Workshop 116 — Building a Complete Original Piece from Nothing: An Integrated Workflow from First Seed to Finished Piano Composition
Workshops 103–115 have gradually separated composition into its component systems:
- motif and melody
- harmony
- bass
- inner voices
- form
- transitions
- development
- harmonic rhythm
- texture/register
- chord-progression grammar
- groove
- spontaneous-sounding melodic development
Now we deliberately put them back together.
The question is:
If you sit at the piano tomorrow with absolutely nothing written, how do you actually create a complete original piece without becoming lost among all these possibilities?
This workshop is not about learning another sophisticated harmonic device.
It is about workflow.
The central lesson is:
Do not compose everything simultaneously. Establish one strong constraint, discover the piece's identity, create structural landmarks, and add musical dimensions in passes—always preserving the possibility of changing earlier decisions when a later layer reveals something better.
116.1The blank-page problem is really a decision problemAt the piano you could choose:116.2So reduce possibilities immediatelyBefore playing anything, choose perhaps:116.3Do not choose twenty constraintsYou need:116.4The first seed can come from five placesYou already know them:116.5Route A — melody firstInvent:116.6Ask what gives motif identityIs it:116.7Route B — rhythm firstTap:116.8Route C — bass firstWrite:116.9Route D — harmony firstChoose:116.10Route E — texture firstPerhaps:116.11For today's main exampleLet's use a melodic seed.116.12First questionIs this motif:116.13The one-minute testAfter one minute away:116.14Give motif one defining featureOur seed:116.15Do not harmonize yetCreate:116.16Phrase 1G-A♭-C-G.116.17This is importantWe got:116.18Pass 1 — structural melodyWrite only:116.19Mark phrase destinationsBar 4:116.20Pass 2 — contour mapWrite:116.21Pass 3 — climax reservationHighest note currently:116.22Composition begins controlling scarcity earlyDo not spend:116.23Pass 4 — bass skeletonDo not fill chords yet.116.24Does it work?If yes:116.25Bass alternativeTry:116.26Keep alternatives temporarilyComposition does not require committing instantly.116.27Pass 5 — harmonic functionMark bass/phrase areas:116.28Pass 6 — harmonic rhythmHow often should harmony change?116.29Version A — activeone chord/bar.116.30Do not automatically choose richer versionThe melody decides.116.31Pass 7 — guide-tone lineChoose one inner voice.116.32Pass 8 — voicingNow distribute:116.33Pass 9 — textureOpening could be:116.34Pass 10 — articulationCharacter word was:116.35Pass 11 — rhythm/grooveMaybe piece is not groove-driven.116.36Pass 12 — play the entire A sectionNow stop.116.37B must respond to that knowledgeNot:116.38Choose B's jobMaybe:116.39Choose two dimensions to changeFor B:116.40B could move to E♭ majorRelated key.116.41Or A♭ majorMore distance.116.42Or remain C-centered but DorianIntroduce:116.43Choose oneFor example:116.44Transition designA ends:116.45Reinterpret the same chordA ends on Cm.116.46This is composition at a higher levelOne chord does two formal jobs:116.47B motif transformationOriginal:116.48Don't worry if exact intervals differ slightlyMotivic relationship can preserve:116.49B rhythmic developmentA used:116.50B registerMove melody:116.51B bassInstead of long roots:116.52But do not change everythingIf B has:116.53Pass 13 — create B structural melodyAgain:116.54Pass 14 — B bassDesign separately.116.55Pass 15 — B harmonyDo not overthink.116.56Now ask: where is global climax?If total piece:116.57Suppose climax is late BReserve:116.58Build toward climax by three dimensions onlyFor example:116.59Avoid “everything ramps”If every dimension increases continuously:116.60Example buildBars 9–10: higher register.116.61Pass 16 — decide what returnsAfter B:116.62Because B was activeA' could return:116.63Return strategyUse:116.64The return does not need new melodyRecognition is enough.116.65Pass 17 — decide what *not* to returnMaybe original accompaniment never returns.116.66Pass 18 — endingBefore writing coda, choose final:116.67That final G is 5thStable.116.68If you want stronger closureEnd melody:116.69Pass 19 — write backward from endingWhat precedes final Cm?116.70Suppose final cadence is simpleG7 → Cm.116.71Pass 20 — coda materialDo not invent new motif.116.72We now have macro formIntro: perhaps none or 2 bars.116.73Now the piece exists structurallyBut it still needs revision.116.74Revision Pass 1 — melody onlyPlay melody from start to end.116.75Delete weak melodic fillerEspecially:116.76Revision Pass 2 — bass onlyPlay entire bass line.116.77Revision Pass 3 — outer voicesMelody + bass.116.78Revision Pass 4 — inner voices onlyExtract middle lines.116.79Revision Pass 5 — harmony strippedReduce every chord to:116.80Revision Pass 6 — harmonic rhythmWrite chord-duration map.116.81Revision Pass 7 — rhythmTap melody rhythm alone.116.82Revision Pass 8 — textureMap:116.83Revision Pass 9 — registerMark:116.84Revision Pass 10 — climaxIdentify:116.85Separate some peaks if desiredMaybe:116.86Revision Pass 11 — transitionsPlay only:116.87Revision Pass 12 — beginningDoes opening reveal:116.88Revision Pass 13 — endingPlay last 8 bars repeatedly.116.89Revision Pass 14 — repetitionMark every repeated idea:116.90Revision Pass 15 — noveltyMark every completely new idea.116.91Revision Pass 16 — complexityCircle:116.92Revision Pass 17 — silenceWhere does nothing happen?116.93Revision Pass 18 — physical pianismIs the piece:116.94Composition is not improved by impossible hand geometryUnless writing for:116.95Revision Pass 19 — memoryStop playing for several hours.116.96Revision Pass 20 — emotional neutrality checkDo not ask only:116.97Now let's formalize a reusable composition workflowChoose:116.98Stage 2 — SeedCreate:116.99Stage 3 — Identity testWhat survives if:116.100Stage 4 — Small phraseTurn seed into:116.101Stage 5 — AnswerCreate:116.102Stage 6 — Structural melodyRemove ornament.116.103Stage 7 — BassCompose:116.104Stage 8 — Functional/harmonic skeletonName:116.105Stage 9 — Inner lineAdd:116.106Stage 10 — TextureChoose:116.107Stage 11 — Groove/pacingDefine:116.108Stage 12 — Formal expansionDecide:116.109Stage 13 — Section jobsWrite one sentence:116.110Stage 14 — transitionsFor each boundary:116.111Stage 15 — developmentUse existing DNA.116.112Stage 16 — climax budgetingReserve:116.113Stage 17 — ending-first refinementChoose exact final:116.114Stage 18 — reductionStrip:116.115Stage 19 — recompositionNow selectively add:116.116Stage 20 — performance versionDetermine what is:116.117A score does not need to specify everythingFor your own piece, you can create:116.118This may suit your goal especially wellBecause you're aiming not merely to:116.119Define three layers of certaintymelody landmarks form major cadences climax * ending116.120Level 1 performance versionPlay exactly written.116.121Level 2 performance versionMelody fixed.116.122Level 3Melody skeleton fixed.116.123Level 4Form/landmarks fixed.116.124Level 5Motif/form only fixed.116.125But first establish baselineYou cannot meaningfully vary:116.126The canonical version is not necessarily definitiveIt is:116.127Write a lead-sheet version of your original pieceInclude:116.128Then write piano-performance annotations separatelyExamples:116.129Now transpose itThis is a crucial test.116.130What survives?If motif depends on:116.131Transpose bass as degreesC-B♭-A♭-G116.132Transpose harmony functionCm-Fm-G7-Cm116.133Transpose motif relationallyG-A♭-C-G in C minor corresponds:116.134The transposition test exposes hidden understandingIf you can transpose by:116.135Now play it in three stylesSame composition:116.136If the melody collapses under another styleMaybe its identity is strongly rhythmic.116.137If harmony collapses under simplificationMaybe color is structural.116.138If piece survives everythingThen its skeleton is robust.116.139The “too many good ideas” problemSuppose while composing B you discover:116.140Store unused ideasCreate:116.141A good idea can be wrong for the pieceImportant.116.142The “I got bored” problemAfter playing A 50 times while composing:116.143Take time awayFresh ears are part of composition.116.144The “it sounds too simple” problemSimple does not mean:116.145The “it sounds derivative” problemMost early sketches resemble:116.146The originality hierarchyLess important:116.147The “I can't find B” problemDo not write new melody immediately.116.148The “transition is awkward” problemIdentify:116.149The “ending feels arbitrary” problemWrite ending:116.150The “climax isn't strong” problemUsually not:116.151The “piece is too long” problemDelete:116.152The “piece is too short” problemDo not add unrelated material.116.153The “harmony is boring” problemBefore reharmonizing ask:116.154The “melody is boring” problemBefore adding notes ask:116.155The “groove is boring” problemBefore adding syncopation ask:116.156The “everything sounds equally important” problemFix:116.157Composition is largely hierarchy managementWhich note matters?116.158The 80/20 composition ruleOften 80% of identity comes from:116.159Signature-detail mute testRemove one distinctive feature.116.160ConverselyIf you remove a clever reharm and nothing changes:116.161Practice Loop A — 30-minute compositionSet timer.116.162Practice Loop B — three beginningsWrite three completely different openings:116.163Practice Loop C — one seed, three piecesUse:116.164Practice Loop D — compose without chord symbolsFirst 20 minutes:116.165Practice Loop E — compose without melody firstWrite:116.166Practice Loop F — two-voice draftEntire first version:116.167Practice Loop G — reduction/rebuildTake finished draft.116.168Improvisation Challenge 1 — compose in real timeBefore starting know only:116.169Challenge 2 — three takesSame rules.116.170Challenge 3 — capture accidental good ideaIf a surprising chord/phrase appears:116.171Challenge 4 — no correction during takeMistake occurs.116.172Challenge 5 — one-page map onlyPerform original piece from:116.173Challenge 6 — transpose immediatelyAfter playing original:116.174Ear Drill — skeleton hearingPlay complete piece.116.175Ear Drill — form memoryWithout piano:116.176Ear Drill — harmonic alternativesAt one phrase ending:116.177Ear Drill — bass alternativesKeep melody fixed.116.178Ear Drill — texture imaginationWithout touching piano:116.179Mini-project 1 — Your First Complete Composition BlueprintBefore playing, write only:116.180Mini-project 2 — 24-Bar Two-Voice CompositionWrite:116.181Mini-project 3 — Three Complete Drafts in One WeekNot three polished pieces.116.182Mini-project 4 — One Piece, Three Arrangement LayersAfter composition is stable:116.183Mini-project 5 — Flexible Original PieceDefine:116.184Mini-project 6 — Composition from an Inner VoiceStart only with:116.185Public-domain “Happy Birthday” as composition modelTake its formal DNA:116.186Example abstract blueprintPhrase 1: short statement.116.187Conceptual lesson from “Mack the Knife”Its repeated formal identity demonstrates:116.188Conceptual lesson from “The Bare Necessities”A piece can have strong identity because:116.189The deeper melody lessonDo not polish melody in isolation forever.116.190The deeper harmony lessonHarmony should often be:116.191The deeper bass lessonBass is your quickest diagnostic layer after melody.116.192The deeper inner-voice lessonInner voices are where:116.193The deeper rhythm lessonA piece begins living when:116.194The deeper texture lessonTexture should express:116.195The deeper form lessonForm is not:116.196The deeper development lessonBefore inventing new content:116.197The deeper improvisation lessonA composition can remain alive after completion.116.198The deeper editing lessonThe final version should contain:116.199The deep compositional questionNot:116.200A complete 20-pass composition checklistWhen starting an original piece:116.201Do not perform all 20 passes rigidly every timeEventually they merge.116.202When stuck, separate dimensions againIf composition becomes confusing:116.203When the piece is going well, don't interrupt it with analysisRecord.116.204Your ear has veto powerIf theoretical logic says:116.205But intuition also benefits from disciplineIf you love an unusual chord:116.206The finished-versus-perfect ruleFinish a version.116.207The draft hierarchyDoes the piece exist?116.208The composition journalFor each original piece keep:116.209Personal style emerges through repeated decisionsYou may notice you repeatedly prefer:116.210Then deliberately challenge itIf you always write:•Workshop 116 mini-test1. Why is blank-page composition often primarily a decision-overload problem?