Tibetan Ta ཏ

Ta is the ninth of the thirty consonants and opens the third line. It writes a voiceless unaspirated stop made with the tip of the tongue at the teeth, further forward than the English t, which is formed on the ridge behind them. In modern Lhasa speech it carries high tone and has fallen together with tha.

Ta is a root letter with no orthographic office of its own, though it takes superscripts readily and takes the ra and wa subscripts.

The line it opens sits at the most conspicuous gap in the whole inherited scheme. In the Indic order from which the Tibetan arrangement descends, the retroflex letters come at exactly this point, between the palatals and the dentals, and in Tibetan they are simply not there. The line runs straight from nya to ta. That absence is a decision, not an oversight: Tibetan has no retroflex consonants and had no use for the five letters that write them. When the language later needed to transcribe Sanskrit accurately, the retroflexes were supplied by a separate device rather than restored to the recitation, so the gap in the alphabet has stayed open for thirteen centuries with the remedy standing outside it.

Because it opens a line, ta also serves the grammars as their specimen dental, in the way that ka serves as their specimen velar. A rule about the dentals is typically stated on ta and then generalised to tha, da and na, which is one more consequence of an alphabet arranged as a grid rather than as a list.

To write Sanskrit retroflexes, the sounds made with the tongue curled back which Sanskrit distinguishes and Tibetan does not have, Tibetan uses mirror images of its own dental letters. Ta reversed writes the retroflex, and so in turn do reversed tha, da, na and sha. They are known as the log yig, the reversed letters.

The device is economical and it is honest about what it is doing. It says, in effect, that this is the dental letter and something has been done to it, which is exactly the relationship the two series have. It also keeps the reversed forms visibly outside the alphabet proper: a reader can see at a glance that a word containing one is a transcription rather than a Tibetan word, in the same way that italics mark a foreign term in an English text.

They are not part of the sum cu. A Tibetan child recites thirty letters and does not recite these, and they are taught later and to a narrower group, since a reader of ordinary Tibetan will meet them chiefly in mantras, in Sanskrit names and in the technical vocabulary of Buddhist scholastic texts.

They are encoded separately in Unicode as well, which reflects the same judgement: the reversed letter is its own character, not a formatting variant of the letter it mirrors, because the two are contrastive and a search that conflated them would fail.