Every time you write name, house, moon or time, you are spelling them the way they sounded around 1400 — naah-muh, hoose, mone, teem. Then, across three centuries, every long vowel in English changed quality — the Great Vowel Shift — while the spelling, newly frozen by the printing press, stayed put. The gap between how English is written and how it is said is the fossil of a sound that walked away.
English vowel letters don’t say what their European cousins say. In Italian or Spanish, i is “ee” and a is “ah”; in English, i in mine is “eye” and a in name is “ay”. That is not English being lawless. The letters were given their Continental values in the Middle Ages, when name really did rhyme with “calm-a” — and then the sounds moved and the letters did not. Tap a word to drop beneath its spelling to the sound it remembers.
Every one of these is the same trick: the digraphs (ee, oo, ou) and the silent final -e are not decoration. They are length marks and vowel values from before the shift, kept on the page after the mouth let them go. Drill into the shift itself below.
Here is the whole shift on one instrument — the phonetician’s vowel chart, a map of the mouth: high vowels at the top, low at the bottom, front on the left, back on the right. Each dot is a long vowel of English, labelled with an everyday word. Drag the timeline from 1400 to 1700 and watch what happened: the long vowels all climbed one step higher, and the two that were already at the ceiling — the vowels of time and house — had nowhere to go, so they broke loose and fell inward as diphthongs.
The motion is schematic — the dots show direction, not a frame-by-frame recording, and the dates are approximate (see the apparatus at the foot). But the endpoints are real and sourced: where each dot starts is the late-Middle-English value, where it lands is the modern one. The pattern — a coordinated raising with the top two breaking away — is exactly what makes linguists call it one shift.
The vowel chart above is a map of the mouth — where the tongue sits. But a vowel is also an acoustic object: feeding a buzz from the voice-box through the shifting shape of the mouth carves out resonances called formants, and the two lowest — F1 and F2 — are almost enough to name the vowel. F1 rises as the mouth opens; F2 rises as the tongue moves forward. So the chart’s two axes are really the two formants in disguise. Here are the published mean formants of the seven Middle-English vowels — press play to hear each one, and read it against where it sits on the chart.
| vowel | IPA | mouth (chart) | F1 Hz | F2 Hz | hear |
|---|
This is the deepest reason the shift is one shift and not seven accidents: it is a coordinated walk through a two-dimensional acoustic space, the high vowels crowding the top until the two at the ceiling have to break for the exits. You can see it and now you can hear it, and — the point of this venue — the picture and the sound are independently sourced and they agree.
Look at the spelling: we write meet with ee and meat with ea, and that difference is not an accident. They were two different vowels — meet a close-e (eː), meat an opener one (ɛː). In the shift, the close-e climbed to iː first; the open-e climbed after it and, in standard English, caught up and merged. Now they are the same sound. But a handful of ea words escaped — they stopped one step short, which is why great rhymes with bait and not with beet.
Here is the cleanest evidence that the shift is real and datable: English kept borrowing French words after it was over, and those late arrivals never went through it. So the very same letters say two different things depending on when the word entered English. mine shifted; machine, borrowed later, did not — and so its i still says the old “ee”.
| letter | native — shifted | sound | late loan — untouched | sound | entered |
|---|
A spelling reformer’s nightmare and a historian’s gift: the i of mine and the i of machine are a 300-year date stamp, hiding in plain sight.
How can anyone know how a vowel sounded before recordings? Three kinds of witness, and they agree. First, the poets: when a rhyme that the spelling still pairs no longer rhymes in the mouth, the vowel moved between the poet and us. Shakespeare’s couplets are full of these.
Second, the puns — wordplay that only lands on the older pronunciation. When the joke goes silent, you have caught a sound mid-flight.
Third, the orthoepists — Renaissance spelling-reformers and pronunciation teachers who, trying to fix English spelling, wrote down in detail how words were actually said. They are our direct phonetic witnesses, and they let us date each move.
Read these lines aloud the way the chart says to read them — long a as “ah”, the final -e sounded — and you are hearing the vowels the modern spelling still records. This is the sound the spelling forgot.
What this is. Standard, textbook historical linguistics — the Great Vowel Shift in English long vowels, c. 1400–1700 — made into something you operate. Nothing here is an original linguistic finding. The only thing new is the form: the shift made runnable on a vowel chart, the spelling shown to be a fossil, and the evidence (rhymes, puns, loanwords, orthoepists) made tappable.
What the values are. Every Middle-English and Modern-English phonetic value is in IPA and sourced to standard scholarship — Roger Lass in the Cambridge History of the English Language (vol. III), Baugh & Cable’s A History of the English Language, and Wells’ Accents of English for the mergers — cross-checked against the consensus account. Every Shakespeare and Chaucer line is transcribed verbatim from a public-domain edition and cited in place. The Original-Pronunciation readings of the puns are David Crystal’s, attributed as interpretation, not fact.
The voice — what you are hearing. Every vowel you can play is formant synthesis: a buzzing glottal source (a ~120 Hz sawtooth) driven through resonators tuned to the vowel’s first two formants, F1 and F2 — the two numbers that almost entirely fix a vowel’s identity. Those numbers are published mean male-voice values, not invented: Catford’s A Practical Introduction to Phonetics (2nd ed., 2001, p. 154) for the seven continental qualities that stand in for the Middle-English long vowels, and Peterson & Barney (1952, Table II) for /ɪ/ and /ʊ/, the diphthong endpoints — each number checked digit-for-digit by an independent fact-check before publishing. The third formant is a fixed nominal 2500 Hz for naturalness and is not a claim. This is an acoustic sketch of each vowel target, never a recording of a speaker: no playable recording of any voice predates 1877, so all Middle-English pronunciation is reconstruction, and the synthesiser only renders the resonances the modern symbols denote — your ears judge the timbre, the numbers carry the claim. And the page’s own acoustic cross-check is real: ordered by the chart’s height the sourced F1 climbs every step, and front-to-back the F2 falls every step — the drawn chart and the published table are two independent witnesses to the same seven vowels.
What the check does — and does not — prove. verify.mjs runs internal-consistency checks (all
green): that the vowel chain is a coherent raising with exactly two breaks, that the merger and its named
escapees land where the page says, that every loanword pair contrasts a shifted native value against an
unshifted late loan, that every Middle-English vowel carries a sourced F1/F2 and that those formants agree in
order with the drawn chart (F1 rising with openness, F2 falling with backness), and a drift guard that the
data embedded in this page is byte-identical to the source of truth. It does not re-derive the phonology, and it cannot: that the long vowels held these
values is the cumulative work of two centuries of philology, which this page demonstrates and cites but does not
re-prove.
The honest caveat, kept in view. The tidy seven-arrow chain is an idealization. The changes were staggered across centuries, varied by region, and whether they form a single causally-linked “chain” at all is debated. The instrument shows the directions and the sourced endpoints — not a claim that seven vowels moved at one instant, or for one reason.
The backing data, the verifier, and the sources live in research/great-vowel-shift. ◆ Part of the Language seam, beside The First Sound Shift.