How Effective Is Mental Practice for Memorizing New Pieces?

Subscribe to the weekly “audio edition” via iTunes

Many well-known musicians have described using mental imagery to increase confidence, and hone and refine their skills over the years. There’s a lot of research support out there for imagery too, suggesting that mental practice can indeed be a useful supplement (though not replacement!) to your regular physical practice.

So…if visualization can help us increase the level and consistency of our motor skills away from our instruments, could it also help us memorize music away from our instruments?

I mean, intuitively, you’d think that the answer would be yes. But what does the research say?

16 pianists…

A team of German and Italian researchers (Bernardi et al., 2013) recruited 16 pianists, ranging in age from 18 to 36 (mean=26), with anywhere from 15-26 years of formal piano training and experience (mean=20 years).

They were all given two technically non-challenging and easily sight-readable excerpts of music to memorize – specifically, the first halves of Scarlatti’s C major (K72) and A major (K113) sonatas – and asked to memorize one of them via mental practice and the other via physical practice, spaced about 5 days apart.

Mental practice…

On mental practice day, participants were placed at a table, and given only the score, a pencil, and a MIDI recording of the piece. 

They were told “You can freely use whatever practice method you prefer, except for physically playing a real piano.” And then, they were given 30 minutes to see how much of the piece they could memorize, without access to a piano.

When the time was up, they were asked to perform, from memory, as much of the piece as they could.

Then, they were given 10 more minutes to memorize the piece – but with access to a piano – and then once again asked to perform from memory, as much of the piece as they could recall.

Physical practice…

On physical practice day, participants were given the score, a pencil, MIDI recording of the piece, a piano, and the instructions: “We ask you to focus on physically practicing the piece, ignoring any mental images you have as you practice. Do not to stop to mentally rehearse the music and avoid formal analysis of the piece.”

As on mental practice day, the pianists were given 30 minutes to memorize as much as possible, and then asked to perform as much of the piece as they could from memory.

Then, they were given 10 more minutes of memorization time, before performing the piece once more.

So…how did mental and physical practice compare?

Memorization rating

When it came to the number of notes participants were able to recall correctly, 30 minutes of physical practice did enable participants to memorize significantly more of the piece than 30 minutes of mental practice – 326 notes for physical practice vs. 242 for mental practice.

However, the extra 10 minutes of physical practice ended up being something of an equalizer.

How so?

Well, just 10 minutes of physical practice after 30 minutes of mental practice took the participants’ note total from 242 to 319 notes. Which boosted them up into essentially the same level of performance as 30 minutes of physical practice (326 notes) and 40 minutes of physical practice (349 notes) did. And yes, 326 and 349 is indeed more than 319, but these differences were not statistically significant.

All this to say, mental practice by itself led to a level of playing that the researchers note was about 40-60% of the same number of minutes of physical practice. Which by itself isn’t bad – but with that initial mental work under their belt, it took only 10 minutes of physical practice to get to essentially the same level of performance as 30 full minutes of physical practice.

Which is pretty cool – but what about the quality of their playing, beyond just playing the right notes? Like, when it came to articulation and phrasing, dynamics and expression, and so on, did one type of practice lead to more compelling performances than the other?

Performance rating

Three professional musicians rated the quality of the participants’ recorded performances.

And much like with memory, 30 minutes of mental practice with 10 minutes of physical practice seemed to result in a level of performance that was pretty much the same as 30 minutes of physical practice.

On the other hand, 40 minutes of all-physical practice did lead to a slightly higher level of performance than 30 minutes of mental practice and 10 minutes of physical practice. But I think that’s kind of what you’d expect to see anyhow, and the researchers made a point of noting that the difference was quite small.

So this all seems rather promising! But does the study yield any clues in terms of what mental practice strategies might work best for memorization in particular?


What mental practice strategies were used?

At several points during and after practice, participants were asked which mental practice strategies they used. Such as mentally hearing the sounds of notes, feeling the movement of fingers/hands, visualizing the score, harmonic analysis, etc.

Interestingly, some of these strategies were associated with better performance, while others were associated with worse performance.

Formal analysis – like harmonic, melodic, and rhythmic analysis of the score – was associated with better performance ratings. There was also a trend towards better note recall performance too, but it just missed the cutoff for statistical significance.

The better-developed the participants’ aural skills1(the researchers administered an aural skills assessment too)2, the better their performance ratings were, and they tended to recall more notes too.

On the other hand, motor imagery (i.e. imagining the feeling of fingers moving), actually miming the movement of fingers, and singing out loud seemed to have no particular performance or memory benefit.

And visualizing the movement of fingers/hands was associated with worse memory performance, while spending more time listening to the MIDI recording was also linked to lower performance quality ratings (which kind of makes sense though, as a MIDI recording isn’t the sort of auditory model you’d want to emulate!).

So what are the big takeaways?

Best mental practice strategies for memorization?

Well, the researchers make a few recommendations.

For one, they note that auditory imagery – or the ability to hear the notes on the page – seems to be an important factor in whether mental practice memorization works or not. Which speaks to the importance of developing aural skills (oops…I guess I should have taken that class a little more seriously back in school!).

They also note that formal analysis and figuring out the harmonic, melodic, and rhythmic structure of the piece is also important (argh…so I guess I should have paid better attention in theory class too?!).

The authors also suggest that “external models” like listening to recordings can certainly be a useful practice aid – but more to help “build up an auditory/structural mental representation that holds even when the model is no longer present.” Kind of like how a metronome can be a useful practice aid, but you want to be sure to keep your eye on the eventual goal of cultivating a stronger internal sense of pulse, and not come to depend on the external clicks as a crutch.


Well…as much as I hate to end on a downer, I think it is important to note that the way the study was set up, what we’re seeing is not so much long-term memory performance, but short-term memory performance. Kind of like cramming for a quiz before class (and forgetting everything 24 hours later) vs. studying in a way that enables you to remember everything a week later. 

The results are still really intriguing and useful, but in an ideal world, it would have been cool to see a follow-up test one day, and maybe one week later, to see how much of their memory was retained in the long-term. Maybe something for a doctoral student in need of a dissertation topic to explore…?


Bernardi, N. F., Schories, A., Jabusch, H.-C., Colombo, B., & Altenmüller, E. (2012). Mental Practice in Music Memorization: An Ecological-Empirical Study. Music Perception, 30(3), 275–290.


Ack! After Countless Hours of Practice,
Why Are Performances Still so Hit or Miss?

It’s not a talent issue. And that rush of adrenaline and emotional roller coaster you experience before performances is totally normal too.

Performing at the upper ranges of your ability under pressure is a unique skill – one that requires specific mental skills, and perhaps a few other tweaks in your approach to practicing too. Elite athletes have been learning these techniques for decades; if nerves and self-doubt have been recurring obstacles in your performances, I’d like to help you do the same.

Click below to learn more about Beyond Practicing – a home-study course where you’ll explore the 6 skills that are characteristic of top performers. And learn how you can develop these into strengths of your own. And begin to see tangible improvements in your playing that transfer to the stage.


4 Responses

  1. As a military musician, I often have to memorize new marches. I used to spend lots of time physically playing the piece to learn it. Now, I often spend the time while out running (easy and long runs only) memorizing things. I do run/walk so I check a screenshot on my phone in walk breaks to check my success or if I’m stuck on remembering a note. This works best with music I am familiar with from performing non-memorized or having memorized another part before, but if I haven’t done either I listen to a recording and also play through once or twice before I start the memorization. I find myself harmonically analyzing while running to more easily recall the notes to play. As a trumpet player, it’s easy enough to simulate finger patterns while running. I follow up one of these sessions with physically playing the newly memorized piece and reinforce over several days as well. I find the results to be far more consistent than what I used to get – probably because my aural skills have improved and I understand the harmonic tendencies better. Interesting study!

  2. I always love these articles.
    For all the benefits of this study, though, it appears to miss the mark pretty significantly, as the ‘mental practice’ group wasn’t given concrete instructions or strategies to employ to mentally study and memorize the music. Unless those students have a clue how to do that, there will probably be no benefit.

    My understanding is that Nadia Boulanger insisted her students so thoroughly study their repertoire that they be able to write out their scores from memory before ever attempting to play them-someone should research that (I learned this from Ms. Marianne Ploger, a musicianship teacher now at Vanderbilt, I believe; also Emile Naoumoff at IU was a student of Boulanger and may have insight).
    I’ve done this with my own repertoire with much better success in every way (technical, expressive, and secure memory) than the ‘play it a bunch’ approach that we all grew up with.
    It’s a game-changing strategy if one knows how to apply it, and I wish I would have learned it sooner in my studies.

  3. I’d think the benefit would depend on the nature of the passage. Some passages, from Chopin in particular, derive from a complex but exact pattern. So exact that I could write a Python script to return the notes! (E.g., the highly organized nervous breakdown in the middle of Opus 10 No. 3). Mentally working through that pattern away from the piano is definitely useful for me — especially since I find those same passages harder to memorize in the first place.

  4. Intriguing, but I’m going to be even more of a Debbie Downer than you Noa 🙂 and say that you can’t really say *anything* based on this study.
    The sample size is 16. If we’re assuming a normal distribution for the results, and a 95% confidence interval, the margin of error is 25%. Unfortunately, this huge margin of error swamps the reported numerical results.
    Also, the pieces of music were different in the two phases of the experiment.

Leave a Reply

Your email address will not be published. Required fields are marked *