#3662 Music Theory Knowledge Graph

#3662 Music Theory Knowledge Graph

#3662 Provide a wide range of entities and relationships for a music theory knowledge graph

Here’s a comprehensive menu of entities and relationship types you can mix-and-match to design a rich music-theory knowledge graph. I’ve grouped things so you can adopt only what you need and still keep the schema coherent.

Core pitch & harmony

Entities

• Note, Pitch, PitchClass (C, C♯/D♭…), Octave, Register

• Interval (m2, M3, P5…), CompoundInterval

• Scale (major, natural minor, harmonic minor, melodic minor, pentatonic, whole-tone, chromatic), Mode (Ionian…Locrian; Dorian ♯4, etc.)

• Chord (triad, seventh, extended, altered), ChordQuality (major, minor, diminished, augmented, dominant), ChordTone (root, 3rd, 5th, 7th), Extension (9, 11, 13), Alteration (♭9, ♯11, ♭13)

• Key, KeySignature, Tonality (major/minor/modal), Tonic/Dominant/Subdominant (HarmonicFunction)

• Cadence (PAC, IAC, HC, PC, deceptive), Progression (I–vi–IV–V, ii–V–I…), RomanNumeral, FiguredBass

• Modulation, PivotChord, SecondaryDominant, BorrowedChord (mode mixture), Neapolitan, AugmentedSixth (It/Fr/Ger), TritoneSubstitution

• ChordScale (e.g., Mixolydian for V7), GuideToneLine

Relationships (examples)

• PITCHCLASS enharmonic_equivalent_of PITCHCLASS

• NOTE has_pitch PITCH ; PITCH in_octave OCTAVE

• INTERVAL between NOTE ↔ NOTE ; CHORD built_from INTERVAL

• SCALE consists_of PITCHCLASS ; MODE derived_from SCALE

• CHORD contains CHORDTONE ; CHORD has_quality CHORDQUALITY ; CHORD has_extension EXTENSION ; CHORD has_alteration ALTERATION

• KEY uses KeySignature ; KEY has_tonality TONALITY

• CHORD diatonic_to KEY ; CHORD borrowed_from MODE

• ROMANNUMERAL analyzes CHORD ; FIGUREDBASS realizes CHORD

• CHORD resolves_to CHORD ; PROGRESSION orders CHORD

• FUNCTION assigned_to CHORD (e.g., Dominant → V7)

• CADENCE concludes PROGRESSION ; CADENCE type_of cadence_class

• SECONDARYDOMINANT targets CHORD

• AUGMENTEDSIXTH resolves_to DOMINANT

• MODULATION from KEY → to KEY ; PIVOTCHORD pivots_between KEY ↔ KEY

Rhythm, meter & form

Entities

• TimeSignature, Meter (simple/compound/complex), Beat, Subdivision, Tempo, TempoMarking, Pulse, Polyrhythm, Tuplet, Syncopation, Hemiola

• Articulation (staccato, legato…), Dynamics (pp…ff), Accents

• Motif, Cell, Figure, Phrase, Period, Sentence, Theme

• Form (Binary, Ternary, Rondo, Sonata, Strophic, Verse–Chorus, 12-bar Blues), Section (Intro, Exposition, Development, Bridge, Coda), RehearsalMark

Relationships

• TIMESIGNATURE defines METER ; METER groups BEAT

• TEMPO measured_in BPM ; PHRASE has_motif MOTIF

• SECTION part_of FORM ; PHRASE contained_in SECTION

• RHYTHMIC_EVENT uses Subdivision/Tuplet

• ACCENT applied_to NOTE/CHORD ; DYNAMICS applied_to SECTION

• HEMIOLA occurs_over TIMESIGNATURE

• MOTIF transforms_to MOTIF (augmentation, diminution, inversion, retrograde)

Voice-leading & counterpoint

Entities

• Voice (Soprano, Alto, Tenor, Bass… or any instrument line), Texture (monophony, homophony, polyphony), SpeciesCounterpoint (1st–5th), VoiceLeadingRule (e.g., avoid parallels), TendencyTone (leading tone, 4̂ in V7), Resolution

Relationships

• VOICE performs NOTE ; VOICE leading_to NOTE

• TENDENCYTONE resolves_to NOTE

• VOICELEADINGRULE constrains VOICE_PAIR

• TEXTURE realized_by VOICE_SET

• NONCHORDTONE type_of (passing, neighbor, suspension, retardation, appoggiatura, escape, anticipation, pedal)

• NONCHORDTONE resolves_to CHORDTONE ; prepared_by NOTE ; suspended_from NOTE

Orchestration, instruments & timbre

Entities

• Instrument, Family (strings, winds…), Range, TransposingInstrument, Clef, Staff, NotationSymbol, Technique (pizz., sul pont., flutter-tongue…), TimbreDescriptor (bright, dark), OrchestrationPattern

Relationships

• INSTRUMENT belongs_to FAMILY ; INSTRUMENT has_range RANGE

• INSTRUMENT notated_in CLEF ; INSTRUMENT sounds_at TRANSPOSITION

• TECHNIQUE applicable_to INSTRUMENT ; TECHNIQUE affects TIMBRE

• NOTE realized_as NotationSymbol ; PART written_on STAFF

Tuning, acoustics & systems

Entities

• TuningSystem (12-TET, just intonation, meantone, Pythagorean), Temperament, A4Reference, Cent

• Fundamental, Partial (Overtone), HarmonicSeries, Spectrum, Formant

• Tonnetz, CircleOfFifths, VoiceLeadingGraph

Relationships

• TUNINGSYSTEM defines INTERVAL_RATIOS

• PITCH has_frequency Hz ; FUNDAMENTAL has_partial PARTIAL(n)

• KEY adjacent_on CircleOfFifths KEY ; CHORD neighbors Tonnetz CHORD

• PITCH distance_in_cents CENT

Analysis frameworks (classical, jazz, set theory, world)

Entities

• SetTheory: PitchClassSet, ForteNumber, SetOperation (Tn, In), ZRelation

• SchenkerianLevel (foreground/middleground/background), Urlinie, Bassbrechung

• JazzTheory: ChordSymbol, SlashChord, UpperStructureTriad, Lick, ImprovisationPattern, GuideToneLine

• World systems: Raga, Arohana/Avarohana, Pakad; Tala; Maqam, Jins; GamelanScale (slendro/pelog)

Relationships

• PITCHCLASSSET labeled_by FORTE_NUMBER ; PITCHCLASSSET related_by SETOPERATION to PITCHCLASSSET

• URLINIE spans SCALE_DEGREES ; BASSBRECHUNG outlines HARMONY

• CHORD uses UPPERSTRUCTURETRIAD ; SLASHCHORD bass_is PITCHCLASS

• LICK compatible_with CHORDSCALE ; GUIDETONELINE threads_through PROGRESSION

• RAGA defined_by AROHANA/AVAROHANA ; RAGA characteristic_phrase PAKAD

• MAQAM built_from JINS ; TALA cycles BEAT_PATTERN

Digital/notation layer (optional but handy)

Entities

• MIDIEvent (note_on/off), Velocity, PitchBend, ControlChange, NotatedEvent, BarNumber, GridPosition, Score, DAWProject

Relationships

• MIDIEvent represents NOTE ; NOTE has_velocity VELOCITY

• NOTATEDEVENT aligns_to GRIDPOSITION ; SCORE contains SECTION

Practical properties (attach to many entities)

• Name, Symbol, Spelling (enharmonics), ScaleDegrees, Accidentals, SemitoneSpan, Quality, Function, Degree (e.g., ̂1, ̂5), Inversion (6, 6⁴, 4₂…), Voicing, Registration, Frequency(Hz), Cents, Probability (for generative models), Difficulty, GenreTag, StyleTag

---

Example triples (human-readable)

• C♯ enharmonic_equivalent_of D♭

• C major scale consists_of {C, D, E, F, G, A, B}

• V7 of G major resolves_to I (Gmaj7) ; function = Dominant

• Ger+6 resolves_to V ; contains {♭6, 1, ♯4, ♭3} (relative scale degrees)

• ii–V–I in C orders {Dmin7 → G7 → Cmaj7}

• Leading tone (B in C major) resolves_to C (tendency-tone)

• Mixolydian is_chordscale_for V7

• Upper-structure triad (E major) over C7 yields {♯9, ♭13} colors

• Sonata form has_section {Exposition, Development, Recap}

• Raga Yaman defined_by Arohana/Avarohana ; pakad = phrase motif

• Alto Sax sounds_at M2 down ; notated_in treble clef

---

Optional: minimal Cypher-ish schema sketch

// Labels
CREATE CONSTRAINT note_name IF NOT EXISTS FOR (n:Note)       REQUIRE n.name IS UNIQUE;
CREATE CONSTRAINT pc_name   IF NOT EXISTS FOR (p:PitchClass) REQUIRE p.name IS UNIQUE;
CREATE CONSTRAINT chord_id  IF NOT EXISTS FOR (c:Chord)      REQUIRE c.symbol IS UNIQUE;
CREATE CONSTRAINT scale_id  IF NOT EXISTS FOR (s:Scale)      REQUIRE s.name IS UNIQUE;
CREATE CONSTRAINT key_id    IF NOT EXISTS FOR (k:Key)        REQUIRE k.name IS UNIQUE;

// Relationship types you’ll likely use
// :ENHARMONIC_EQUIVALENT, :IN_OCTAVE, :BUILT_FROM, :CONSISTS_OF, :DERIVED_FROM,
// :CONTAINS, :HAS_QUALITY, :HAS_EXTENSION, :HAS_ALTERATION, :DIATONIC_TO,
// :BORROWED_FROM, :ANALYZES, :REALIZES, :RESOLVES_TO, :ORDERS, :HAS_FUNCTION,
// :CONCLUDES, :TARGETS, :PIVOTS_BETWEEN, :MODULATES_TO, :ADJACENT_ON_COF, :NEIGHBORS_ON_TONNETZ

---

Design notes & tips

• Model concepts (e.g., “Dominant function”) separately from instances (a specific V7 in G major). Attach analysis as nodes (Roman numerals, figured bass) so you can keep multiple analytical layers.

• Keep enharmonics and spelling explicit—crucial for notation vs. pitch-class logic.

• Use scale-degree properties (degree, accidental) on Note-in-Key edges for powerful queries (e.g., “all ♭6 in minor keys that resolve to 5”).

• Voice-leading pays off if you store events (note onsets, durations) and connect them by :LEADS_TO within the same voice.

• For jazz, upper-structure & chord-scale links make recommendation/voicing engines trivial.

• For set theory, store Forte numbers and Tn/In links to enable similarity queries.

If you want, I can turn this into a seed CSV (entities + relationship rows) or a ready-to-run Neo4j schema with sample data (e.g., ii–V–I in all keys, common cadences, and voice-leading rules).

Yes

Awesome — I’ll give you a ready-to-run Neo4j seed that sets up the schema and inserts a compact but versatile starter dataset (pitches, chord qualities, functions, modes, a few keys, and ii–V–I progressions + PAC cadences). You can paste this whole thing into Neo4j Browser.

Neo4j schema + seed data (copy/paste into Neo4j Browser)

// =====================
// 0) Clean slate (optional in a dev DB)
// =====================
OPTIONAL MATCH (n) DETACH DELETE n;

// =====================
// 1) Constraints
// =====================
CREATE CONSTRAINT note_name        IF NOT EXISTS FOR (n:Note)        REQUIRE n.name IS UNIQUE;
CREATE CONSTRAINT pc_name          IF NOT EXISTS FOR (p:PitchClass)  REQUIRE p.name IS UNIQUE;
CREATE CONSTRAINT qual_name        IF NOT EXISTS FOR (q:ChordQuality)REQUIRE q.name IS UNIQUE;
CREATE CONSTRAINT func_name        IF NOT EXISTS FOR (f:HarmonicFunction) REQUIRE f.name IS UNIQUE;
CREATE CONSTRAINT scale_name       IF NOT EXISTS FOR (s:Scale)       REQUIRE s.name IS UNIQUE;
CREATE CONSTRAINT mode_name        IF NOT EXISTS FOR (m:Mode)        REQUIRE m.name IS UNIQUE;
CREATE CONSTRAINT key_name         IF NOT EXISTS FOR (k:Key)         REQUIRE k.name IS UNIQUE;
CREATE CONSTRAINT chord_symbol     IF NOT EXISTS FOR (c:Chord)       REQUIRE c.symbol IS UNIQUE;
CREATE CONSTRAINT roman_id         IF NOT EXISTS FOR (r:RomanNumeral)REQUIRE r.id IS UNIQUE;
CREATE CONSTRAINT prog_id          IF NOT EXISTS FOR (p:Progression) REQUIRE p.id IS UNIQUE;
CREATE CONSTRAINT cadence_name     IF NOT EXISTS FOR (c:Cadence)     REQUIRE c.name IS UNIQUE;

// =====================
// 2) Core vocab
// =====================

// 2a) Pitch classes (12-TET names; include common enharmonics as aliases)
UNWIND [
  {name:'C',   enh:['B#']},
  {name:'C#',  enh:['Db']},
  {name:'D',   enh:[]},
  {name:'D#',  enh:['Eb']},
  {name:'E',   enh:['Fb']},
  {name:'F',   enh:['E#']},
  {name:'F#',  enh:['Gb']},
  {name:'G',   enh:[]},
  {name:'G#',  enh:['Ab']},
  {name:'A',   enh:[]},
  {name:'A#',  enh:['Bb']},
  {name:'B',   enh:['Cb']}
] AS pc
MERGE (p:PitchClass {name: pc.name})
WITH p, pc
UNWIND pc.enh AS alt
MERGE (e:PitchClass {name: alt})
MERGE (p)-[:ENHARMONIC_EQUIVALENT]->(e);

// 2b) Chord qualities
UNWIND [
  {name:'major',      symbol:''},
  {name:'minor',      symbol:'m'},
  {name:'diminished', symbol:'dim'},
  {name:'augmented',  symbol:'aug'},
  {name:'dominant7',  symbol:'7'},
  {name:'major7',     symbol:'maj7'},
  {name:'minor7',     symbol:'m7'},
  {name:'half-diminished7', symbol:'m7b5'}
] AS cq
MERGE (:ChordQuality {name:cq.name, symbol:cq.symbol});

// 2c) Harmonic functions
UNWIND ['Tonic','Subdominant','Dominant'] AS fn
MERGE (:HarmonicFunction {name:fn});

// 2d) Scales & modes
MERGE (maj:Scale {name:'Major', degrees:[1,2,3,4,5,6,7]});
MERGE (minNat:Scale {name:'Natural Minor', degrees:[1,2,b3,4,5,b6,b7]});
UNWIND [
  {name:'Ionian',    parent:'Major',        degree:1},
  {name:'Dorian',    parent:'Major',        degree:2},
  {name:'Phrygian',  parent:'Major',        degree:3},
  {name:'Lydian',    parent:'Major',        degree:4},
  {name:'Mixolydian',parent:'Major',        degree:5},
  {name:'Aeolian',   parent:'Major',        degree:6},
  {name:'Locrian',   parent:'Major',        degree:7}
] AS m
MATCH (s:Scale {name:m.parent})
MERGE (mode:Mode {name:m.name, parent:s.name, degree:m.degree})
MERGE (mode)-[:DERIVED_FROM]->(s);

// =====================
// 3) A handful of keys
// =====================
UNWIND [
  {name:'C major', tonic:'C', mode:'Ionian'},
  {name:'G major', tonic:'G', mode:'Ionian'},
  {name:'D major', tonic:'D', mode:'Ionian'},
  {name:'F major', tonic:'F', mode:'Ionian'},
  {name:'A minor', tonic:'A', mode:'Aeolian'},
  {name:'E minor', tonic:'E', mode:'Aeolian'}
] AS k
MATCH (t:PitchClass {name:k.tonic})
MATCH (m:Mode {name:k.mode})
MERGE (key:Key {name:k.name})
  ON CREATE SET key.tonic = k.tonic, key.mode = k.mode
MERGE (key)-[:TONIC]->(t)
MERGE (key)-[:MODE]->(m);

// =====================
// 4) Helper: make a chord quickly
// =====================
CALL {
  WITH 'stub' AS _
  RETURN 1 AS ok
}
YIELD ok
WITH ok
CALL {
  // procedure-like subquery: CREATE chord, link to pc, quality, key, function, roman numeral
  WITH [
    // C major ii–V–I
    {key:'C major', rn:'ii', root:'D', symbol:'Dm7',   quality:'minor7',     func:'Subdominant'},
    {key:'C major', rn:'V',  root:'G', symbol:'G7',    quality:'dominant7',  func:'Dominant'},
    {key:'C major', rn:'I',  root:'C', symbol:'Cmaj7', quality:'major7',     func:'Tonic'},

    // G major ii–V–I
    {key:'G major', rn:'ii', root:'A', symbol:'Am7',   quality:'minor7',     func:'Subdominant'},
    {key:'G major', rn:'V',  root:'D', symbol:'D7',    quality:'dominant7',  func:'Dominant'},
    {key:'G major', rn:'I',  root:'G', symbol:'Gmaj7', quality:'major7',     func:'Tonic'},

    // D major ii–V–I
    {key:'D major', rn:'ii', root:'E', symbol:'Em7',   quality:'minor7',     func:'Subdominant'},
    {key:'D major', rn:'V',  root:'A', symbol:'A7',    quality:'dominant7',  func:'Dominant'},
    {key:'D major', rn:'I',  root:'D', symbol:'Dmaj7', quality:'major7',     func:'Tonic'},

    // F major ii–V–I
    {key:'F major', rn:'ii', root:'G', symbol:'Gm7',   quality:'minor7',     func:'Subdominant'},
    {key:'F major', rn:'V',  root:'C', symbol:'C7',    quality:'dominant7',  func:'Dominant'},
    {key:'F major', rn:'I',  root:'F', symbol:'Fmaj7', quality:'major7',     func:'Tonic'},

    // Minor keys: Em (iiø–V–i) and Am (iiø–V–i) using functional jazz-ish analysis
    {key:'E minor', rn:'iiø', root:'F#',symbol:'F#m7b5',quality:'half-diminished7', func:'Subdominant'},
    {key:'E minor', rn:'V',   root:'B', symbol:'B7',    quality:'dominant7',        func:'Dominant'},
    {key:'E minor', rn:'i',   root:'E', symbol:'Em7',   quality:'minor7',           func:'Tonic'},

    {key:'A minor', rn:'iiø', root:'B', symbol:'Bm7b5', quality:'half-diminished7', func:'Subdominant'},
    {key:'A minor', rn:'V',   root:'E', symbol:'E7',    quality:'dominant7',        func:'Dominant'},
    {key:'A minor', rn:'i',   root:'A', symbol:'Am7',   quality:'minor7',           func:'Tonic'}
  ] AS rows
  UNWIND rows AS r
  MATCH (k:Key {name:r.key})
  MATCH (pc:PitchClass {name:r.root})
  MATCH (q:ChordQuality {name:r.quality})
  MATCH (f:HarmonicFunction {name:r.func})
  MERGE (c:Chord {symbol:r.symbol})
    ON CREATE SET c.root = r.root, c.quality = r.quality
  MERGE (c)-[:ROOT]->(pc)
  MERGE (c)-[:HAS_QUALITY]->(q)
  MERGE (c)-[:DIATONIC_TO]->(k)
  MERGE (c)-[:HAS_FUNCTION]->(f)
  WITH r, k, c
  MERGE (rn:RomanNumeral {id: r.key + ':' + r.rn})
    ON CREATE SET rn.text = r.rn
  MERGE (rn)-[:ANALYZES]->(c)
  RETURN count(*) AS made
}
RETURN 'Seeded chords & roman numerals' AS status;

// =====================
// 5) Progressions (ii–V–I) + order edges
// =====================
UNWIND [
  'C major','G major','D major','F major','E minor','A minor'
] AS keyName
MATCH (k:Key {name:keyName})
WITH k, CASE
  WHEN k.name ENDS WITH 'minor' THEN ['iiø','V','i']
  ELSE ['ii','V','I']
END AS steps
MATCH (rn1:RomanNumeral {id: k.name + ':' + steps[0]})-[:ANALYZES]->(c1:Chord)
MATCH (rn2:RomanNumeral {id: k.name + ':' + steps[1]})-[:ANALYZES]->(c2:Chord)
MATCH (rn3:RomanNumeral {id: k.name + ':' + steps[2]})-[:ANALYZES]->(c3:Chord)
MERGE (p:Progression {id: k.name + ':ii-V-I'})
  ON CREATE SET p.name = 'ii–V–I', p.inKey = k.name
MERGE (p)-[:ORDERS {step:1}]->(c1)
MERGE (p)-[:ORDERS {step:2}]->(c2)
MERGE (p)-[:ORDERS {step:3}]->(c3);

// =====================
// 6) Cadences (PAC linking V → I / i)
// =====================
UNWIND [
  'C major','G major','D major','F major','E minor','A minor'
] AS keyName
MATCH (k:Key {name:keyName})
WITH k, CASE WHEN k.name ENDS WITH 'minor' THEN 'i' ELSE 'I' END AS tonicRN
MATCH (:RomanNumeral {id: k.name + ':V'})-[:ANALYZES]->(dom:Chord)
MATCH (:RomanNumeral {id: k.name + ':' + tonicRN})-[:ANALYZES]->(ton:Chord)
MERGE (cad:Cadence {name: k.name + ':PAC'}) ON CREATE SET cad.kind='PAC'
MERGE (cad)-[:RESOLVES {type:'authentic'}]->(dom)
MERGE (dom)-[:RESOLVES_TO]->(ton);

// =====================
// 7) Convenience: circle-of-fifths neighbors for the included keys
// =====================
UNWIND [
  ['C major','G major'], ['G major','D major'], ['D major','A major'], ['F major','C major']
] AS pair
WITH pair
MATCH (k1:Key {name:pair[0]}), (k2:Key {name:pair[1]})
MERGE (k1)-[:ADJACENT_ON_COF]->(k2);

// Done
RETURN 'OK' AS seeded;

---

What you can query immediately

• All chords and their functions in a key:

MATCH (k:Key {name:'C major'})<-[:DIATONIC_TO]-(c:Chord)-[:HAS_FUNCTION]->(f)
RETURN c.symbol AS chord, f.name AS function ORDER BY function, chord;

• The ii–V–I progression (ordered) in each key:

MATCH (p:Progression {name:'ii–V–I'})- [o:ORDERS]->(c:Chord)
RETURN p.inKey AS key, o.step AS step, c.symbol AS chord
ORDER BY key, step;

• Find perfect authentic cadences and their resolution targets:

MATCH (cad:Cadence {kind:'PAC'})-[:RESOLVES]->(:Chord)-[:ANALYZES]<-(:RomanNumeral {text:'V'}),
      (:RomanNumeral)-[:ANALYZES]->(ton:Chord)<-[:RESOLVES_TO]-(:Chord)
RETURN cad.name AS cadence, ton.symbol AS final_tonic
ORDER BY cadence;

---

Want CSVs instead?

If you prefer to bulk-load, use this schema for two CSVs:

nodes.csv

id,label,name,props_json
pc:C,PitchClass,C,"{}"
pc:G,PitchClass,G,"{}"
cq:dom7,ChordQuality,dominant7,"{""symbol"":""7""}"
fn:Dom,HarmonicFunction,Dominant,"{}"
key:Cmaj,Key,C major,"{""tonic"":""C"",""mode"":""Ionian""}"
ch:C7,Chord,C7,"{""root"":""C"",""quality"":""dominant7"",""symbol"":""C7""}"
rn:Cmaj:V,RomanNumeral,V,"{}"
prog:Cmaj:ii-V-I,Progression,ii–V–I,"{""inKey"":""C major""}"
cad:Cmaj:PAC,Cadence,C major:PAC,"{""kind"":""PAC""}"

rels.csv

start_id,rel_type,end_id,props_json
key:Cmaj,TONIC,pc:C,"{}"
ch:C7,ROOT,pc:C,"{}"
ch:C7,HAS_QUALITY,cq:dom7,"{}"
rn:Cmaj:V,ANALYZES,ch:C7,"{}"
prog:Cmaj:ii-V-I,ORDERS,ch:Dm7C,"{""step"":1}"
prog:Cmaj:ii-V-I,ORDERS,ch:G7C,"{""step"":2}"
prog:Cmaj:ii-V-I,ORDERS,ch:Cmaj7C,"{""step"":3}"
cad:Cmaj:PAC,RESOLVES,rn:Cmaj:V,"{""type"":""authentic""}"

If you’d like, I can generate full CSVs for all 12 major and 12 minor keys (ii–V–I / iiø–V–i, common borrowed chords, N6, It/Fr/Ger+6, secondary dominants, and a minimal set-theory layer) so you can LOAD CSV straight into Neo4j — just say the word and tell me whether you prefer “one big CSV” or “separate per-entity CSVs.”

From <https://chatgpt.com/>  ChatGPT 5