The double-neighbor figure: the tiny five-note pattern at the root of vocal warm-ups

Sing this out loud, slowly, on the syllables do-re-do-ti-do. Then transpose up a half-step and sing it again. Keep climbing.

You have just performed something with at least four different names in four different traditions:

In music theory, it is a double-neighbor figure — an embellishment that begins and ends on a stable tone with one step above and one step below in between.
In Italian bel canto pedagogy, it is one of the basic five-note vocalises — short patterns transposed by semitone through the voice’s comfortable range, the bread-and-butter of Vaccai and Concone.
In Edwin Gordon’s Music Learning Theory, it is a tonal pattern — specifically a tonic-function pattern with the resting tone at its center.
In Kodály-influenced classroom singing, it is a pentachord pattern sung on movable-do solfège.

The convergence is not coincidence. The double-neighbor figure is one of the smallest possible coherent musical gestures that contains all the sub-skills a developing ear and voice need to practice together. Looking at why it keeps showing up — and what each tradition adds to the picture — is a useful lens on how singing skill is actually built.

What’s in this five-note shape, structurally

The pattern (anchor → upper neighbor → anchor → lower neighbor → anchor) does five things at once:

Establishes a stable reference pitch. The first note is the home tone you are circling.
Tests upward stepwise motion. Singing accurately a step up is a distinct skill from singing the same note again — it requires a small, controlled change in vocal-fold tension.
Tests return to home. Coming back to the same pitch you just left is a separate skill: it asks the singer to remember the anchor while having moved away from it.
Tests downward stepwise motion. Going below the anchor recruits different muscular adjustments than going above it.
Tests return to home a second time, from the other direction. This is the consolidation move — the singer must arrive at the same pitch from above that they previously arrived at from below.

Five tasks in five notes, all anchored on one tone. That is dense pedagogical packaging for what looks, on the page, like a trivial little doodle ^[1].

The vocalise tradition: short patterns, chromatic transposition

In the 19th-century Italian school, singers worked through compilations like Nicola Vaccai’s Practical Method of Italian Singing (1832) and Giuseppe Concone’s 50 Lessons. These were not exotic études — they were graduated drills built around short, recognizable patterns: 1-2-3-2-1, 1-3-5-3-1, neighbor figures, simple turns. The crucial pedagogical move was that the student would sing the pattern in C, then in C♯, then in D, and so on, with the teacher transposing at the piano ^[2].

This is the convention in nearly every vocal pedagogy that has been documented:

The pattern is the same; the key changes.
Transposition is chromatic, by semitone, not by cycle of fifths. (Cycle-of-fifths transposition is a piano convention — voice has no fingerings, so the natural axis is the semitone climb through the singer’s range.)
The climb continues until the top of the comfortable range is reached, then descends back through the start to the bottom.

Why this convention? Three reasons. The pattern trains a relationship between scale degrees, not absolute pitches — anchoring it to one fixed key would teach absolute pitch instead of relative pitch. The chromatic climb covers the singer’s whole physiological range, which is the literal point of a warm-up. And it adapts to each individual voice — the comfortable range for a baritone is not the same as for a soprano, so the starting note must be tailored.

What Edwin Gordon adds: the resting tone

Gordon’s Music Learning Theory, developed across the second half of the 20th century, made one observation explicit that the bel canto tradition assumed: a tonal pattern is only meaningful relative to a perceived tonal center. Gordon called that center the resting tone, and the very first skill he asked students to develop was simply hearing and singing it ^[3].

In MLT, before any pattern is sung, the teacher establishes tonality by singing the resting tone or a short cadential figure. Then short patterns — typically two or three notes long — are echoed by the student. Gordon’s pattern sequence introduces tonic-function patterns first (those that begin and end on or near the resting tone), then dominant-function patterns (those built around the fifth scale degree), then mixtures.

The double-neighbor on the tonic — do-re-do-ti-do — sits at the foundation of this sequence. It is unambiguously a tonic-function pattern: every note relates back to the home pitch, and the figure begins and ends there.

Why your ear-training app should care

Two implications for self-directed practice.

First, the double-neighbor figure is a great candidate for a “first sing-back exercise” because it isolates relative pitch from absolute pitch. A learner singing do-re-do-ti-do in twelve keys cannot rely on having memorized the absolute frequency of the start note — the start note keeps changing. What stays constant is the relationship: up a step, back, down a step, back. The skill being trained is the one that transfers to every real piece of music in every key.

Second, transposition is essential, not optional. A sing-back exercise that always starts on the same pitch teaches the singer that pitch and accidentally calibrates them to it. Modern ear-training apps that randomize the starting key per round (EarMaster, Tonedear, ToneGym, the Functional Ear Trainer, Fifths) get this right; ones that don’t are leaving the most important skill on the table.

A note on terminology

If you have come up through classical music theory, you may have encountered the double-neighbor figure described as a “non-chord tone embellishment” — a way to decorate a single chord tone with two passing dissonances ^[1:1]. That is the harmonic role of the figure inside a multi-voice composition.

The melodic role — the standalone gesture, sung in isolation against an established key — is the older one, and it is the role that matters for ear training. When you sing do-re-do-ti-do, you are not embellishing anything. You are practicing the smallest complete melodic statement that requires you to commit to a key, hold a reference pitch, leave it in two directions, and find your way home twice.

That is what makes it a good five seconds of practice.

References

Hutchinson, R. Music Theory for the 21st-Century Classroom — chapter on Double Neighbor figures. https://musictheory.pugetsound.edu/mt21c/DoubleNeighbor.html. See also LibreTexts: Equipping the Musical Ear — Neighbor Tone. https://human.libretexts.org/Courses/Sierra_College/Equipping_the_Musical_Ear/13:_Melodic_Embellishments/13.03:_Neighbor_Tone ↩︎ ↩︎
Vaccai, N. (1832/2000). Practical Method of Italian Singing, ed. John Glenn Paton. G. Schirmer. ISBN 9780793551200. See also Library of Congress, NLS Music Notes: Classic Italian Vocal Exercises: Oldies but Most Definitely Goodies (2022). https://blogs.loc.gov/nls-music-notes/2022/05/classic-italian-vocal-exercises-oldies-but-most-definitely-goodies/ ↩︎
Gordon, E. E. (2012). Learning Sequences in Music: A Contemporary Music Learning Theory. GIA Publications. See also the GIML overview of Music Learning Theory: https://giml.org/mlt/audiation/ and Sequencing Pattern Instruction in a MLT Classroom: Tonal Edition — SingtoKids. https://singtokids.com/sequencing-pattern-instruction-mlt-classroom-tonal-edition/ ↩︎