Can You Practice Too Much? (Brainjo Bite)

YouTube player


Welcome to another episode of Brainjo bites. So recently I released an episode that was about the ideal length of a practice session. So both the minimum time that’s needed for it to be effective, meaning that we are a little bit better the next time we grab our instrument, and whether or not it’s possible that we could do too much. Or if there is such a thing as a practice session that’s too long.

I received a number of questions after that episode asking about multiple practice sessions over the course of one day – specifically whether or not it was okay to do multiple sessions in a day and, if so, if there was an ideal amount of time to space between those sessions. So those are the questions I’m going to be tackling in this episode.

And as usual, we’re going to look at what the neuroscience research has to say on this topic, which I think you’ll find both really interesting and very useful in terms of answering this question and kind of understanding the best way to practice. And then I’ll conclude with my own thoughts of how to structure your practice sessions in a way that incorporates this research if you’re doing multiple sessions over the course of a single day, so that you get the most out of that practice time and so that you don’t undermine those efforts. And you’ll see why that is in just a moment.

So first to give you the necessary background here, remember that our goal, when we practice is to stimulate our brain to change specifically, we want to build new neural networks that support the knowledge or the skill that we wish to acquire.

And whenever we first start learning something, whether it’s the elements of the periodic table or the hand movements that are needed to play, Mary had a little lamb on the piano, those initial memories and the neural networks that support them are very fragile meaning that they can be easily disrupted or broken and over time with practice and repetition, those memories and those neural networks that support them, it gets stronger and stronger. And this process of strengthening those connections when a memory goes from weak to strong is known as consolidation. And ultimately what happens over time is that those networks become so strong and so stable that they’re essentially yours for good. This is where the phrase it’s like riding a comes from. So that means that those neural networks for how to ride a bike are so entrenched and so strong that they can remain there for many, many years, even if you never ride a bike or activate those networks over that period of time.

And as we’ve discussed in previous episode, this process of learning and consolidation is primarily about strengthening the synapses, or the connections between neurons. So when you’re building new neural networks, your brain is forming connections that didn’t previously previously exist. So in 1996, a landmark study was published in the journal of nature.

Now, as I said, when we first form new memories, they’re quite fragile. And what the researchers wanted to know was if you spend time practicing some new motor tasks you’re trying to learn, and then you practice a different one shortly thereafter, will that disrupt the process of consolidation? So this has been demonstrated prior to this in other domains of memory where this kind of thing happened. So if you were to memorize a list of words, and then a few minutes later, you’re given another list of words, it’s going to disrupt your ability to memorize the first list.

So in other words, the question was would a second motor task disrupt the consolidation of the first one. So in this particular study, they created this apparatus that subjects had to control with a joystick. So it was a joystick attached to a metal arm that had a couple of joints, and it was linked up to control a cursor on a computer screen and the subjects in this trial were given a little bit of time to learn how it worked and to learn how to move the cursor around. And then what they did in the actual test trials was that the arm they were controlling was hooked up to some motors so they could change the forces that were on the arm as the subjects moved around. And they did it in predictable ways so the subjects could learn how it was going to how the forces would change as they moved and they learn how to move it around and hit the targets on the screen.

Again, the specifics of this aren’t too important, but the basic idea is that they were learning how to operate a new device, which required learning a new motor skill. Now, what they did was for the control group, they had them do one of these trial runs on the first day, and then they tested them again on the same task with the same disturbances in the force of the controlling arm, 24 hours later. And what did they find for them? Well, they found that their performance had improved significantly over the course of 24 hours. So they were better at that motor task, 24 hours later than they’d been ever during the previous day. Again, that alone illustrating the principle that the reason we practice is so we can get better in the future, not right then and there. So get better when we’re not practicing again, that’s illustrated perfectly by this study and many, many, many others.

And of course, what had happened during that practice session was that it had stimulated the brain to start changing to form neural networks that controlled this new motor task. But the part that’s relevant to our question today is what they did for the test subjects. Remember I just described the control group. So what they did for them was after that first test, they then had them perform a second trial that involved a different set of forces on the joystick, meaning that those movements that they learned in order to accommodate for those forces on the first trial were no longer of use. So for one of the test groups, they had them perform this second test immediately after with the different forces, and they had another group do it five minutes after another, another do it one hour after, and another do it four hours after.

And then they tested them all on the original task, just like the control group 24 hours later. And so what did they find? Well, here’s a beautiful graph from this paper that illustrates these results. So what they found was the group that performed the second task either immediately five minutes later, or one hour later, did not demonstrate any significant improvements on the original task. The following day. In other words, trying to learn a second motor task after that first one interrupted the learning of the first task. So it disrupted those fragile neural networks and disrupted the process of consolidation.

Additionally, those who performed the second task within an hour of the first performed worse on the second task than those who waited four hours. So learning the first task also disrupted their performance on the second one. And now for the group that waited four hours between the first test trial and the second one, they actually did demonstrate improvement a day later on that first trial.

So I mentioned that this was a landmark study, and that was because it was the first one to demonstrate the phenomenon of interference with respect to motor learning. So what we now describe as motor interference, again, it had been previously shown in other types of memory formation, but this was the first one to show it that it applied to the formation of motor memories. And there have been many subsequent studies since this one, which have not only confirmed these findings, but have provided additional details about the neurophysiology of what’s going on here.

And we now know that this phenomenon of motor interference works in both directions. So you have retrograde interference, which refers to when learning something new interferes with something that you’ve previously learned. So in this case, performing that second task prevented the subjects from learning the first one.

And then there’s also, what’s known as anterograde interference, and that works in the opposite direction. So when learning the first task interferes with your ability to learn a second one at a later point in time. Now, as I mentioned, the timeframe here definitely matters. Remember in this study that the amount of interference diminished considerably at the four hour mark, the research has also shown that neurobiologically what’s happening is that this interfering task is arrested or stopping the process of synaptic strengthening that is needed to form new networks. And a related concept that has emerged in this area to help explain these findings is that of a “synaptic modification range,” and the basic gist of that is that you can only do so much modifying of the strength of synapses at any given time. And finally, another point that is very relevant to our original question is that the type of task performed matters as well.

And what I mean by that is if, instead of having the subjects practice another motor task with that joystick, you gave them a set of words to memorize the formation of that original memory. That motor task memory will not be disrupted. And that’s almost certainly because those kinds of memories involved entirely different brain regions and entirely different sign apses. So there’s little to no overlap between the synapses or the connections in networks for things like semantic memories, remembering lists of words, and so forth, and those for motor memories. And as a general rule of thumb interference applies to learning and memory when there is overlap in those synapses.

Alright, so what is the take-home message from this research? How should we incorporate this into how we, how we practice remember that our original question was whether it was okay to do multiple practice sessions over the course of the day.

And as you can see, this is a really good question because the research shows that it’s entirely possible to prevent ourselves from improving by practicing more. So imagine the scenario where you practice a new tune or a set of challenging measures in a tune in a particular song for 20 minutes or so. And then an hour later, you try to tackle an entirely different tune or set of challenging measures from an entirely different song because of this phenomenon of motor interference, you’re likely to end up worse off than you would if you just done that single practice session. And it’s important to remember that what we’re talking about is when you’re trying to learn something new, when you have a practice session dedicated to acquiring some new skill or knowledge, rather than playing things that are already familiar. So back to the original question here, what I would recommend based on the available research.

So when you’re trying to learn new motor skills, so new movements of your hands, whatever they’re being applied to aim for one session of about 20 to 25 minutes in length. And then I would also suggest right before bed, even when you’re laying in bed visualizing, whatever it is you’re working on during that particular session that will help provide the brain an extra cue that you want it to work on that that night. And then if you want to do another session of new learning, make it in a domain that doesn’t have synaptic overlap with the first one. So I like to think of music learning in three broad categories. We have our technical skills, which are largely those motor components. And that’s the thing that we’ve studied the most with respect to these interference phenomenon. And then you have your development and you have knowledge and concepts.

So if you’re doing another session of new learning, make it from one of these other two categories. Since these are going to have little synaptic overlap with motor skill learning, I find that most people focus a disproportionate amount of their practice time on the technical elements and not enough on these other two categories of ear training and knowledge and musical knowledge and concepts yet it’s those categories that not surprisingly are what separate the experts from the amateurs rather than technical skills. So if you want to put in multiple practice sessions a day, use it as an opportunity to work on these often neglected areas. And then if you do want to work on some new motor skill, that’s unrelated to the first thing that you work on. So another song or another, a few measures or whatever, then make it so that it’s at least four hours after the first one.

Thanks so much for listening to this episode of Brainjo bites, to make sure you catch future episodes. Be sure to subscribe to the podcast on your podcast player and consider leaving a rating and review in iTunes, which helps other folks to funding it as well. To learn more about music courses based on the Brainjo method of instruction, head over to