Tibetan tradition tells the story precisely: in the seventh century CE the king Songtsen Gampo, wishing to bring the dharma and statecraft to Tibet, sent his minister Thonmi Sambhota to India, and the minister returned to devise the Tibetan script on the model of Indian letters, together with a grammar. Modern scholarship values the story while handling it cautiously, since the earliest sources for Thonmi are centuries later than the events, and some historians treat him as legendary; what is secure is that the script appears in the seventh to eighth centuries, modelled on a northern Indian Brahmi-derived script of the Gupta type, and that its creation is bound from the beginning to the project of translating Buddhism.
The alphabet arranges thirty consonants and four vowel signs on the Indic phonetic grid, adapted to Tibetan sounds. Because the spoken language moved on while the orthography froze after a ninth-century reform, written Tibetan preserves consonant clusters that no longer sound, making the script a fossil record of the old language. Two great styles divide use: uchen, the headed letters of carving, printing and scripture, and ume, the headless cursive of correspondence and notes, with dbu-med shorthand traditions between them.