Guidance, and the student's own voice

There is a moment in nearly every lesson where I have to make a very small decision very quickly. A student plays a phrase and it is not what I would have done — slower than I expected, or leaning on a note I would have left alone — and I have about two seconds to work out whether I am hearing a mistake or a choice. It sounds trivial. It is not. Get that judgement wrong often enough in one direction and you end up with a student who plays beautifully and cannot practise without you. Get it wrong in the other direction and you end up with years of vague habits that nobody ever named. I have made both errors, and after two decades of teaching in Zurich I still think this is the hardest part of the job — harder than fingering, harder than pedalling, because unlike those there is no chart for it.

Two ways to get this wrong

The over-guided student is easy to recognise. They play well, often very well, and after every phrase their eyes come up to check your face. Ask them why they made a diminuendo there and the honest answer is that you asked for one three weeks ago. Give them a new piece and nothing happens for a week, because the piece has not yet been told what to do. The under-guided student is the mirror image: full of opinions, comfortable at the instrument, and quietly stuck, repeating the same physical habit for three years because nobody wanted to interrupt the flow. Of the two, over-guiding is far more common, and I think that is because it feels like good teaching. It fills the lesson. The parent in the next room hears a great deal of instruction. The piece improves visibly within forty minutes. Under-teaching is uncomfortable and looks lazy, which is exactly why it takes nerve to leave something alone.

Some things are not negotiable — most things are

It helps enormously to be explicit, with yourself and with the student, about where the line actually sits. On my side of it there are three categories, and only three. The text: the notes, the written rhythm, the rests, the articulation the composer actually put there rather than the one the student half-remembers. The body: how you sit, how the arm transfers weight, anything that produces tension or risks injury. And method: how to take apart a difficult bar, how to practise slowly with a purpose, how to memorise something so it survives a stage. Everything else — the tempo within a plausible range, the colour of a chord, the shape of a ritardando, which voice comes forward in a texture, how much pedal, where the highest point of a phrase falls, how long a silence is allowed to last — genuinely belongs to the player. That second list is much longer than most students assume, and simply telling them so changes the way they practise. Decisions they are allowed to make are decisions they start listening for.

How much correction is too much

I use three rules, none of them elegant, all of them useful. First: one thing at a time. If I have heard four problems, I say one, and I keep the other three in my notebook — a student holding four instructions applies none of them properly. Second: at least once per lesson the student plays something all the way through and I do not stop them, whatever happens. I write instead of speaking. Playing through interruption-free is a skill in itself, and a student who is never allowed to finish learns to expect rescue. Third: no more than three things leave the room. If my list is longer than three, my diagnosis was too shallow — there is almost certainly one underlying cause I have not found yet. There is also a fourth habit I try to keep, which is silence immediately after a correction. Say the thing, then let them try it twice without commentary. The temptation to refine it while they are still moving is enormous, and it robs them of the part where they work out how the instruction feels from the inside.

Ask instead of telling

The single most useful sentence in my teaching is "what did you want there?" If the student has an answer, my job for the next five minutes is to help them get what they wanted — not to substitute my version, even if mine is better. If they have no answer, then I have found the real gap, and it is not a gap in their hands. From there, a few small moves do most of the work. I play a passage two ways, deliberately, and ask which they prefer and why; the reasoning matters more than the choice. I ask them to write their own decisions into the score in pencil, in their own handwriting — I stopped writing in students' scores years ago, except for fingering, because a score covered in my marks is a score they read instead of hear. And at the end I ask what they intend to do this week, and let them build at least part of that plan, then adjust it rather than replace it. Sometimes I let a decision I disagree with stand for a week. Hearing your own idea not quite work is a better teacher than being told in advance that it would not.

Independence arrives in stages

None of this means handing a beginner the keys. Freedom without craft is not freedom; it is just noise with confidence, and a student who has never been shown how a phrase can be shaped has nothing to choose between. So the amount of room grows with the toolkit. Early on, the choices are small and real: pick the dynamics for these eight bars, choose which of these two pieces we learn next. In the middle years, a student should be finding their own fingering and able to justify it, and drafting their own practice plan for me to react to. Later, they should arrive with an interpretation already formed and be able to defend it when I push. The pattern differs by person, too. Adult beginners often over-defer — "just tell me what's correct" — and mostly need permission and a bit of insistence. Teenagers frequently arrive with strong instincts and no craft to execute them, and for them the guidance is not about taste at all; it is about giving them the technique to do the thing they already want to do.

The signal I watch for is a modest one. A student sits down and, before playing, tells me what they decided to try this week and why. At that point the balance is roughly right, and my job becomes smaller and considerably more interesting — reacting to a musician rather than assembling a performance. The aim was never that they should play the way I play. The aim is that in ten years they play something that would not have occurred to me, and that it works, and that they can tell me exactly why they did it.