One of my longest-standing research interests is in ear training pedagogy. Conventional wisdom says that the point of ear training class is not to train students’ ears in the classroom, exactly—everyone realizes that the majority of work must be done outside of class (though it’s hard to convince students that this is a good use of their limited time when ear training is a 1-credit-hour course)—rather the point is instead to give students methods for practicing. But what also takes up a large portion of class time is assessment. This assessment comes in the form of solo singing in front of their peers, and dictation tests. Both of these assessment activities are not ideal, as most pedagogues will admit. So then why do they persist? I think mostly practical reasons, involving who teaches ear training and what’s simplest to do in the classroom, which I don’t want to get into here.
Although I’m not teaching ear training this semester, and I’m not currently in a position to try out any of this, I often fantasize about my ideal ear training set up, which would attempt to improve on the typical assessment tools. My own successful ear training journey had almost nothing to do with singing quizzes or dictation exams (though being compelled to learn moveable-do solfège in college undoubtedly was an improvement over my previous ad hoc methods, even if I was too stubborn to recognize this at the time). I learned mostly on my own, by learning to play songs by ear.
In light of this, I would like to eliminate dictation tests in favor of transcription work. Transcription teaches many of the same skills as dictation: connecting sound with notation, and synthesizing pitch/rhythm/harmony relationships. Transcription has the advantage of affording more freedom to the student to manipulate the recording as they see fit.
Each student will experience different challenges in creating the transcription. Letting them control their “hearings”—the length and the quantity of them—allows each student to tailor their experience to their own needs. Students can be taught to do this in their own minds during a dictation exam, of course—this is often referred to as developing musical memory—but this is a highly advanced skill, one that maybe shouldn’t be taught before students really understand how to connect sound and notation. In my experience, developing musical memory will come organically while students work on transcription. As they try to become more efficient with their transcriptions, they will learn to memorize larger chunks.
Further, in a transcription assignment, the student is free to check themselves as they go along using any instrument of their choice—piano, guitar, piccolo, marimba, whatever. This makes the whole process more relatable to performing on their instrument. More importantly, this teaches students to self-correct rather than rely on the teacher to tell them what was wrong.
I would also incorporate learning to play songs by ear. This could involve both full harmonic textures, in which case students must perform on the piano or guitar, as well as melodic textures, which could be performed on any instrument.
The primary issue with transcription assignments is the problem of cheating, i.e., copying from a published score. Some cheating is probably inevitable, but here are some ways I would work against it:
- De-incentivize cheating through pass-fail grading, or other generous grading, and the opportunity to revise. Lowering the stakes of homework assignments often helps reduce cheating.
- Choose obscure pieces, withhold information, make recordings myself, etc. to make finding a score more difficult.
- Make at least a portion of the transcriptions from popular songs or other music that does not rely on notation, rather than art music. This would also serve to expose students to alternative repertoires.
I would also tentatively consider the possibility of a transcription “exam,” administered in a music lab so that students could access resources (keyboard, headphones, notation software, etc.), which could determine whether the student passes or fails the course. This would help ensure that students were not unethical in their submission. I have concerns, though, about having such high stakes riding on a single activity.
I recently gave a transcription assignment in my Analysis of Post-1950 Popular Music seminar, a graduate course. While many students agreed that the assignment was difficult, many students (many of the same students!) also agreed that the assignment was very valuable. Students reported totally losing track of time during the assignment because they were so interested and engrossed in the work, and they found it valuable to have to make a final decision about the best way to represent ambiguous cases.
Transcription is a better assessment tool than dictation because it is a more holistic approach to teaching the skill of connecting sound to notation. Transcription involves “real” music, with lots of rhythmic/textural variation. Also, in my experience as a professional musician, transcribing is an important job skill: music educators, composers, music therapists, musicologists, and performers alike could all potentially need to transcribe something in their careers. In this situation, they would have access to all kinds of resources, like online chords, peer collaboration, instruments, notation software with playback, and so on. Why restrict them to 6 hearings, no instruments, and even no humming? This is like transcription on extra-hard mode, and it’s not the most efficient way to teach people ear training, even if it may be the most efficient way to grade it.