Bridges to traditional music theory

Spatial composition extends, departs from, and occasionally refuses traditional Western music theory. Here is where the analogy holds and where it breaks, useful both for musicians arriving with a theory background and for educators considering whether this fits a music curriculum.

Spatial composition is not a replacement for music theory. It is an additional lens. Most of what classical and jazz pedagogy teach about voice leading, contour, harmonic function, and counterpoint maps onto things you can do in a graph; sometimes one-to-one, sometimes with a translation. Where the analogy breaks down, it tends to break in interesting directions: spatial composition has vocabulary for things traditional theory does not name.

Voice leading lives in harmony nodes

A harmony node with the nearest chord tone above or nearest chord tone below transform is doing voice leading on the fly. When the chord changes, the modifier moves the incoming pitch by the smallest interval that lands on a tone of the new chord, which is one core voice-leading principle that classical theory teaches: keep melodic motion smooth between chord changes. (Classical voice-leading is a deeper body of rules than this: parallel-fifth avoidance, soprano/bass voicing, leading-tone resolution, and so on, but smallest-interval-motion is the foundational instinct, and the platform implements it directly.) Set up two harmony branches with different nearest-tone transforms and you have automatic two-voice writing.

Counterpoint becomes branching topology

Independent voices that share a harmonic center are exactly what Splitters and parallel branches produce. A single Source feeding two Harmony branches at different intervals—say third above and fifth below—produces three-voice texture from one origin. The branches stay independent because each transform operates on its own signal.

Contour theory drives the agent

Cognitive musicology distinguishes melodic contour—the up-and-down shape—from absolute pitch. The ComposerAgent exposes nine contour modes: free, ascending, descending, oscillate, gravity, motif, sequence, pedal, and wave. Ideas that once lived in an analysis textbook become controls you can hear on a Source.

Harmonic focus speaks in scale degrees

The agent's harmonic focus uses familiar vocabulary: chord tones, scale tones, color tones, pentatonic, root focus, and guide tones. The third and seventh—the guide tones—are the harmonic skeleton bebop players learn to outline. Set a Source to guide tones with a gravity contour and the melody begins to lean into that language.

Key, scale, and mode are first-class

Western tonal music is held together by a few shared decisions: a root note, a scale, and a mode. Sources, Chord Nodes, and Harmony transforms read the same global settings. Change C to F and the graph transposes coherently. Switch major to harmonic minor and procedural voices re-quantize on the next signal. Choose Dorian or Mixolydian and the network shares the modal vocabulary without making you re-spell every chord.

The supported scales reach beyond the major / minor pair: chromatic, dorian, phrygian, lydian, mixolydian, harmonic minor, melodic minor, pentatonic major and minor, blues, whole tone, plus presets that capture more specific colors. The point is not exhaustive coverage: it is that the system speaks the same language as a music theory student. Mode shifts in spatial composition are the same operation a theorist would describe in writing.

Beyond twelve-note Western defaults

The current platform is strongest when it speaks the language of Western tonal theory: scale degrees, Roman numerals, chord function, and 12-tone equal temperament. That is useful, but it is not the whole world of music. Spatial composition should become a tool for exploring non-Western harmonic systems, non-Western scales, tuning systems outside 12-TET, and pitches that sit between the notes on a piano.

This work has to be done carefully. Traditions such as raga, maqam, gamelan tuning, and many other bodies of practice are not just alternate presets. They carry their own theory, pedagogy, instruments, history, and cultural memory. The goal is to make the platform wider without flattening that heritage into surface color. Expect updates on this front in the coming months as the pitch and harmony model grows beyond Western defaults and becomes a compositional tool that can respect more cultural disciplines.

Chord progressions, generated and manual

Above the scale layer sits the chord progression: the sequence of harmonies that gives a piece its emotional arc. Spatial composition supports two modes for this. Manual progressions let a creator set a Roman-numeral sequence such as I, IV, V, or vi and lock the progression in place. Generated progressions use a Markov chain over tonal functions, then bias the result with tonic influence, dominant influence, and deceptive-cadence chance. The generated mode is built on functional-harmony logic that classical and jazz pedagogy teach: tonic / predominant / dominant motion, leading tones pulling toward resolution, and V to I as the strongest cadence.

Every chord node and every procedural agent reads the progression as it advances. When generated mode lands on a deceptive cadence and resolves V to vi instead of V to I, voice-leading transforms move to the new target. Chromatic applied dominants such as V/V need a richer chord type than the current scale-degree model, so that work belongs to the next stage of the harmony system rather than the current generator.

Form lives in song structure

Song structure defines named sections—intro, verse, chorus, bridge, outro—with measure counts and per-section properties. Boundaries become triggers: Sources can fire on sectionStart, sectionEnd, or sectionCycle. A hit can land on every chorus entry, while an instrumental section swaps in its own progression.

Form becomes a first-class object, not a long sequence of mute decisions. The chorus topology can wake when the chorus arrives. An intro Source can fire once and stay quiet for the rest of the piece. Form is structure, structure is topology, and the graph reads them together.

How composer agents connect to Western harmony

The composer agent is not a black box generator. Every parameter on it maps to a concept Western theory already names. Contour maps to melodic-contour analysis. Harmonic focus maps to chord-tone vs scale-tone vs color-tone hierarchy. Voicing style maps to standard chord voicings (close, drop2, shell, quartal, rootless). Voice leading bias and leap recovery shape melodic connection. Beat targets and approach style support walking lines and chromatic approaches. Phrase behavior supports cadences and one-agent call/response. Swing amount maps to jazz swing.

A music theorist analyzing a spatial composition has a vocabulary for nearly every choice the agent is making. A composer setting up an agent is making the same kinds of choices a theory-trained writer makes: just at the level of parameter rather than the level of note. The generated music is tonal, voice-led, contour-shaped, and metrically swing-able exactly because those traditions are encoded in the parameter space. The procedural and the traditional are not in tension; the procedural is the traditional, expressed as a tunable surface.

Where the analogy ends

Some things in spatial composition have no clean parallel in traditional theory. Cycles are not fugue subjects or canons; a signal loops back through the graph and keeps firing until the feedback budget stops repeated routing. Probability splits are not fixed score and not improvisation in the traditional sense; they are parametric arrangement. Audio reactivity has no analog in the score-based tradition at all because the graph can respond to the loudness of the music it just produced. Spatial composition needs new vocabulary here, and that vocabulary is one of the things this knowledgebase exists to develop.

The bridge is bidirectional. Students of spatial composition can speak both languages. Theorists who engage with spatial composition gain a new analytical lens, one that takes seriously the idea that topology is structure and that some musical phenomena are easier to describe as processes on a graph than as events on a timeline.