There are only a handful of things in life that have the ability to cross social, generational, cultural, and economic boundaries. Music is one of them! Whether you are on stage in a symphony hall, a sold-out concert arena, or in the bathroom shower, there is likely a musician in all of us.

But what separates the true musical geniuses from us shower-screeching professionals? Maybe it’s the ability to discriminate between pitches and the ability to identify musical intervals. A music interval refers to the distance in pitch between two tones and is a fundamental component of all musical structures.  Melodies and harmonies can be thought of as sequences of underlying intervals. Given the importance of recognizing musical intervals, there are a myriad of tutorials, classes, schools, and workshops dedicated to teaching and perfecting interval identification.

Here is one of them:

Interestingly, to a cognitive psychologist, musical-interval learning is merely a complex categorization task: the ability to hear two successive tones and assign the couple to the appropriate musical label (e.g., minor second or perfect fourth). Evaluating the best training procedures to improve musical-interval learning is therefore a question of auditory cognition.

In a recent article in the Psychonomic Society’s journal Attention, Perception, and Psychophysics, David Little, Henry Cheng, and Beverly Wright sought to investigate how different training regimes influence musical-interval perception depending on the configuration of the training blocks.

Specifically, Little and colleagues compared the effectiveness of four different three-day training regimes on a musical-interval comparison task: all-practice regime, a practice + exposure regime, a practice + silence regime, and an exposure + silence regime. Each of the four training regimes consisted of various combinations of three different block types: practice blocks, exposure blocks and silence blocks (see figure below).

During practice blocks, participants completed a musical-interval comparison task in which they listened to two musical-intervals and were asked to decide which interval was associated with the appropriate interval label (e.g., a “fourth”).

During exposure blocks, participants passively listened to musical-intervals while they completed a visual distractor task where they matched symbols to numbers on the screen. Importantly, during the exposure trials, participants did not make judgments about the intervals they heard.

During silence blocks, participants completed the same musical-interval comparison task used during the practice blocks, except they were not presented with the tones. They were told that they would not hear any sounds and were instructed to ‘guess’ which interval corresponded with the given interval label, even though they never heard the tones. They were also given feedback on their performance similar to the practice blocks.

The training experiment was carried out over four days and comprised four phases, a familiarization period, a pretest, training, and a posttest. The familiarization, pretest, and first training regime were carried out on day 1. The next two training regimes were carried out on day 2 and day 3. The posttest was carried out on day 4.

During the familiarization period, participants learned about the concept of musical intervals. Using schematic diagrams, participants learned that the musical interval referred to the relationship (i.e. the ratio) between two tones played in succession. They were told that the ratio of the tones was the critical feature to focus on. They were then given verbal instructions for the musical-interval comparison task. After hearing the instructions, they completed two sets of training trials (i.e. two blocks of practice trials).

During the training phases, listeners were divided into four different groups, one for each of the four training regimes (i.e. all-practice, practice + exposure, practice + silence, exposure + silence). Importantly, for a single participant, the training regime remained the same across all training sessions.

The pretest and posttest both consisted of several blocks of practice trials. The only difference was that the posttest also included identification of additional musical intervals (i.e. triangle tones) that were not used during training. The pretest only required identification of pure musical intervals.

There were a number of interesting results from the musical-interval training experiment. Here we note a few. First, at the group level, the practice + exposure training was the only regime to yield learning (as evidenced by an increase in accuracy) between day 1 and day 4 of training. Participants in the practice + exposure group were 69% accurate at day 1 and 88% accurate at day 4. None of the other three training regimes resulted in significant improvements training across days.

Although there was no significant difference in accuracy across training groups at day 1, by day 4, the practice + exposure group was roughly 16% (or higher) percentage points greater than the other training groups.

Lastly, accuracy for the untrained intervals (e.g. the triangle tones) during the posttest was better for the practice + exposure group compared to the other three training groups. (See figure below. Panel A shows average accuracy across days for each training group. Panel B shows individual subject performance).

The moral of the story, is that for fine-grain discrimination, like the musical-interval task used in this study, it is not just practice that makes perfect (or improves learning), but practice plus passively experiencing the stimulus (exposure) that best improves performance.

There truly is something to be said for not just actively practicing, but also sitting back periodically and taking it all in.

Featured Psychonomic Society article:

Little, D., Cheng, H. & Wright, B. (2018). Inducing musical-interval learning by combining task practice with periods of stimulus exposure alone. Attention, Perception and Psychophysics, DOI 10.3758/s13414-018-1584-x.