This article records tradition as it has been passed down and reported. Its sources are not yet part of the atlas's verified catalogue.
Tibetan Letters gathers the thirty root consonants of the sum cu, recited in eight lines from the seventh century onward, together with the four vowel signs, gi gu, greng bu, na ro and zhabs kyu, that mark i, e, o and u on a consonant that otherwise carries the inherent vowel a.
The consonants follow the Indian model of five classes by place of articulation, moving from the back of the mouth to the front, but Tibetan pruned the Sanskrit scheme to fit exactly thirty, dropping distinctions such as the retroflex series that classical Tibetan had no use for; a separate set of reversed letterforms was added later, outside the recited thirty, purely to transliterate Sanskrit. The first four letters, ka kha ga nga, stand for the whole alphabet in ordinary speech, much as ABC does for the Latin one, and grammars reach for ka whenever a rule needs a specimen consonant.
Modern Lhasa speech has collapsed much of the classical system into tone: ka and kha, once a plain and an aspirated stop, are now both an aspirated k distinguished only by pitch, and a voiced letter such as ga is heard as an unaspirated k with low tone. Spelling still carries the older distinctions the ear no longer makes. A further complexity belongs to a subset of the letters alone: several, ga among them, serve a second life as prefixes or suffixes in the orthography, silent markers of a syllable's shape, while others, ka and cha among them, are root letters only and never take that orthographic role.