianists who start naturally curved and get an instinctive feel for going from there to a less curved position can learn the ease of going from a slightly closed position to a more open one. However, pianists who start curved and stay locked into that (or, even worse, who try to close up further still) while producing sound, end up working extremely hard. Even if you don't go to the extreme of fully lengthened fingers, consider that getting straighter will still progressively improve ease of balance (just as the positions get progressively easier when you go slowly from a very deep squat to an upright standing position). Even pianists who don't go to full length with their fingers can be quite thoroughly dependent on this "standing up" action of lengthening out. I am not dictating the exact hand position that would be considered "correct", but rather being sure to provide the easiest way of feeling the type of action that reduces the workload involved in balance. By experiencing the whole range of the "standing up" movement you can come to decide for yourself which part of it you prefer to balance in (regardless of whether it's necessarily the one I demonstrate here)- and be sure you're making an informed choice rather than sticking to the limits of your experience.

Anyway, once you have options, the pianistic equivalent of the slightly bent squat position is far from off limits- even though it involves slightly greater effort to maintain balance. Not everything is about saving effort- when you have a musical image to attempt to execute. However, remember that the greater the bend in your balanced position, the more work it takes to maintain your balance. If you rarely play with significantly lengthened fingers, it's worth asking yourself whether you're actually subjecting your hand to the pianistic equivalent of this type of workload?

.
.