The Geometry of Vowel Sounds
Every vowel you produce is shaped by the position of your tongue inside your mouth. Raise it high and forward, and you get /i/ (as in 'beet'). Lower it and pull it back, and you get /ɑ/ (as in 'father'). The acoustic fingerprint of each position is captured by two frequencies — F1 and F2 — which together create a map where every vowel has a unique coordinate. This is the vowel space.
Why Languages Choose Their Vowels
Of the infinite possible tongue positions, languages select only a handful of vowels — typically 5 to 7. But the choices are not random. Dispersion Theory, proposed by Lindblom in 1972, explains that languages space their vowels to maximize distinctiveness: the further apart two vowels are in acoustic space, the easier they are to tell apart. The universal 5-vowel system /a, e, i, o, u/ represents the most efficient packing of maximum contrast into minimum inventory.
Formants: The Physics of Vowels
When you speak, your vocal cords produce a buzz of many frequencies. Your tongue, lips, and jaw shape the vocal tract into a resonating tube that amplifies certain frequencies (formants) and dampens others. F1, the lowest formant, drops as the tongue rises. F2, the second formant, rises as the tongue moves forward. Together they explain over 90% of vowel perception — which is why a simple 2D plot can represent the entire vowel space.
Cross-Linguistic Patterns
The most striking finding in phonetic typology is the regularity of vowel systems. Three-vowel languages almost always choose /a, i, u/ — the three corners of the vowel space. Five-vowel languages add /e/ and /o/. Seven-vowel languages fill in further. Each addition maximizes the distance from existing vowels, as if languages are solving an optimization problem. This simulation lets you see that optimization in action.