Swimming in water
Educational culture sometimes behaves as if it were enough to read the right book and understand the right formula. Knowledge really can be got from a text; skill appears only in action.
You cannot learn to swim by watching a pool with a well-organised set of notes. You cannot master writing by only reading advice for writers. You cannot learn to speak in public while your speeches happen exclusively in your own head, where the audience is disciplined and applauds in the places provided.
Practice at first often reduces performance.
A new way of typing is slower than two fingers. A new tool makes you think about every movement. Automating one task takes longer than doing it by hand. This creates the temptation to go back to the familiar method, which today is faster.
What has to be compared, though, is not today’s speeds but the total cost of a series of repetitions.
If the skill will be needed once, a long period of learning may not pay off. If the action lies ahead hundreds of times, a temporary loss of speed becomes an investment.
Deliberate practice differs from mere repetition: the task slightly exceeds your current level, the result can be checked, the mistake becomes visible, there is feedback, difficulty rises gradually and the skill is applied under varying conditions.
It helps to train separately the element that is holding back the whole job: touch typing, mental arithmetic, reading diagrams, pronunciation, handling a tool, searching documentation.
At the same time training must not turn into permanent preparation. Sometimes a person spends years improving their starting technique in order not to start the race.
A moment comes when you have to get into the water.
Practice becomes deliberate when it trains separately the very element that keeps making you go around the difficult spot.
A musician skips the awkward transition. A driver picks a route without parking. A specialist does the familiar part of the work well and hands colleagues everything they feel unsure about. Spoken language develops up to the point where you have to talk to a live person, after which a great interest in grammar tables suddenly appears.
A weak link is rarely strengthened by general repetition. It has to be singled out and trained on its own.
For that it helps to:
1. name the difficulty precisely;
2. slow down;
3. break the action into elements;
4. repeat the problem element;
5. get feedback;
6. put it back into the whole;
7. raise the difficulty gradually.
Not everything unpleasant has to be trained. Dislike may be telling you that the action is unimportant, badly organised or should not be done by this person at all. But if the skill stays incomplete without it, avoidance only fixes the boundary in place.
A professional is not distinguished by loving every part of the job. They know how to do the hard part without negotiating with their own mood.
Repetition without feedback entrenches what is already there. Someone who has typed with two fingers for twenty years has accumulated enormous experience of typing with two fingers. A driver with thirty years behind the wheel parks confidently in the one way they learned in their second week. Experience here is measured not in years but in the number of occasions when it became visible that the result was worse than it could have been. Someone with long service may have had no such occasions at all.
So the most expensive hours of practice are the first ones after changing method. They look like getting worse, they feel like humiliation, and they are paid for long before any result appears.
Automatism is needed most where the real situation will leave no time to think: an emergency procedure, first aid, driving, public speaking, a complex movement in sport.
Sources and further reading
Ebbinghaus, H. (1885). Über das Gedächtnis. Duncker & Humblot.
Roediger, H. L., III, & Karpicke, J. D. (2006). Test-enhanced learning: Taking memory tests improves long-term retention. Psychological Science, 17(3), 249–255. https://doi.org/10.1111/j.1467-9280.2006.01693.x
Chi, M. T. H., Bassok, M., Lewis, M. W., Reimann, P., & Glaser, R. (1989). Self-explanations: How students study and use examples in learning to solve problems. Cognitive Science, 13(2), 145–182. https://doi.org/10.1207/s15516709cog1302_1
Wason, P. C. (1960). On the failure to eliminate hypotheses in a conceptual task. Quarterly Journal of Experimental Psychology, 12, 129–140. https://doi.org/10.1080/17470216008416717
Ericsson, K. A., Krampe, R. T., & Tesch-Römer, C. (1993). The role of deliberate practice in the acquisition of expert performance. Psychological Review, 100(3), 363–406. https://doi.org/10.1037/0033-295X.100.3.363
Adams, D. (1979). The hitchhiker’s guide to the galaxy. Pan Books.
Autio, C., Schwartz, R., Dunietz, J., Jain, S., Stanley, M., Tabassi, E., Hall, P., & Roberts, K. (2024). Artificial intelligence risk management framework: Generative artificial intelligence profile (NIST AI 600-1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.AI.600-1
OECD. (2026). OECD digital education outlook 2026: Exploring effective uses of generative AI in education. OECD Publishing. https://doi.org/10.1787/062a7394-en
Cruces, G., Fernández Meijide, D., Galiani, S., Gálvez, R. H., & Lombardi, M. (2026). Does generative AI narrow education-based productivity gaps? Evidence from a randomized experiment (NBER Working Paper No. 34851). National Bureau of Economic Research. https://doi.org/10.3386/w34851
Part III