What is harmonic dictation?
Harmonic dictation is writing down music in more than one part: the bass line, the melody above it, and the voices in between. like this one is four melodies sounding at once, and also a I–IV–V–I progression. Melodic dictation captures a single voice, and harmonic dictation captures several, including how they line up into chords.
Harmony can be heard vertically, as a series of chords with qualities and functions, or horizontally, as separate lines moving through time. But knowing the progression is I–IV–V–I doesn't say which chord tone the soprano sings, and hearing the soprano and bass clearly doesn't tell us the inner voices that complete the chords.
How do we hear more than one voice at once?
The auditory system sorts a mixture of simultaneous sounds into separate streams, grouping notes that are close in pitch, start together, move together, and share a timbre (Bregman, 1990). So, two voices are heard separately when they maintain enough independence. Otherwise, they fuse. tend to blend into a single hollow line, but voices that are further apart, like , remain distinct. Voice-leading rules are, among other things, instructions for keeping voices audible as separate voices. Huron (2001) showed that the traditional rules of voice leading (avoid parallel fifths and octaves, avoid crossed voices, and keep each voice close to its previous note) follow from the same perceptual principles.
| Motion | What the voices do | Listen |
|---|---|---|
| Parallel | Move in the same direction by the same interval | |
| Similar | Move in the same direction by different intervals | |
| Contrary | Move in opposite directions | |
| Oblique | One voice holds while the other moves | |
| Parallel fifths | Move together a perfect fifth apart, and tend to fuse | |
| Parallel octaves | Move together an octave apart, and fuse almost completely |
Why are the outer voices easier to hear?
When multiple melodies play at once, a change in the highest one produces a larger automatic response in the brain in both musicians and nonmusicians than the same change in lower ones (Fujioka et al., 2005). The top voice benefits from this built-in pitch advantage, and that's part of why the highest voice so often carries the melody. Listeners also detect timing deviations better in a lower voice than in a higher one (Hove et al., 2014), and the lowest note of a chord determines its inversion and usually its root. Consistent with both, listeners notice changes to a melody more readily in the outer voices than in the inner ones (Palmer & Holleran, 1994).
Why do inner voices disappear?
Listeners asked to count the voices in a texture of similar timbres are accurate up to about three and increasingly undercount beyond that (Huron, 1989), and the voices most likely to go missing are the ones in the middle. An inner voice has neither the top voice's advantage in pitch nor the bass voice's advantage in time and harmony, and it sounds between the voices that do.
Inner voices also typically move as little as possible by design, holding common tones and stepping to the nearest note of each new chord. That smooth motion stays coherent as a line, but nothing in it draws attention, so an inner voice recedes behind the outer voices unless we listen for it deliberately.
How does voice leading make inner voices predictable?
Listening for an inner voice is much easier when we know roughly what to expect, and the same conventions that keep voices separate also make them predictable. In common-practice voice leading, a tone shared by two consecutive chords is usually held in the same voice, the other voices move to the nearest available chord tone, and tendency tones resolve. So, the leading tone rises to the tonic, and the seventh of a dominant seventh chord resolves down. Once the bass and the chords are known, the inner voices are heavily constrained, and a plausible guess can be checked against what our ears actually hear.
In a , the tenor holds G through the V–I, the alto steps from D up to E, and the soprano's B, the leading tone, resolves up to C. None of those motions had to be heard as isolated notes, because each is what the chord change implies. Whether it's better to hear inner voices directly or to infer them from the harmony and confirm by ear is a matter of teaching tradition rather than evidence, and strong dictation usually does both anyway.
How many voices are needed for a harmony?
Strictly speaking, harmony requires at least two pitches sounding together. In practice, an orchestra might double a set of pitches across dozens of parts, and even a single unaccompanied melody can strongly imply harmony through its contour and arpeggiation.
Harmonic dictation is usually practiced in two, three, or four voices. The outer voices, referred to as soprano and bass, define the harmonic framework. Three voices add an inner voice, enough to sound every tone of a triad, though three-part writing often leaves out the fifth so each line can move smoothly. Four voices, the traditional chorale texture, allow for doubled tones in triads or fully voiced seventh chords.
| Texture | Voices | Listen |
|---|---|---|
| Two voices | Soprano and bass | |
| Three voices | Soprano, one inner voice, and bass | |
| Four voices, major key | Soprano, alto, tenor, and bass | |
| Four voices, minor key | Soprano, alto, tenor, and bass |
Why do we make octave errors?
An octave error identifies the correct pitch class but places it in the wrong register. This is one of the most common mistakes in harmonic dictation because notes operate on both their pitch height and scale degree. For example, a listener fixated on recognizing a note's function in the scale might misjudge or ignore its physical height. This ambiguity happens most often in the bass and inner voices, where overlapping registers make it plausible for a single note to belong to multiple parts.
It's tempting to assume notes an octave apart simply sound like the same note, but directly perceiving this equivalence can be difficult for listeners, though it improves with training (Schmidt & Williams, 2025). So octave errors come less from octaves sounding identical than from listening for a note's function instead of its height. We might hear and as the exact same sequence and approximate their placement on the staff. When we fail to track this height, spacing conventions offer a structural safety net to rule out impossible voice placements, but relying on conventions does not effectively substitute for hearing the physical height of the notes.
How does a trained ear take harmonic dictation?
The plan of hearings from melodic dictation carries over. The first hearings focus on the soprano and the bass, which form the frame everything else is built on. A bass line that is heard and understood in its key implies the chords above it, and the harmony can be worked out from the bass and confirmed by ear. The inner voices come last, predicted from the voice leading and checked against the sound. Hearing and reasoning complement each other throughout.
Many listeners find that singing one voice of a chorale while playing another, then swapping which is sung and which is played, trains the ear to hold one line steady while a second sounds against it. A progression can also be learned from the bottom up by playing each chord as a close-position arpeggio from the bass, singing its tones from the bass up and back down, then singing the whole progression while playing only the bass notes. Once a dictation is corrected, playing or writing it again in another key without looking at the page shows whether the harmony was understood or only transcribed. These are conservatory practice habits rather than tested methods, but each one trains the ear to hear several lines at once and understand them as harmony.
Harmonic dictation in real music
Harmonic dictation is how a choir part is written down from a recording, how a string quartet's inner lines are transcribed, and how a pianist figures out voicings under a melody. In real music, voices drop in and out, double at the octave, and get split between instruments. The texture is rarely a neat four-voice chorale, but the same frame still applies.
Harmonic dictation draws on nearly every other listening skill at once: chord quality, chord progressions, scale degrees, and single-line dictation in both melody and rhythm. Our ear training hub maps how those skills fit together and what perceptual learning research says about building them.