Last spring, some of my piano students participated in the annual Vancouver Music Festival Workshop, a non-competitive festival at which the students receive immediate oral feedback from the adjudicators. My students, in the youngest festival classes, had the privilege of working with the lovely and encouraging Yvette Rowledge, an experienced adjudicator with a knack for expressing comments in a way that young children can understand. The festival took place at a church, and the students played on a beautiful grand piano. Of course, every piano is a little bit different, and it takes a little while to adjust to playing a different instrument, especially for these young pianists. Yvette acknowledged this to the students, often by using the phrase: “This piano doesn’t know your piece yet”. This was a cute way of saying that the students’ brains hadn’t adapted to playing on this different piano.
Changing our movements to adapt to different environmental conditions is known as motor adaptation, an important category of motor learning. Motor adaptation is believed to work through feedback modification of existing motor programs. Think about it this way: when you go to play something on the piano, you automatically move your fingers, hands, and arms a certain way, with a certain amount of force, because you have learned that the keys give a particular amount of resistance and are a certain distance apart, etc. You have learned that making these particular movements on the piano will lead to certain sensory results: your fingers will feel the resistance of the keys and your ears will hear the resulting pitches at a certain tempo and volume level. This mapping of motor commands to sensory consequences is called a forward model – the brain takes the motor commands that have been generated, and predicts forward what the sensory result will be.
The part of the brain that has been shown to do this prediction is the cerebellum, a structure the size of a slightly flattened baseball, at the lower back of the brain, just above where the back of your head attaches to your neck. The cerebellum not only predicts what the sensory result of actions should be, it also compares the real sensory result with the predicted one. So if you’re playing a different piano and it has a stiffer action than you’re used to, the sensory result will be different than you expect: your fingers will feel more resistance, the keys will move more slowly, and the sounds you produce will be quieter and perhaps uneven. In other words, what you feel and hear will be different from what you expected. The cerebellum will register that there is some kind of error, some difference between the predicted result and the actual result. The cerebellum will then alter the forward model to try to reduce that error. In this example, it will change the model to take into account that more velocity is required to move the piano keys.
This is what motor adaptation is: the updating of a forward model, so that the prediction about sensory results better matches the actual sensory results of the movement. Sensory feedback allows our brain to learn the precise forces, directions, and velocities of movement required in a particular situation. This updating of the model is an ongoing activity, happening while we perform a movement. It’s not like you play something on the piano and then, after you finish, your brain figures out what went wrong. Motor adaptation is happening while you’re playing: as you press down the key, you immediately feel and hear a difference in how the piano responds, and you adapt our movements continually. Of course, getting the adaptation exactly right takes a bit of time, so that your playing improves throughout the whole piece.
Motor adaptation learning occurs all the time, in all sorts of situations, whenever you learn to change the force and/or direction of your movements in response to changes in the environment. For example, every time you drive an unfamiliar car, you have to adapt to how it handles. To turn a corner, you may need to exert more or less force on the steering wheel than in the car you are used to driving. Similarly, if the brakes in the unfamiliar car are more sensitive than the brakes in your own car, your stops may be jerky at first. But you quickly learn, without thinking too much about it, what forces are required for steering and braking in this car, and adapt your movements so that you turn and brake smoothly and automatically.
Motor adaptation certainly plays an important role in performing music. As we've seen, every time a musician plays on a different instrument, she must adapt to the different forces needed. When we play in a different hall with different acoustics, we adapt. When we need to play more quietly because someone is sleeping or watching TV in another room, we adapt our movements. If the bench is too close to the piano or an orchestral player is crammed into a tiny orchestra pit, we need to adapt our movements.
On the practical side, studies have shown that it’s possible to improve at adaptation. If we practice playing on lots of different pianos, we are better equipped to make the necessary changes for each individual one. This occurs because our cerebellum will get an idea of what types of possible errors are out there and how to correct for different errors.
The ability to adapt to different pianos is especially important for my beginner students, some of whom practice at home on cheap digital keyboards with unweighted keys. (I don’t recommend it, of course, but I understand parents’ desires to keep costs down while they’re uncertain about whether their children will continue with piano). When these children are playing on an acoustic piano, and actually have to press the keys down hard enough to make the hammers hit the strings, the forces required are quite different, and I hear the familiar refrain of “It was better at home”. They need to learn to adapt. And I think the more different pianos they play on, the better for motor adaptation. That way, every piano they play will “know their piece”.
Krakauer, J.W., and Mazzoni, P. (2011). Human sensorimotor learning: adaptation, skill, and beyond. Curr. Opin. Neurobiol. 21, 636–644.
Shadmehr, R., Smith, M.A., and Krakauer, J.W. (2010). Error correction, sensory prediction, and adaptation in motor control. Annu. Rev. Neurosci. 33, 89–108.