Before reviewing the authors, a brief review of a 1986 ‘what do we know about learning music’ report. (Overpriced at Amazon; your library can get you a copy.)
- Divided practice sessions are more effective than one long session to learn and retain
- Practicing with hands together is more efficient than hands alone (Matthews disagrees)
- Score analysis/study before practice improves learning efficiency (Viloteau agrees vigorously)
- Once something is learned, only a little repetition is needed for maintenance (assumes both high savings and high retrieval)
- Visualizing a score with eyes closed improves memorization
- Familiar elements in new music aid memorization
- Awareness of related or altered patterns in music aids memorization (means, e.g., this section carries the melody through six diminished chords in different inversions, circle of fifths descending, ah, so it’s just a bunch of diminished chords moving in alternation down the neck, I know how to do that)
- Compositional and structural awareness aid memorization (which seems a ‘partner’ to nos. 3 and 6; means, e.g., knowing this section is a III – VI – II construct)
No. 1 is spacing. Spacing feels more frustrating and less productive, but for many things, music included, it’s more effective. To be a classical guitarist requires a high level of persistence. This is necessary but also a risk. Hammering at something until you get it right is likely less effective than practicing twenty minutes, taking a break, practicing something else, taking a break, and perhaps coming back (or even not coming back until the next day or even the next week). The insistence of most (all?) of the authors on breaks is well-supported by research on spacing. It probably needs emphasis because our persistence (stubbornness) carries a bias to ‘just keep doing it until we get it right.’ Massing, that is, rather than spacing.
On #5, score-reading and visualization, learning researchers distinguish savings (how deeply you know something) and retrieval (whether you can recall it at will). This recommends we memorize a picture of the song. Presumably since the written score is so completely orthogonal to the physical and musical acts of playing, this is a method of reinforcing retrieval. Aided, one thinks, by combining both sensual (visual) and semantic (score content) aspects. Opportunity cost is not mentioned: if visualizing a score takes away from other practice, is learning improved?
(Is there any other discipline where one memorizes a visual image of semantic content? Mathematicians don’t memorize pictures of equations, nor actors a picture of the script, I don’t believe.)
Remarkable for its absence: knowing a song, in the simplest, most quotidian way. Being able to sing or whistle it accurately. If it came on the radio, knowing just how it goes and where it’s going. Score analysis may improve learning efficiency, but it’s hard to believe that hearing and remembering the song is so unimportant as to not merit mention. Score analysis is, after all, constructing an analytic description of a representation of a song. That’s two levels of indirection from the actual music, and omits the embodiment, which is essential. Perhaps for those who can instantly translate a score into accurate mind’s-ear musical tones? An impressive skill, but it’s hard to imagine this is a requirement for elite playing; certainly it’s not for amateurs, and every level of indirection adds cognitive burden.
As remarkable, the failure to discuss goals and self-testing. It’s possible discussion of goals was ‘invisible’ to the symposium because everyone in the population with which they were concerned—college/university music students—does it. (Or have teachers set goals for them.) There are gradations in skill at setting goals at multiple levels: someone might be good at high-level goals and yet be poor at decomposing them, for instance; the opposite is often seen as well. There are similar or even greater gradations in skill at self-assessment. There’s substantial literature demonstrating that self-testing is a powerful positive influence on learning. This is a gaping hole in the Symposium’s findings.
The Crane Symposium summary makes an implicit assumption of steady forward progress. Biological systems, as with most complex nonlinear systems, work in thresholds and plateaus. While, for example, the Royal Conservatory and London College of Music provide graded steps that are continuous, as typically are methods, we learn through (for lack of better terms) disorganization into organization, struggle followed by sense-making. The continuous-improvement-to-the-edge-of-capacity view biases practice toward technical proficiency, because it is much more easily measured than musical understanding and expression. We can come to value the subgoal over the goal.
Finally, the product of the Symposium assumes a constant supply of motivation. (Plus presumably a rigorous discipline to bridge gaps in motivation.) A different way of looking at this is, what elements in the structures of learning nourish motivation? What deplete it?
The short version of the Crane Symposium is, spacing is better than massing, and the more paths to the song (score, structure, similarity to other songs, &c), the more reliable both the encoding and the retrieval.
Learning music on guitar 