Please note that posts appear from the most recent first, rather than in consecutive order.

Please click here for a full chronological index of posts

Tuesday, 3 November 2015

Piano Technique, Weight in Motion, Boxing, Taichichuan and The Cherry Tomato- an objective analysis

Although I hope to get some of my more specifically themed core posts completed in the near future, for the moment here's a collection of musings on a whole range of issues that was inspired by this article, by jazz pianist Mark Polishook.

The primary theme here is of weight (and the limitations of what it can accurately account for in objective reality), although the relationships between pianism and martial arts/tennis may well be of interest to some. In short, I can agree with very little asserted within the article and thus wish to debunk numerous claims within, via simple practical illustrations. I certainly hope that this post won't end up as a tirade of negative vitriol. However, I'm publishing it specifically because of how interesting I consider some of the off-shoots of my disagreement to be, with regard to my own alternative conclusions about how the analogies can best be related to pianism (in a manner that is both simple to understand and consistent with credible science). Also, I have no personal beef with the author and would like to start by making it clear that my intense disagreement with the validity of what he draws from these analogies should not be interpreted as any direct sleight upon his capabilities as either a teacher or as a pianist. I've often witnessed teachers (whose verbal or written explanations of technique I would dispute) using practical approaches that are so effective that the direct plausibility (or lack of) in the explanation itself doesn't in any way interfere with what they draw out of students. My disagreement is in no way intended to read as any kind of professional condemnation. Indeed, if he wishes to provide any follow up comments to what I have to say here, in the spirit of two-way dialogue, it would be most warmly welcomed.

However, the following arguments were triggered for two primary reasons. Firstly, there are numerous issues being portrayed as if having some kind of real life grounding in science- most notably the outrageous pseudoscience (nb I don't use such a word lightly, and will substantiate that assertion fully) that is quoted from boxer Jack Dempsey. I have a big problem with ultra-subjective imagery portrayed directly as a reality for two reasons. Firstly, a subjective method does not become any more effective on a practical level, due to having flawed analytical logic tacked on. It if does offer something useful on a practical level (which it may very well, to those who come at it from the right starting point), there will be ways to trigger the same practical effects, without also creating confusion. But above all, if it is not a real truth, even on a practical level it's not only very unlikely to work for everyone, but may well be least effective of all for those who best take explanations at face value. A great many of the ideas dealt with here are extremely prevalent beliefs in modern pianism, yet represent just about everything which I consider to have severely held my technique back over a period of many years. Indeed, while I'll go on to give theories as to why scientifically implausible ideas about (supposed) use of weight might well help a boxer or tennis player, I'll also show quite how easily they could hinder some of the most important fundamentals of pianism (unless the pianist has already been lucky enough to have developed them). 

Anyway, the first assertion which I take overwhelming issue with is as such:

"Third, body weight rather than finger strength gives a much greater sense of control at the piano – weight-based technique returns more feedback to the body than finger technique alone."

Firstly, we have a false dichotomy. It's not a simplistic choice between one option or the other. As I illustrate in my post on balance, intending to use weight through the finger often leaves it with no choice but to work considerably harder, compared to when the finger itself is proactive about connecting out of free will. As for the claim about sensitivity, I'm simply baffled. Let's introduce an excellent analogy that I heard pianist and teacher David Kuijken use. If you want to scratch your nose, do you do so by moving the whole arm back and forth from your shoulder, or do you move the finger itself? Well obviously the arm carries the finger to your nose. But I certainly don't get as much feedback or sensitivity when I try to scratch by moving the whole arm, while the finger only seeks to maintain "structure" (rather than get on with performing the scratch for itself). By extrapolating further on this idea, it also gives an excellent representation of how narrow it is to offer but two polarised options of "finger technique alone" or weight. If I lie on my back and involve arm weight in the scratching, it's the last thing that could add to either control or sensitivity of feedback about the movement. So, I certainly don't use any weight and neither does it help in either respect. But do I scratch my nose with the finger alone?

Well, it depends what you mean. The arm always plays a role of some kind, whether here or at the piano. Just not one of having to either generate the basic movement or apply notable weight. My arm is entirely responsive to the finger movement, but it just doesn't need to take an active role in creating the movement. Likewise at the piano. If "finger alone" means locking the arm stiffly, it's a bad idea whether you're scratching your nose or playing the piano. But a responsive arm is something else altogether than a weighted one. To imply just two options doesn't bring us to the real issues. I get the most feedback when my finger strokes the key and the arm is made light enough to respond freely to resulting reactions. It's not finger technique alone. But neither is it anything to do with weight-based technique. If I weigh down notably on every finger, my arm stifles my freedom to perform that finger movement, and doesn't "breathe" with the reactions. With notable weight, things tighten up in the forearm when you try to move the finger with anything resembling a deliberate quality- because the application of weight starts to create a blockage against the reaction, thus working muscles harder. It's not unmanageable for a single note, but it's not half as sensitive as what I experience by stroking the finger and keeping the arm light enough to respond freely.

See the difference in practise at the piano here:

Which looks more sensitive to you? In the first execution, my specific goal is to use "weight-in-motion" to move the keys. So my arm thus sinks down while moving each individual key, before having to lift back up to repeat. Otherwise weight is not "in motion" And as Polishook also claims "Again, no need to use strength – arm weight alone is sufficient". Sufficient to move a piano key, yes. Sufficient to give the finest control over every sound, or to sculpt a longer phrase though? Not really. Perhaps the author will say I have followed his instructions wrongly in some way. However, I would welcome him to clarify precisely which points I am disobeying, if so. The reason this style of playing doesn't work well for a phrase is specifically because I followed the explicit reference to "weight alone"- and thus ended up producing sound by plopping the weight of my arm down on each key, without moving the finger. Not only was there no word from the author on the need for to do so, but we were also directly told that that the weight would be sufficient. I'm not trying to play Devil's advocate here. This is the genuine mistake I made in the past, as a result of such one-sided descriptions. I'd be glad if he were to clarify whether I have done as he intended, or whether he feels I've gone against something he states. But seeing as we're being told that weight is supposed to be in motion, that's exactly what I am indeed doing in the first example (while I instead keep it balanced in the second, and produce the tone with the most wilful motion of my fingers).

Also, which looks closer to the fluid energy of tai-chi, visually? Bobbing around to set weight into motion, or stroking the finger through the key until it gradually evolves into a "standing" position that is still brimming with life and vitality? Which looks more like smooth walking? Dropping down through individual notes is a good exercise for a very tense student to loosen their arms up, but to literally produce sounds one by one with a series of individual arm motions is but an occasional tool in wider technique. This is absolutely not the fundamental building block of normal playing. Today, there are occasions when I use that type of motion, deliberately. But I generally find far more sensitivity and control when my standard intention is to stroke the keys via generous finger motions, while the arm follows the phrase horizontally and without any intention to weigh down on each key. The arm responds freely to every finger action- meaning that I am neither using "finger technique alone" (as the arm is abundantly involved in the organisation through a phrase) nor employing either gravitational weight nor arm momentum to provide the basic motion of the key. 

Anyway, to see whether it is fair to assert that there is more sensitivity through a weight-based technique, try the exercises within this video away from the piano. Start with the hand in something resembling a playing position and either rest down on it, or drop down from a small height. Then start again with the hand resting down only very gently, fairly flat, and lighten the arm all the further still while the fingers and thumb gently draw in together, to open the arch. Do you experience more sensation in the contact of your fingertips when resting weight down into the hand (or to put it another way- when the fingers are having to resist the weight that you are applying to them)? Or when gently stroking the fingers inwards while the arm is light enough to drift up and away with the reaction? I definitely favour the latter. If I should switch to wishing to rest weight down, after having finished stroking the fingers in, the loss of sensitivity that occurs can be absolutely phenomenal! Just the most gentle connection is enough, when the hand finds a living balance through a gradual path of unrestricted movement. When feeling the soft fur of a cat, would consciously striving to rest the weight of your arm upon the cat offer more quality of sensation than a lighter and more gentle contact? 
It's worth noting that if you also should attempt the exercise while starting from significant weight (I mean the prior one to that- leave the poor cat alone please), you'll experience just how stifling weight can be upon the hand's ability to expand safely. Can you spot on which occasions I was moving freely and easily and when I was having to apply a little strength, to perform the movement against the resistance of extra weight? Find out for yourself both what gives you the most feedback and what feels the most effortless, when you play around with the options. Finally, try experimenting with a similar style of recitative passage at the piano, as I demonstrated earlier (I used a passage from Tchaikovsky's Romance op 5, if you want to try the same one).

"So finger strength isn’t the issue at all. Using the resources of the body efficiently – the weight of the whole mechanism all connected together as it is – that’s the issue. Athletes use whole-body weight-based technique all the time. Which is why tennis players talk about getting body weight behind the racquet. Same in baseball with bats or really in any sport where a stick or a bat hits a ball. Weight-based technique is why balls get “hit out of the park” so to speak.(...)

The advantages of weight-based technique are why boxers put their body into a punch."

I barely know where to get started here, as it's not only remarkably one sided but gives us no significant clue as to what the word "weight" is even supposed to mean to the reader. To illustrate this, let's think what weight can add to a punch in the only situation where it's notably literal, firstly. Sit on a bed or soft chair and put your hand in the air. Then relax your arm and allow your hand to fall freely down. Okay, you probably wouldn't want your youngest daughter to have been lying right underneath that, but would you say that you could have knocked out Mike Tyson, had you managed to channel that into a punch? After all, we all remember that scene in Rocky II in which he raised his fist in the air, let it fall on top of Apollo Creed's head and knocked him out cold (or am I misremembering?). Gravity is really not all that quick to generate speed. Drop a ball from a height and see if you can then outrun it with your hand to catch it again. Not exactly an astounding feat, because the earth's g-force is nothing terribly fancy as far as rates of acceleration go. It's especially lame compared to the levels of muscle generated acceleration that boxers put into a genuinely powerful punch. And that's to say nothing of the fact that it's impossible to even send any notable weight into generating the vast hand speed associated with a horizontal punch.

Jack Dempsey's diagram of a sled shows how we can redirect acceleration via weight to work in a different direction, supposedly illustrating that it can also be directed to a punch. Well, Jack, I'm afraid you're no heavy-weight when it comes to intellect. Now, I suppose you could grease up a slide and rest your fist on it. At least I have no alternative idea how else we might redirect energy of a descending fist into a horizontal path- and he doesn't seem to have even tried to conjure up any theory of his own. You could let your relaxed fist slip down it and then see how much speed it generates before contacting a stack of bricks at the bottom, although I'm less than convinced that you'd impress sufficiently within the local dojo to trigger an emergency black belt award ceremony there and then. While I probably wouldn't fancy my chances against Dempsey in a back alley, in a one-on-one bout of bare-knuckle science he'd probably have been lucky to see the end of the first round against an average high schooler, if he thought a punch might in any way be linked to either falling babies or speeding sleds. The fist can neither fall for long enough to gather the kind of speed that either a baby or sled builds up (not even if dropping vertically) nor is there anything to redirect the utterly unimpressive speed into a horizontal path. In fact, to put these issues squarely into perspective, the fastest recorded punch is 44 miles per hour- which would require a fall of some 130 or so feet, for such a speed to be reached via falling weight!!! So what is all this talk of "weight" supposed to mean- given that gravity was so bad at the task of generating a serious punch, even in the ridiculously contrived situations where it could make a true contribution?

Well, some serious clarification is due, if anyone is to make sense of such a colossal discrepancy between the myth and scientific reality of "weight". Here, I'd like to expand on a style of analysis I've previously used at the piano regarding relative speeds- although I'll illustrate it through a practical exercise. Firstly hold your elbow slightly bent and swing your hips to the side, while keeping the arm in a held position. Here we used the whole body to accelerate the fist. A formidable blow? Hardly. The fist speed was limited to the speed generated by the body, which was very small. Turning the whole body through the strike is almost certainly what people mean when using "weight" in this particular context. The word tends to be used in a colloquial fashion, as a substitute for the more accurate term "momentum". Polishook may have intended this colloquial meaning in much of his article- although it's very confusing if so, as the passages he quotes from Dempsey are about literal weight in relation to gravity. That is an altogether different issue. Literal weight can create acceleration instead of muscles (although we've seen how dubious this was). Colloquial weight, (technically momentum) in boxing, tennis or baseball is generated almost exclusively by muscular acceleration of the body and thus has zero in common with his pianistic idea of using weight rather than muscle, for energy. As far as I can gather, he appears to be switching freely between the two different interpretations of the word "weight", as if they were just a single concept, rather than entirely different ones (if not, I'd warmly invite him to clarify what he does actually mean, as it's not at all clear to me).

Anyway, as we've seen, using the body "weight" (or rather, momentum) to generate the fist movement achieves next to nothing on its own. We accelerated the fist to a small speed, but divorced from more important movements it doesn't even register as a real punch. So, this time turn your hips and also turn your shoulder. On top of the speed achieved by the hips, we have added some speed to the fist from the shoulder movement. The fist is now separating out from the body in what I call "positive movement" and thus reaching a greater speed than that of the body itself. Although we didn't really add much yet. So this time, add the extension of the arm via the elbow. Finally, we have a real punch. The arm adds a truly notable level of acceleration to the fist, away from the shoulder. This speed is now being added to speed generated by the body. The most important ingredient in accelerating the fist was by far the arm. Clearly it makes more sense to build up a good jab first. The body can then be added, to supplement momentum behind what is already a good fist speed. But if the fist itself is not accelerating out and away from the moving body (with plenty of speed of its own) the body has been conclusively demonstrated as incapable of compensating for the absence.

Try applying the same concepts to a tennis forehand. Once again the body produces pitiful results, when merely coupled with the arm. If the arm itself is swinging freely and easily, you can turn the body through, to achieve some extra racquet head speed, plus considerable extra momentum through contact. But you cannot possibly accomplish results by pretending that the arm is only there to make a structure that can pass on the power from the hips. No, it's there to perform the stroke and it must do so without fail. The turn of the torso can again add extra power, but the arm must always individuate out from the body, with plenty of its own. It's not about unifying the body into one piece that merely transmits motion from the start of the mechanism to the very end. It is a matter of parts both receiving speed from the prior links in the chain AND individuating out from them by actively generating independent movement.  Successive parts need to make their own meaningful contributions to motion, or the whole thing fails badly.

Now, in either of these scenarios, I see no problem from thinking about weight (whether in the colloquial or literal sense). It's abundantly obvious that you won't achieve results from locking up your arm and then turning the hips. It's equally obvious that drooping limp body parts towards the floor is not going to generate control over horizontal motion. The imagery of gravity doing work can loosen up muscles in readiness for unrestricted freedom of movement, without creating a genuine distraction from key elements. In no way would pretending that gravity is responsible discourage those individuations of body parts, while generating acceleration.

But let's return to pianism for a while here, in an entirely comparable fashion. When a pianist who hasn't first learned to move his fingers properly thinks of using arm weight to set the keys in motion, it's like telling a tennis player who turns his body (without swinging the arm a jot) that he needs to forget his arm and use his torso more. Okay, the tennis situation is ludicrous and unlikely to ever occur, but the pianistic equivalent to such a situation is exceedingly common. Instead of necessarily freeing the mechanism in readiness for essential hand movements (which is what I believe happens under the surface, as the explanation for success stories), weight ideals at least risk causing distraction from them. At worst they may even gradually program an impressionable mind to actively intend to avoid finger movements (remember how much strenuous it was to attempt notable finger movement during armweight, in the earlier exercise!). I suffered badly from this and I've encountered numerous students who can sink weight, yet have rigid/collapsing hands that have never learned to apply the necessary baseline of movement to a key. Under arm weight, a hand that has inadequate capacity for movement will buckle under the weight into what I call negative movement.

Think back to a decent punch. The aspect that people almost always stress is the role of the body. However, the fist is carrying the greatest speed found anywhere in the mechanism, when it individuates out and away from that turning body. It could not correctly be seen as merely something that transmits the speed of the body. It must not only add extra movement on top of that- but a truly significant amount. Likewise in tennis, the fastest speed generated must be passed specifically to the racquet head, when the arm individuates out from the turning motion of the body. None of these examples are about coupling parts, but about separating them out from each other during the action. In the diagram above, the knuckle ends up travelling faster than the fingertip (due to the finger being compressed from both ends, between the force of arm-weight and the key's resistance). Thus the fastest speed peaks at the knuckle. Not only does the fingertip fail to individuate out with a higher speed, but it doesn't even manage to move as fast as the knuckle (which is gaining on the tip and thus certain to be moving faster than it). The extra energy carried in this speeding knuckle joint merely crashes into the keybed as an aftershock, without reaching either fingertip or key.

This time, the finger is generating at least some motion of its own (just as the arm had to be in the punch/tennis stroke), meaning that the fastest speed is unquestionably channelled specifically into the key itself. The fingertip is now gaining further away from the knuckle- just as the fist was gaining away from the body in a suitable punch. Even if we are intent on playing via individual arm movements, the full potential won't ever see the business end of events unless the hand can make at least some small positive contribution to slice through the key resistance. Oh, and in case you're wondering why this matters (given that pianism isn't all about generating the fastest possible key speed) this post explains why directing your greatest acceleration towards the point of contact has as much to do with precise control over the very softest sounds, as it does to with producing the loudest ones efficiently. Although, for a more a concise example, picture aiming a ball at a near target. It's every bit as important that your hand is accelerating out from your body for a gentle but precise throw, as it is when aiming for raw distance. 

Moving on, if we still insisted on using the weight to generate the key motion, how fast could weight generate repetition? Firstly, lets look again at how effective gravity really is as an accelerator. Throw a ball in the air to around eye level and watch as it changes direction. Do you notice how it seemingly "hangs" in the air for a while?

Okay, it's not quite as extreme as Wile E Coyote, in the seconds before he looks down and finally plunges into the gorge, but using weight to produce movement is not any basis for fast reversals of motion. Consider also that a piano key moves but a cm or so. How much speed was generated in the first cm of the drop? Very little at all, even in free fall (and remember that a piano key will also be putting up notable resistance that lowers the rate of acceleration). The falling baby of Dempsey's diagram could scarcely be less relevant, because in even moderately fast pianism we don't have time for long falls, over which to gather acceleration. In slower and particularly in chordal pianism it's indeed possible to gather speed over a bigger drop, as Rubinstein famously took to matters of extreme. But it's objectively inaccurate to attribute the possibility of achieving more than around three (probably a generous estimate) or so key movements per second to free-falling weight. Even if you pull back very sharply after each drop, we're looking at six individual reversals of direction per second. Take a small object and drop it from a cm or so then pick it up and repeat, as fast as you can. Even if you should choose a bouncy object, it won't be very fast. This should give you a pretty good idea of how relatively slowly weight creates speed, in the early moments of a fall- and of how utterly implausible it is to attribute key movements to weight, during even a moderately brisk scale. 
"So, again, no need to use strength – arm weight alone is sufficient. Which doesn’t mean muscles aren’t necessary. Because they are. The role of the muscles is to help the body maintain an efficient structure – so we can use and apply body-weight-in-motion."

To say that the above is a case of the most extraordinarily one-sided wording would be generous. I'm afraid it's not technically even correct, unless we're talking about an extraordinarily narrow style of limited piano technique, that would rule out all manner of possibilities. As we saw in boxing, weight in motion cannot possibly explain the world record punch speed. Given that it wasn't generated by a boxer having plunged from the necessary 130 ft in the air, the likelihood that his muscles created the acceleration is a safe enough bet. Neither can weight come close to explaining many of the most routine ingredients of pianism. Aside from mere "structure" muscles have an unequivocally important role of generating the movement that causes keys to go down. I've already illustrated how few repetitions are possible per second, by setting weight in motion. If the author didn't intend the literal interpretation of the term that he himself chose to employ, I could only invite him to consider alternative wording- as I don't know how myself or any other reader might be able to deduce what he really is alluding to. But the statement, in the form he asserts, it is incorrect beyond any reasonable doubt.

It's interesting that the video selection is of Errol Garner- who certainly does depend on plenty of literal use of falling weight to move keys. Now, I take no issue with Garner's phenomenal musical talent, but does his fondness for bouncing the arm represent everything you need in pianism? In case you're second guessing, this is certainly not a jazz v classical issue, mind. I've posted it before, but I'll post it again here:

These kinds of runs likely approach some 12 or more notes per second. The arm is neither falling 12 times per second, nor being picked up ready to fall again on an additional 12 occasions within the very same second. As I said, 6 reversals per second is probably a rather generous estimate for what is possible from weight in motion. 24 per second would be plain ludicrous. Muscles generate motion in his fingers, beyond question (regardless of whether it is the ones in the forearm, or the intrinsic muscles in the palm) while the arm moves at a perfectly leisurely rate. Yes, there are places where he does use dropping gestures- but exclusively in places where he has the time to do so, and where it's conducive to the type of sound he wants. For smoothness and for speed, it's abundantly visible that results are produced by moving the arm laterally, while the fingers produce the motion of the keys (without letting the arm's weight collapse down, other than as the exception). This very simple style of motion is the foundation of piano technique. It works every bit as well for Tatum style licks, legato melodies in Chopin, or passagework in Mozart concertos.

What Tatum exemplifies is how truly simple means can produce the most extraordinary results. The only thing we should regard as "remarkable" about his finger technique is the sheer speed and control. But far from involving any kind of radical motions, he simply mastered elements that any pianist can already be learning in their earliest lessons. Weight in motion can be simple too, in the right context, but there are places where it precludes results, if it's your basic foundation. The extraordinary things in Tatum couldn't arise from such a premise. Garner achieves some of his most outstanding results via a baseline of dropping weight and I'll certainly not argue otherwise. I doubt if his sense of timing could easily have been achieved in any other way. However, while his runs may not have been quite as super smooth and light as Tatum's notorious virtuosity, there are plenty of moments in the video where he too uses the slower gliding horizontal arm motion, while the fingers individuate to produce the key motions. No matter how many moments of authentic weight in motion are indeed on display, he certainly wasn't some one trick pony who only knew how to drop weight.

I'll stress again that I sincerely have no ill will towards the author. However, I've pulled no punches in these honest responses to his assertions, because I write as one of the many pianists who has previously been duped into assuming that such phrases as "arm weight alone is sufficient" exist in reality, rather than purely in some pianists imaginations. Such ideas seem so enticing, that's it's hard not to be swayed by a picture of having to use very little effort. However, owing to my deficient ability to apply finger movement, attempting to use weight to achieve the impossible ultimately made me work harder and more strenuously. For all I know, the author may teach far more about cultivation of finger movements, when teaching in person. However, by writing an article without a single word to acknowledge that that fingers and muscles must generate not only "structure" but also movement of keys (just as the arms must make such vital contributions of movement in punches or tennis strokes) I see a picture which is not only objectively flawed on an analytical level, but potentially misleading to anyone who should take the ideas at face value, without having first learned extremely advanced finger agility.

I'll finish by giving a brief taster of a concept that will likely be fleshed out more in an upcoming core post on the basics of generating sound. Alan Fraser uses Tai-Chi walking as an excellent analogy for the action of fingers in pianistic legato. However, I wonder if a step machine is an even better one still- given that it introduces something analogous to the actual depression of the piano keys? If you've ever used one, consider the style of motion on a medium to higher setting (so the pedals don't put up a huge resistance to being moved). Do we simply stand on them and allow weight to push the pedals down while the leg keeps structure? No. Do we try to lock the torso into a fixed position and shove down with the legs, as if they were a whole separate entity from a rigidly fixed body? No. Like the title of the machine suggests, we are supposed to imagine that we are trying to walk UP some stairs. By trying to push the body up and away over each step, we actually tend to balance the torso into a pretty steady equilibrium. It doesn't collapse notably down on top of each foot and neither is it weight that drives the pedals. The legs move the mechanism via muscular engagement and there's no question that they could be seen as a mere "structure" to pass on weight in motion. Although a conscious effort to weigh the upper body down through the legs only provides greater burden (just the same as such a mindset only burdens when climbing real stairs) there should be a clear feeling of responses in the torso, to the action of the legs. What if we take a similar image of leading with motions of the fingers that try to push up at responsive knuckles and arms (which end up nicely balanced like the body), when seated at the piano? 

EDIT- I initially neglected to deal with most remarkable analogy of all, namely the tomato. Firstly, after all the talk of body weight in motion, it's a curious turnaround, if he suddenly wants us to hold back and keep the weight extremely subtle. However, according to his follow-up comment, it turns out that the intended point was to show how gently weight can be applied. But it's beyond me why we're even told to apply any- regardless of how gently. It take quite some religious devotion to armweight, to spy the problem of pushing through a fragile tomato without bursting it- and then start with the assumption that we should look to rest armweight upon it. Hell, why not go back to his original ideas and rest the body weight in too, if we're going down such a remarkable path? Gentle or not, I really don't recommend that you start from such an irrational approach to the problem as to reach straight for the big guns, in the expectation that they will offer the greatest precision.

Sunday, 28 September 2014

The mechanics of chaos vs the mechanics of predictability- what quality of movement offers the best tonal control over ppppp, FFFFF and every possible shade in between?

Nb. This post is all about the consistent factors behind control over sound. After outlining some very basic issues about what defines control/the lack of, I'm also going to talk a little of the great artist Shura Cherkassky's methods- and illustrate quite how neatly his approach to the keyboard corresponds with what mechanics suggests to be necessary for close control of sound. Whether we're talking about getting the softest sounds possible or the fullest ones, the essential basis for control is scarcely different. Likewise, staccato and legato have a lot more in common than most people realise. Much the same basic issue dictates whether the precise sound produced is primarily defined by artistic choices, or by mere chance factors that pollute the result. This was originally the end of my previous post on the concept of "positive movement" vs "negative movement". However, as that post was rather long (and as the issue of tonal control is quite so important) I decided to expand that segment into a separate standalone post. Please start by reading the definitions in that post- so you can appreciate what I am actually referring to via these original terms. However, in order to understand why these issues are at the very core of what defines the possibility of controlling your sound, it's not necessary to read the full post right now, but only the definitions. Also, please note that additional illustrative videos will shortly be added to each of these posts.

Why is "negative movement" (a squashing of the distance between knuckle and fingertip during the key depression, as per the diagrams) the single biggest enemy to a pianists control over sound?

Well it's very simple to prove. Basically if two bodies are in contact and one is accelerating the other from behind, as soon as the one at the back starts to slow down a little, the other will start to escape on existing steam. At this point it is coasting and you have given up your means to manipulate it any further. Even if you should catch up again, merely by bobbling in and out of contact, you have already lost the simplicity and directness associated with staying in touch continuously. When a hand collapses as shown, during key depression, it can be said with objective certainty that the knuckle is travelling faster than the fingertip- (which is why the distance between them is becoming progressively smaller). This suggests (with near certainty) that the fingertip is losing available speed while moving the key, rather than accelerating continuously through- which renders it impossible for that fingertip to be passing on speed both smoothly and continuously, in the most predictable fashion. The transfer of energy is necessarily "chaotic"- which means that a very small change in one variable can result in a very significant change to the final result in sound.

If that's not making sense yet, there's a more specific situation that's very easy to imagine, that should readily demonstrate both the truth and significance of this broader principle. Imagine standing still behind a line and pushing a supermarket trolley into motion (without being allowed to take any steps forward, beyond the line). Let's say there's a target a certain number of metres away and you hope to roll it into stopping there, or at least come as close as possible. It's not a miniscule distance but it's not especially far either- say 5m. Is it easier to give a short sharp jab or is it easier to connect clearly with the trolley and then gradually push through until you can stay with it no further? Well, the greater the distance of travel during which you can stay in steady contact, the more margin of error you have (I won't bore you by detailing the proof through physics or associated graphs, but any physicist worth their salt will affirm to you that this is an absolute objective truth, with reference to "impulse"). So, for control, you should push continuously over a longer duration, rather than apply an identical level energy through a brief prod. Incidentally, for direct equivalence to the concept of negative movement within a finger, imagine if the distance between your hands and shoulders were to become squashed, rather than grow out through the push. Think quite how hopeless this would make your control of the trolley!!!

Another example I used before is golf- where all good golfers aim slowly but surely through the ball for short putts- and never hold back or start to slow the club as it reaches the ball! The simple reason is that slowing down during the movement, rather than accelerating, means there is collossal margin for error. A tiny difference in timing means a huge difference in the result, as per chaos theory. There is more inherent predictability when a small but increasing level of force is applied steadily over a longer distance of contact. To put that more simply in layman's terms- you need to putt by aiming steadily THROUGH the ball, rather than move sharply but with intent to stop the club upon contact. Nobody could succeed with a technique that is objectively unfit for the purpose of achieving a high level of accuracy. This isn't a matter of subjective preference- as evidenced by the fact that a golfer who jerks at his putts by slowing down the club is literally unheard of at professional levels. If you care to do well, you need to use a technique that gives you decent odds of getting as close as possible to intensity of kinetic energy that you had actually aimed for.

Going back to the trolley, if we only wanted to move it by a mere 2cm then a shorter prod might be more accurate- but only for the reason that a longer contact would necessarily apply too much energy and would thus mean overshooting the target. However, remember that even the softest pianissimo playing must swing the hammer all the way to the strings. No note that sounds involves the barest minimum. Silent key depressions are the bare minimum- and those are really not something we want when we're supposed to be producing tone! By analogy we're not in a situation where prolonging contact will automatically mean overdoing it. Moving the key too fast is the only way to overshoot your intended volume, so there is no associated benefit to trying to keep the contact short (at least, not when striving for a very soft tone- I wouldn't rule out sharper attacks for specific tonal effects at higher intensities). Making a lasting contact adds scope for control of tone, but does not in any way mean sacrificing the possibility of softness. The classic mistake is to become less positive when trying to play soft. However, moving steadily but with a positive quality is the very thing that makes it possible to play truly quietly- without either risking missed notes, or compromising the level of softness for the sake of avoiding them. 

At the piano, try either pecking at a note and then pulling the finger right back (rather than aiming through a longer contact) or allowing a finger to buckle into negative movement.  This will tend to allow the hammer to escape from your influence sooner, just the same as if you prod abruptly at a trolley. When you stay in contact over a longer distance (by starting gradually and accelerating right through to release) you have more scope to judge exactly how much energy and therefore tone you will pass on. It's important to note that the piano has a point called "escapement", after which the hammer always gets released from your direct manipulation. However, while you can't change anything after that point, what you could very well do is lose consistent contact with the hammer before that point is reached- should your fingertip start slowing down rather than accelerating. It's very easy to overlook this, but it is quite possible to lose influence prematurely.

This is why you should NEVER think of making the fingers more passive or floppy for soft playing! Fingers with give mean you cannot accurately predict either how long you are manipulating the hammer for or precisely how much speed it will receive during that time- before it escapes you. But don't stiffen the fingers either! You can achieve all the "firmness" you need by preventing negative movement in the same way that you always should.  Reliable control of soft tones comes when you bond with a hammer by starting very slowly against the key's resistance but then use "positive movement"- so you are always accelerating the fingertip through subtle resistance until the hammer is set free. The acceleration will be more slight of course, when speaking of soft playing, but you have to stick with it until the hammer is already gone, if you want fine control. This is equally true of any dynamic- which is why the most polished performers move so effortlessly between levels. They're not switching between radically different concepts but doing much the same basic actions, with simply more or less intensity to correspond with the dynamic. In soft sounds, the trick is not to be pulling back with any repression but using a positive and deliberate movement throughout. It's always easier to judge the result if you slowly increase a gentle force, over a longer time, than it is to judge a larger force for an ultra-short duration.  The great pianist Josef Llhevine spoke of the need to ground every key even in soft playing. Although this advice doesn't quite convey all the deeper nuances that I'm looking to reveal here, I have no doubt that it was his way of encouraging this same positive quality of prolonged contact. If you don't ground the keys, it's a very likely indicator that you are repressing and thus failing to maximise contact with the hammer. Once you appreciate the basic concept, looking to finish grounding every key is a very simple way of checking that you're staying suitably positive, without having to retain any more complex thoughts during performance.

This is why so many great pianists are known to have liked practising soft passages loudly, even. What they were achieving is a consistently positive quality of motion, that was not infected by any trace of repression or give in the fingers. Listen to what the great Shura Cherkassky (an artist renowned for the most astonishingly soft yet defined pianissimos) had to say about how he practised, and watch the few seconds of him doing so, that follow.

A pianist who already thinks like a musician simply cannot lose that, by spending some of his time searching for a simple positive quality of movement from each and every finger. He may not have directly analysed the concept that I have labelled "positive movement" but his manner of practise was quite visibly certainly geared towards achieving a simple positivity of motion. When done with sensitivity, this kind of preparation may be rather musically dry, but it trains the reflexes to work consistently in the most effective manner. Such practise isn't where Cherkassky developed his profound musicianship, but it is where he kept up his ability to control the precise sound of every single key. When he put musical intentions in, this is what he could do:

Note just how vigorously and deliberately he GRABS the keys with his fingers, even in the softest accompaniment notes! The secret here is to remember this:

The finest control over soft playing comes when the finger moves every bit as positively as in louder playing, but simply at a far slower speed.

What we see in Cherkassky is a deliberate movement done very slowly- not a short fast movement that is tempered by repression!!! His dry practise may have sounded like a "piano tuner", as he put it, but it trained the deliberate quality that allowed him to reach such extremes without losing control of his sound. At this point, the less specific concept of "positive thinking" is actually closely intertwined with the more specific notion of positive movement. Mental confidence is much more inclined to generate  positive movement than mental uncertainty. With a sense of being even a little mentally tentative or restrained, negative movement is almost certain to be huge a problem. I always like to think of fingers moving gradually but with extreme power (as if they could slowly but effortlessly cut through a brick wall, if need be) for soft playing- so nothing in the mindset implies anything less than a long movement with intent. I may not be on the level of Cherkassky's astonishing pianism, but in recent months I've made a huge amount of progress in terms of how softly I can play without giving up control, by being aware of these defining issues. In the long run, feeling generically mentally positive can start to look after the physical side- provided that you've already ingrained the more specific quality of the physical habits, of course. But if even Cherkassky saw fit to do his practise that way, there's every reason why lesser mortals should give serious consideration to devoting some of their practise time to achieving a similar straightfoward positive quality- rather than kid themselves that they're too artistic to dirty their hands with such mundane practicalities.

Incidentally, the issue of staying in touch with the key's resistance is another reason why I greatly dislike the idea of moving wishing to move a key "fast" for even loud playing. Any golfer, cricketer, tennis player (or other sportsman who uses their body to transmit movement to another object) will not achieve the biggest distance by merely trying to swing as "fast" as they can. To transfer energy effectively, you still have to pace things so the acceleration increases through contact with the resistance- just the same as for quiet playing. If you don't achieve acceleration right there, a reaction force starts to slow you down, so you lose contact too soon to apply serious impulse to your target. Any good sportsman knows this (in instinct, at least, if not necessarily always in analysis) but, sadly, a lot of pianists really don't have either knowledge or "feel" for this. Some misguided methods even encourage things which are likely to obstruct the process of acquiring that feel, if taken at face value. Rather than be slowed down upon connecting with your target, you need to be able to accelerate the fingertip right through the full distance- so the hammer doesn't escape from your manipulation prematurely.

Even in staccato I've recently discovered this to be important. In fact, perhaps especially in staccato- where it's overwhelmingly common for pianists to lose exact control over the intensity, by backing out prematurely. An excellent comparison is the screw shot in snooker or pool. Although your average pub pool player stabs at the ball to achieve backspin, top snooker players will aim low but cue straight through the ball- in line with the same principle of prolonging contact rather than shortening it. Just because you want to allow the damper back to the string, it doesn't mean you should be thinking about trying to back out before you've even finished manipulating the hammer. Finish the job properly with one simple and unpolluted positive intention (the reaction to which should automatically bounce the finger straight back up anyway) and you'll have vastly more control over the tone than when you "peck" at the keys with needless anticipation of pulling back out.

Anyway, back to the very loudest playing. If we return to the trolley example and picture trying to get the biggest distance- would you rather swing a baseball bat at the trolley as fast as possible (and see it bounce straight off a short and unpleasant contact) or use the same technique of connecting to the resistance, before smoothly accelerating your hands through the trolley in a prolonged contact? I know what I'd do. Although I used reference to relative speeds as a basis to prove that negative movements waste energy on impact, I simply don't think of high absolute speeds as being the way to play loud. I think of feeling the key's resistance clearly and then staying in touch with it as far as possible- by continuously accelerating into the escapement (the level at which the hammer necessarily leaves your control, even when accelerating smoothly). I then smoothly continue the same action, so my hand evolves further into a place of openess and comfort. When I'm playing at my best, I know that the process will feeling surprisingly long, slow and effortless during the biggest sounds- not like a short but violent stab. With suitable use of positive movement, there's nothing to relax from- because your movements are taking your hand towards increased freedom from the very outset, rather than headed for the type of compressed impact from which retreat might be desirable. When you have learned the type of action which actively builds upon comfort, through ongoing continuation, even the very loudest FFFFF can be done without fear of impact or stress. In fact, the more positively the hand performs, the safer it actually starts to become.