Functional anatomy and physiology have naturally attended mainly to those functions which occur most commonly in everyday life. Piano playing is a more specialized area, where functions arise which have so far been neglected in medical science. These functions are here described by a pianist (IH) in the hope that medical researchers will respond to fill the gaps. The importance of this lies not only in the understanding of skilled manipulative activity but also in the avoidance of overuse syndrome (OUS) or repetitive strain injury (RSI).

     Medical reference to OUS among pianists began early (21) and continues to the present day (6). We approach this problem, and the converse one of describing skilled playing, by considering the many schools of piano technique. These schools are divided here into two groups: (i) "weight playing" and (ii) all other approaches, to be called "naive playing". Although the term "naive" is used merely for convenience and is not intended as pejorative, we will nevertheless show weight playing as consistent with a high level of manipulative skill, and naive playing as consistent with OUS .

     Weight playing has come to be used by leading artists; we wish to avoid polemics and so will merely give a few quotations from such artists (Appendix A). These quotations are expressed without medical precision, but may serve the present purpose. It is difficult to describe weight playing in words (11 p38)—it is easier by personal demonstration with physical contact or biofeedback—but this must now be attempted.

     The upper arm hangs freely, and therefore vertically. {p. 210} The forearm is horizontal, the height of the chair having been adjusted accordingly. The forearm is not held in position by arm muscles (biceps and triceps) as would be naively supposed; instead it is subject only to (i) the force of gravity opposed by (ii) the resistance of the keyboard (iii) ligamentous integrity and (iv) muscle action (mainly flexion) to stabilize the wrist and fingers. The fingers carry out all the manipulation needed to maintain their posture and to convey themselves from one keyboard location to another. Finger flexion uses the flexors digitorum profundus and sublimus, and these to only a small fraction of their capacity—compare work on parallel bars. Apart from this, most of the work is done by the intrinsic muscles, which accordingly become well developed in such pianists (10 pl3). The lumbricals and interossei play the most vital roles; indeed, "these short muscles are extremely well endowed with special nerve endings which provide them with a positional sense that has no equal elsewhere in the body" (17 p65).

     We now assume a passage of single notes (one finger at a time) for one hand and describe the raising of the finger that is about to play. This does not use the extensors, as would be naively supposed. Instead, the finger which is presently playing is strongly flexed; this action provides a force tending to raise the hand (and with it the forearm), while the non-playing fingers appear to be raised (perhaps about 15 degrees extension at the metacarpo-phalangeal (MCP) joint) with scarcely any work of their own. This small amount of activity by the non-playing fingers is treated in more detail below in the section on the lumbricals. The arm also will be displaced to some extent or tossed around in a ballistic manner by this activity, so that the observer might mistakenly think that the arm is being 'used'; its movements are instead passive.

     Thus the hand, wrist and arm are controlled from the distal end, using the small muscles of the hand, and not from the proximal end. This will be termed the "disto-proximal" theory, by contrast with the naively invoked "proximo-distal" approach. Neurologically too, the disto-proximal direction is the primary one, as the proprioceptive information from the distal parts governs the activity. (Compare the well-known sensory and motor homunculus (20) and see (13 p223)).

     As the forearm is allowed to 'fall', the manipulations are carried out against the resistance of the keyboard. Now activity against resistance is very different from activity 'in mid-air' or with a partially withdrawn forearm. Rather little medical literature is devoted to resisted activity: most writers either assume tacitly that there is no resistance, or that resistance is the exception rather than {p. 211} the rule; in either case, such literature is inapplicable to weight playing.

     The function of the wrist will now be treated briefly. Finger flexion is strongest when the wrist is extended (9 pl62); but constant wrist extension strains the tendons and may contribute to carpal tunnel problems, so relief is provided by frequent vertical wrist movement. Such movement is needed also to allow the passing under of the thumb to facilitate motion across the keyboard. Wrist extension is achieved not by the extensors (together with antagonistic flexion) but by gravity (the "unseen muscle", "the only inexhaustible source of energy" (16 p206,171)) acting on the forearm, while the integrity of the MCP Joint is maintained by flexion. The opposite (upward) wrist movement is achieved by increased flexion against the resistance of the keyboard.

     The above description of weight playing may be applied to perhaps ninety per cent of a pianist's tasks; a full solid chord, for example, may provide an occasional exception with the wrist and fingers somewhat fixated and the triceps supplying additional downward force as accelerator synergist to gravity, analogously to the act of stamping one's foot.

     We refer again to the earlier description of the raising of the finger that is about to play. Present medical opinion appears to be that extension at the MCP joint is a function solely of the extensors. Our hypothesis is that the small amount of such extension occurring in the weight-bearing piano-playing situation is a function not primarily of the extensors but of the lumbricals, with the assistance of the elasticity of the interosseous hood and the proximal gliding of the MCP fibrous girdle.

     The interosseous hood has been described (24 p42) as "the transverse or proximal part of the superficial insertions of all types of interosseous muscles. This structure migrates proximally during digital extension and distally during digital flexion (Fig 1-26)." Now the finger that is currently playing is bearing the weight of the forearm; the resistance thus suffered causes the hood to migrate proximally. At the same time the fibrocartilaginous palmer plates and associated formations are stretched. The location of the point of insertion of the lumbricals of all fingers will thus move dorsally beyond the midline (compare a similar phenomenon at the inter-phalangeal joints (9 p201-2 fig 85)). As a result of the different bias thus set up, the lumbrical will now extend at this joint, whereas in the absence of weight-bearing it would have flexed. The extensors in the forearm thus contribute only as weak accessories, so that a prime cause of OUS is removed.

     {p.212} (A slight difference occurs in pianissimo playing. For in playing above pianissimo level the hand is held well arched, with the MCP knuckles rather prominent (usually referred to as playing "with curved fingers") and the hood migrates as described above. In pianissimo playing, however, the hand is less arched, the MCP knuckles less prominent and the fingers flatter; in this case the hood will migrate less, but correspondingly less extension at the MCP joint is required. The normal tonus of the lumbricals and extensors may alone suffice to produce this minimal extension.)

     The MCP extension of which the lumbrical is thus capable is rather slight in extent, and not powerful. But its contribution to the playing of the next note is merely preparatory; the extension is not an end in itself. This distinction between a preparatory and a deliberate action is important; it may be seen in the action of stamping the foot on the ground—the leg is first raised, but only in the light, weak, preparatory manner, not in the deliberate manner of goose-step marching. A warning may be needed here that the dynamics of fast playing cannot be studied in static situations; the full implications of this remark are reserved for a future paper. Neurologically, the biasing of muscle receptors is a function of the gamma loop (22 p225), (12 p567); again we defer fuller discussion (22 p225), (12 p567).

     In "naive playing" some of the features of weight playing are not present. The forearm is held in position by the arm muscles: it is 'withdrawn' instead of being allowed to drop. Thus finger flexion is opposed antagonistically to extension rather than to the gravitational force. Some activities of flexion which were possible against the resistance of the keyboard are no longer possible, and typically are replaced by activity of the extensors. Thus the fingers are raised by means of the extensors, where pain frequently occurs. There is a relatively close-packed position at the Joints (shoulder, elbow, and wrist) which inhibits mobility.

     A lower level of manipulative skill may also be the result of congenital anatomical differences, in the light of our analysis. For example, in 50% of people the dorsal interosseus inserts into bone, while in the other 50% it inserts into the extensor mechanism (2 p33): in the former case the mechanism proposed for the lumbricals will not be available.

     The above ideas are now illustrated in the playing of repeated 'octaves' by the thumb and little finger of one {p. 213} hand simultaneously playing keys requiring a stretch near the maximum (see Figs 1, 2).

     Fig 1 shows the naive approach: (a) arm activities and (b) wrist extension are pictured, either one or both of which may be used. Here the agonist and antagonist muscles of the shoulder girdle contract to stabilize the humerus in its activities, those of the humerus to stabilize the forearm, and finally those of the forearm to stabilize the wrist and fingers. It is suggested that this may lead to OUS. Since key dip is only 9.5 - 10 mm in total, and 2 - 2.5 mm suffices to return the jack under the roller for repetition, the large movements are redundant. (Regrettably, however such methods are often taught (19 p259-261), (23 p93-99).)

     {p. 214} Fig 2 shows weight playing: here the shoulder is relatively open-packed (which incidentally facilitates quick lateral curvilinear relocation across the keyboard), while extension is reduced to passive synergism (5 p438-9) in relation to powerful flexion against gravity; full bio-mechanical details are however not known at present.

     Digital manipulation when the fingers bear the weight of the forearm (which has been found to be the most advantageous mode in piano playing) has been given little or no attention in the medical literature. A mechanism is proposed for extension at the metacarpo-phalangeal joint; the proposed mechanism depends essentially upon the weight-bearing usage. Overuse syndrome could be prevented (unless a congenital or other physical anomaly is present) by adopting the appropriate usage.

     The authors thank especially Dr. A Ganora, Director, Illawarra Rehabilitation Centre, for his encouragement in these researches over many years; for medical advice Prof. MA Arnold, Dr. WJG Burke, Dr. C McKellar, Dr. PD Middleton, & Dr. L Rassaby; and for assistance the Alfred S White Music Bequest, John Painter, and the NSW State Conservatorium of Music.