"As I have intimated, if the fingers are used exclusively a terribly dry tone must result. The full-arm touch, in which I experience a complete relaxation of the arm from the shoulder to the finger tips, is the condition I employ at most times. But the touches I use are combinations of the different finger, hand and arm touches. These lead to myriads of results, and only the experienced performer can judge where they should be applied to produce desired effects.

"You will observe by placing your hand upon my shoulder that even with the movement of the single finger a muscular activity may be detected at the shoulder. This shows how completely relaxed I keep my entire arm during performance. It is only in this way that I can produce the right kind of singing tone in cantabile passages. Sometimes I use one touch in one voice and an entirely different touch in another voice. The combinations are kaleidoscopic in their multiplicity.

Mechanical Methods Dangerous

"I have never been in favor of the many automatic and mechanical methods of producing touch. They are all dangerous to my mind. There is only one real way of teaching, and that is through the sense of hearing of the pupil. The teacher should go to the piano and produce the desired tonal effect, and the pupil should listen and watch the teacher. Then the pupil should be instructed to secure a similar result, and the teacher should persevere until the audible effect is nearly the same. If the pupil, working empirically, does not discover the means leading to this effect, the teacher should call the pupil's attention to some of the physical conditions leading to the result. If the teacher is unable to play well enough to illustrate this, and to secure the right kind of touch from his pupils, he has no business to be a teacher of advanced students. All the theory in the world will never lead to the proper results.

"Rubinstein paid little or no attention to the theory of touch, and, in fact, he frequently stated that he cared little about such things, but who could hear Rubinstein's touch without being benefited? I believe that in teaching touch the teacher should first give his model of the touch required and then proceed from this positive ideal, by means of the so-called Socratic method of inducing the pupil to produce a similar result through repeated questions. In this way the pupil will not be obliged to resign his individuality, as would be the case if he followed strict technical injunctions and rules.

Students Should Hear Virtuosos

"For the same reason it is advisable for the pupil to hear many fine pianists. He should never miss an opportunity to attend the concerts of great virtuosos. I can frankly say that I have learned as much from hearing the concerts of great performers as I have from any other source of educational inspiration. The pupil should listen intelligently and earnestly. When he hears what appeals to him as a particularly fine tonal effect, he should endeavor to note the means the pianist employs to produce this effect.

"He must, however, learn to discriminate between affection or needless movement and the legitimate means to an end. Consequent upon a relaxed full arm is the occasional dropping of the wrist below the level of the keyboard. A few great players practice this at a public recital, and lo! and behold! a veritable cult of 'wrist-droppers' arises and we see students raising and lowering the wrist with exaggerated mechanical stiffness and entirely ignoring the important end in which this wrist dropping was only an incident.

Methods, and Still More Methods

"I am continually amused at the thousand and one different ways of striking the keys that teachers devise and then attach with the label 'method.' These varied contortions are, after all, largely a matter of vision, and have little effect upon the real musical results that the composition demands. Touch, as I have previously said, all comes down to the question of the degree of weight applied to the keyboard and the degree of quickness with which it is applied. In rapid octave and staccato passages the hand touch is largely used. This is the touch most dependent upon local muscular activity. Aside from this the combination of muscular and weight touch almost invariably obtain.