A guided tour of how the West learned to write music down — and a practical, stage-by-stage lesson in how to read each notation you will meet in a chant book, a mensural manuscript or a lute tablature.
History of Early Music · A Didactic Guide to the Notation
For most of history music was purely oral. The vast chant repertory of the early Church — thousands of melodies — was held entirely in the memory of singers and passed on by ear, cantor to novice, over a lifetime. The first notation was not a recording but a memory aid: a set of reminders scribbled above words a singer already half-knew.
A chanter memorised the entire repertory. Charlemagne wanted one uniform chant across his empire — but sound cannot travel far by memory alone.
The earliest signs only reminded you of a tune you knew. They fixed nothing precisely — contour, not content.
Six centuries of invention slowly added pitch, then rhythm, then metre — until a stranger could sing a tune never heard before, at sight.
The story that follows is the gradual pinning-down of two things sound refuses to hold still: how high and how long.
The first written signs — neumes (Greek neuma, "gesture") — appear in manuscripts around 900 CE. They are cheironomic: they trace the hand-gestures a cantor used to conduct, showing only the contour of the melody. Each neume tells you how many notes fall on a syllable and which way the voice turns.
How to read it: a neume sits over its syllable; read left→right, higher on the page = higher in pitch. But these earliest neumes give no exact pitch and no rhythm — only the melodic gesture. You had to know the tune already.
The great weakness of neumes was pitch. Scribes fixed it in two steps. First they placed the neumes at careful heights above the text (diastematic or "heightened" neumes). Then they scratched a single reference line into the parchment with a dry point, and later drew it in colour — red for F, yellow for C.
How to read it: measure each neume's height against the coloured line. A note on the red line is F; two steps above it is A, and so on. The dry-point / coloured line is the direct ancestor of the modern staff line — one certainty at last.
Around 1030 the monk Guido of Arezzo ruled a whole ladder of lines — his four-line staff — and put a clef at the front to name one of them. A singer could now read an unfamiliar melody at sight. Below, a C-clef fixes the bottom line as C; every other pitch is counted from it.
Hear exactly what the eye reads: eight notes climbing line–space–line from C to C, in medieval Pythagorean tuning
Guido also gave the notes names. From the hymn Ut queant laxis, whose six phrases each start a step higher, he drew the syllables Ut–Re–Mi–Fa–Sol–La — the six-note hexachord. Its one secret: the single semitone always falls between Mi–Fa.
To sing beyond six notes you mutate — slide into an overlapping hexachord mid-scale. Where a B had to be flattened (soft, b molle) or kept natural (hard, b durum) gave us "flat" and "natural", and the editorial sharps/flats we call musica ficta. Guido mapped the whole system onto the joints of the left hand, a memory palace.
Six rising notes on C; listen for the one narrow step, Mi–Fa (E–F), that anchors the whole system
The neume shapes hardened into square notes on Guido's four-line staff. This is the very notation you still meet in a modern chant book (the Graduale). Here is a worked Alleluia incipit — decode it part by part.
The C-clef (or F-clef) names one line. Every pitch is read relative to it — the staff can sit anywhere on the page.
Squares grouped together belong to one syllable. Read a stacked group bottom-to-top, a spread group left-to-right.
None is written. Notes are broadly equal; word and phrase shape the flow. There are no bar-lines.
The little hook at a line's end predicts the first note of the next line — a page-turn cue.
The composers of Notre-Dame (Léonin, Pérotin, c.1200) needed voices to move together in time. Their solution: six rhythmic modes, fixed patterns of long (longa) and short (brevis) — and, crucially, you signalled which mode by the pattern of ligatures (joined note-groups). For the first time, rhythm itself was written.
The lilting long–short of Mode 1 — the same trochaic gait shown in the top row of the chart
How to read it: you do not read the rhythm from a note's shape yet — you infer it from the ligature pattern, which tells you the mode; the mode then tells you the long–short sequence to apply.
Around 1260 Franco of Cologne made the decisive leap in his Ars cantus mensurabilis: a single note's shape now tells you its length — no mode to infer, no context needed. This is mensural ("measured") notation, the true root of our system.
How to read it: learn three shapes — stemmed square = longa, plain square = breve, diamond = semibreve. Longest to shortest, left to right. Triple grouping ("perfect") is the medieval norm, a nod to the Holy Trinity.
Philippe de Vitry's Ars Nova (c.1320) freed rhythm completely: each note could divide into three (perfect) or two (imperfect), and a mensuration sign at the front declared which. The semicircle for imperfect time is the direct ancestor of our modern C time-signature.
Identical pitches, two mensurations: first a lilting triple grouping, then an even duple — the sign at the front decides which
As paper replaced parchment, filled black notes tore the page, so scribes drew them void (outline only) — white mensural notation, the look of a Josquin or Palestrina source. Two things still surprise a modern eye: there are no bar-lines, and singers read from separate part-books, not a score.
How to read it: the shapes are the mensural ones, only hollow. With no bar-lines you keep your place by counting the beat (tactus); with no score, you must listen across the room to fit your part to the others.
For instruments a different logic won out. Lute tablature does not tell you which pitch — it tells you where to put your fingers. The lines are the lute's courses (strings), a letter says which fret to stop, and rhythm flags float above the grid.
The five letters above become five plucked pitches — A–G–F–E–D, a little falling "tear" phrase, on a lute timbre
By the 1600s the last pieces fell into place: filled note-heads with flags for short values, the five-line staff, bar-lines dividing a regular metre, a time signature, and the score stacking every voice so one reader sees the whole. Nothing was invented from nothing — each stage kept something.
Guido's ruled lines, grown from four to five, and his movable clefs.
The mensural family — breve, semibreve, minim — survives as our whole, half and quarter notes.
Ars Nova's perfect/imperfect become simple and compound time; the semicircle became C.
Regular bar-lines and the vertical score are Baroque additions — not medieval at all.
Four practical habits will let you open almost any early-music score — original or modern transcription — and find your way.
Find the clef first. A C- or F-clef names one of the four lines; read every square from there. Groups on one syllable, no bar-lines, roughly equal notes, and a custos at each line-end.
Expect no bar-lines, or only faint mensurstriche between the staves. Note-values are often halved or quartered from the original so the maths looks familiar.
The bracket rule: an accidental printed above a note or in [square brackets] was not in the source — it is the editor's suggested musica ficta, the sharp or flat a Renaissance singer would have added by convention. Take it as advice, not law.
Seven centuries of invention turned a fleeting gesture of the hand into an exact grid of pitch and time. Each stage solved one problem and handed the rest on: neumes caught the contour, Guido fixed the pitch, Notre-Dame and Franco fixed the rhythm, Ars Nova freed the metre. To read early music is to read that memory, thawed.
Neumes → heightened neumes → the coloured line → Guido's staff & clefs. Pitch, made certain.
Rhythmic modes → Franconian shapes → Ars Nova prolation. Duration, made visible.
Five lines, bar-lines, a score — the modern page is a museum of every earlier idea.
"Nisi enim ab homine memoria teneantur soni, pereunt, quia scribi non possunt." — "For unless sounds are held in the memory by man, they perish, because they cannot be written down."
— Isidore of Seville, Etymologiae (c. 625) — the very problem that all notation was invented to solvec. 900 – 1600 · from gesture to grid