Why Doesn’t One Correct Run Prove a Passage Is Learned?

After repeating a passage many times, an adult learner may treat the first error-free run as proof that it is “learned”. More often, it proves only that performance succeeded at that moment. The hands still carry cues from the preceding attempts, and the ear anticipates the spot that was just corrected. Motor-learning research distinguishes immediate performance during practice from retention after a delay. Improvement in the first does not automatically establish the second.

A useful piano test is not an instant repeat. Leave the passage, practise something else, then return a few minutes later from a different preceding phrase. If the opening depends on the momentum of the previous repetitions, or a new entry point causes a halt, the success still belongs largely to that practice context. Being able to recover the intended sound, rhythmic structure and movement relationships after attention has shifted is stronger evidence of retrievable learning. One failure does not prove the passage is unlearned either: fatigue, tension and chance distraction all affect performance. The pattern across separated attempts matters more.

The point is not to invent a magical threshold such as “three correct runs”. It is to change what counts as evidence. Instead of ending with one polished play-through, leave a small retention test: pause, begin from a fresh entry point, and notice where reconstruction is needed. Evaluation then moves from “Did that just feel fluent?” to “Can I rebuild it later?”

When organising practice with <a href="https://apps.apple.com/us/app/scoreflow/id6796205656">ScoreFlow</a&gt;, this delayed single attempt can likewise become the observation point: note the cold start, not only the best run after continuous repetition.

https://bjorklab.psych.ucla.edu/wp-content/uploads/sites/13/2016/07/Soderstrom_Bjork_Learning_versus_Performance.pdf
https://pmc.ncbi.nlm.nih.gov/articles/PMC11237090/