Tibetan Gi Gu ◌ི

Gi gu is the first of the four Tibetan vowel signs, the dbyangs, and writes the vowel i. It is placed above the consonant it modifies and replaces that consonant inherent a, so ka with gi gu upon it is read ki.

It is a dependent sign and not a letter. It cannot stand alone, it is not one of the thirty, and where a syllable begins with the vowel i the sign is written on a chen, the thirtieth consonant, which exists in part to give such vowels something to attach to. Shown on its own, as in a code chart or on this page, the sign is displayed on a dotted circle, which is the convention for a mark that has no independent position.

The four signs are recited after the thirty consonants, and the thirty and the four together are what a Tibetan child learns as the alphabet. Four is a small inventory, and it is small because Tibetan has a small vowel system and does not distinguish length. Everything a longer inventory would be needed for, in particular the long vowels and diphthongs of Sanskrit, is handled by the transcription apparatus that sits outside the alphabet.

The four signs are recited in a fixed order, i, u, e, o, and gi gu comes first. That order does more than govern the recitation. It is also the order in which the vowels are compared when Tibetan words are sorted, once the root letter, the superscript and the subscript have been taken into account, so the sequence is part of how a dictionary is arranged.

Old Tibetan writing has a feature that later orthography abandoned: a form of the vowel sign i written the other way round, mirroring the ordinary shape. It is known as the reversed gi gu, and its presence in a text is one of the clearest signals that the text or its exemplar is early.

What made this observable in quantity was the recovery of a very large body of Old Tibetan documents from a sealed library cave at the Mogao site near Dunhuang, on the eastern edge of the Taklamakan, opened at the turn of the twentieth century and dispersed to collections in Europe and Asia. The cave held manuscripts in many languages, and among them a substantial Tibetan corpus from the period of Tibetan control of the region: contracts, letters, administrative documents, divination texts and Buddhist works.

The importance of that corpus is that it is not a canonical tradition copied and recopied. It is ordinary written Tibetan from a datable period, and it shows the orthography as it was actually used before the great standardisations of vocabulary and spelling. The reversed gi gu, the second suffix da, particular particle spellings and a set of other archaisms all appear there.

What later scribes did with these features varied. Some reproduced what was in front of them and some normalised it, so an archaism in a manuscript is evidence about its exemplar as much as about its date.