Two folk explanations dominate popular accounts of why some adults acquire piano skill faster than others in a beginner class. The first is innate musical talent — some people are “wired for music” and the rest are not. The second is mindset — anyone can learn anything if they believe they can, and the differences come down to grit, not ability. Burgoyne, Harris, and Hambrick (2019) tested both explanations against a third candidate that gets less airtime in popular music education writing: general cognitive ability. The result was clean and not particularly fashionable. General intelligence predicted piano skill acquisition, music aptitude looked predictive until intelligence was controlled for, and mindset did not predict acquisition at all.
The Study That Pinned Down the Comparison
The design was tighter than most music-learning research. The authors recruited 171 undergraduate students with little or no prior piano experience and put each of them through the same procedure: measure cognitive ability, music aptitude, and mindset beforehand; then have them learn a short piano piece guided by an instructional video; then have them perform the piece from memory while a panel of musicians scored melodic and rhythmic accuracy. The skill-acquisition score is the dependent variable; the three pre-existing individual differences are the predictors.
What makes the design decisive is the multiple-regression framework. Each predictor’s contribution is estimated after controlling for the others. If music aptitude correlates with skill acquisition only because high-aptitude people also tend to be high-g, the regression will show music aptitude’s beta dropping toward zero once g is entered. If mindset matters only because it correlates with effort during the lesson, that pathway has to survive controlling for the cognitive predictors. The output is a partition of the explanatory variance into what each construct contributes uniquely — not what each correlates with in isolation.
The three predictors together accounted for 22.4% of the variance in skill acquisition. That is a substantial chunk for a single 25-minute learning episode in beginners, and the partitioning of that variance is what the paper’s contribution rests on.
General Intelligence Did the Heavy Lifting
Of the three predictors, general intelligence carried the largest standardized regression coefficient and was the only one statistically significant after controls. The g composite Burgoyne et al. constructed pulled together working memory, processing speed, and reasoning subtests — the standard WAIS-style cluster.
This is not a surprising result given what learning a piano piece from a video in 25 minutes demands. The learner has to hold a multi-step motor sequence in working memory while watching the next step. They have to decode pattern relationships in the melodic and rhythmic structure to predict what comes next. They have to translate a symbolic-visual representation (the keyboard layout, the finger positions shown in the video) into a coordinated motor output. These are the canonical demands working memory, fluid reasoning, and processing speed each measure. A test of general cognitive ability is, almost by construction, a test of how well someone can handle exactly this kind of compound information-processing load.
The practical implication is that early-stage skill acquisition is bottlenecked by cognitive throughput, not by domain-specific aptitudes the student has not yet had the chance to develop.
Music Aptitude: Real Correlation, No Unique Contribution
Music aptitude — measured with a standard auditory discrimination test — correlated with skill acquisition at the bivariate level. By itself it looked like a predictor. The result that matters is what happens when general intelligence is entered into the regression: music aptitude’s beta drops to non-significance.
The cleanest interpretation is not that music aptitude is unreal. It is that, in beginners, the variance music-aptitude tests pick up overlaps heavily with the variance general cognitive ability tests pick up. Discriminating subtle pitch and rhythm differences is itself a cognitively demanding task; people who do well on it also tend to do well on processing-speed and working-memory tests. Whatever fraction of “music aptitude” might be domain-specific in advanced musicians, in adults who have never sat at a keyboard the construct is largely redundant with g.
This pattern is consistent with what Macnamara, Hambrick, and Oswald (2014) found in their meta-analysis of expertise across music, sports, games, and education: domain-specific predictors get more independent leverage as expertise accumulates, but at the beginner stage, general cognitive ability does most of the explanatory work.
The Mindset Null Result Is Not an Outlier
Growth versus fixed mindset — measured with Dweck’s standard scale — did not predict skill acquisition. The beta was near zero and statistically non-significant.
This finding fits inside a larger pattern. Sisk, Burgoyne, Sun, Butler, and Macnamara (2018) ran two meta-analyses on the growth-mindset literature: one on the correlation between mindset and academic achievement (k = 273 effect sizes), one on the effect of mindset interventions on achievement (k = 43). The correlation was small (r ≈ 0.10). The intervention effect was smaller still (d ≈ 0.08), with most of the apparent effect concentrated in low-achieving and at-risk samples and largely absent in the rest. Burgoyne is a co-author on both papers, which is part of why the 2019 piano study reads as a clean test of an a priori hypothesis: the same research group that documented mindset’s weak overall effect on academics tested whether the same pattern held in a controlled motor-cognitive learning task. It did. The full meta-analytic evidence on growth mindset covers the broader literature.
The implication for music education is direct. Telling a beginner that they can learn piano if they have the right mindset is, at best, a motivational nudge to keep showing up. It is not an instructional intervention that accelerates skill acquisition.
The Beginner-vs-Expert Caveat
The result describes the earliest possible stage of piano learning — first-exposure performance after one guided session. The authors are explicit that conclusions about beginners do not transfer automatically to skilled musicians.
Macnamara et al.’s (2014) meta-analysis estimated that deliberate practice accounts for roughly 21% of variance in music performance — substantial, but well short of the 80% sometimes claimed in popular practice-rules-everything writing. The remaining variance is shared among general cognitive ability, domain-specific aptitudes that develop with practice, motivation, starting age, and (probably) factors no current research captures well. As expertise accumulates, the relative contributions shift. Deliberate practice volume becomes more important. Domain-specific cognitive structures (auditory imagery, motor automaticity for common figurations) develop and start contributing variance that did not exist at baseline.
The Burgoyne result is about who picks up a piano piece fastest in the first hour. It is not about who, after a decade of conservatory training, becomes a concert pianist. Both questions are legitimate; they have different answers.
What This Changes for Beginning-Stage Instruction
If early piano learning is bottlenecked by cognitive throughput rather than by mindset, the instructional moves that should pay off are the ones that reduce cognitive load: shorter sequences, fewer simultaneous demands, decoupling rhythm from pitch from dynamics rather than introducing them all at once, and pairing students of similar starting ability so the pace of instruction matches the pace at which the slower learner can keep up.
Mindset interventions still have a role, but it is a different role. They are useful for retention — keeping a discouraged beginner from quitting after a hard week — not for accelerating skill acquisition. Sold as motivational support, they are reasonable; sold as a substitute for instructional adaptation to cognitive load, they are misleading.
The corollary holds for self-directed adult beginners, who are increasingly the dominant population for online piano instruction. The most useful self-assessment is not “do I have the right mindset” but “what is the cognitive load of my current practice unit, and can I split it into pieces small enough to learn before the next one.” Whether learning music in turn raises general intelligence is a separate question with its own evidence base — and a much weaker one than this study’s main result.
What the Study Does Not Establish
Three limitations are worth keeping in mind. First, the sample was undergraduate, restricting the range of cognitive ability — the result might look different at the tails. Second, the dependent measure was performance after a single learning episode, not a learning trajectory across weeks or months. Third, the mindset measure was the standard Dweck self-report scale, which has been criticized for low test-retest reliability and for poorly capturing the construct as Dweck originally articulated it. The mindset null might reflect measurement limitations rather than a true null effect — though the same critique applies to the broader mindset literature, and the convergent meta-analytic evidence makes the null more plausible.
None of these caveats overturn the main result. They scope it: in undergraduates, in a single-session piano-learning task, with standard predictor measurements, general cognitive ability dominates and mindset does not move the needle.
Frequently Asked Questions
Does intelligence determine whether someone can learn piano?
It predicts how fast someone picks it up at the start, not whether they can learn it. Burgoyne et al. (2019) found general intelligence accounted for the largest unique share of variance in beginners’ skill acquisition over a single learning session. Most people can learn piano; the rate at which they do varies systematically with cognitive ability. That is a different claim from “only smart people can learn piano.”
Why didn’t music aptitude predict skill acquisition?
It did at the bivariate level. The unique contribution disappeared once general intelligence was controlled for, because music-aptitude tests in beginners measure something that overlaps heavily with general cognitive ability. Discriminating subtle pitch and rhythm patterns is itself a cognitively demanding task. In beginners, music aptitude and general intelligence are not cleanly separable constructs.
Is the mindset finding really reliable?
It fits the broader literature. Sisk et al. (2018) meta-analyzed the mindset evidence across academic achievement and found a weak correlation (r ≈ 0.10) and a small intervention effect (d ≈ 0.08). The piano result is one more data point in that pattern, not an isolated null. Mindset is not nothing — it appears to matter modestly for at-risk and low-achieving students — but it is not a general accelerator of skill acquisition.
Does this mean deliberate practice doesn’t matter?
No. Macnamara, Hambrick, and Oswald (2014) found deliberate practice accounts for roughly 21% of variance in music performance across the expertise range. That is a substantial contribution. The Burgoyne study examined the first 25 minutes of piano learning, where practice volume has not had time to differentiate students. Practice matters more, and intelligence likely matters less, as expertise accumulates.
What should music educators take from this?
For early-stage instruction, the rate-limiting factor is cognitive load. Reducing simultaneous demands, chunking sequences shorter, and matching pace to the slower learner in a paired session are likely to outperform mindset-framing interventions for beginning skill acquisition. Mindset support has a role in retention — keeping discouraged beginners from quitting — but should not be marketed as a learning accelerator.
References
- Burgoyne, A. P., Harris, L. J., & Hambrick, D. Z. (2019). Predicting piano skill acquisition in beginners: The role of general intelligence, music aptitude, and mindset. Intelligence, 76, 101383. https://doi.org/10.1016/j.intell.2019.101383
- Macnamara, B. N., Hambrick, D. Z., & Oswald, F. L. (2014). Deliberate practice and performance in music, games, sports, education, and professions: A meta-analysis. Psychological Science, 25(8), 1608–1618. https://doi.org/10.1177/0956797614535810
- Sisk, V. F., Burgoyne, A. P., Sun, J., Butler, J. L., & Macnamara, B. N. (2018). To what extent and under which circumstances are growth mind-sets important to academic achievement? Two meta-analyses. Psychological Science, 29(4), 549–571. https://doi.org/10.1177/0956797617739704