Tibetan Ja ཇ

Ja is the seventh of the thirty consonants and the voiced member of the palatal line. In the classical language it was a voiced affricate, the j of English judge. Modern Lhasa speech devoices it to an unaspirated affricate and carries the difference on low tone instead, which is the trade the whole voiced series has made.

Like the other voiced letters of the recitation, ja is a root letter and carries no orthographic office. It is never a prefix, never a suffix, and never stacked above or below another letter.

Its shape has one further job. Ja with the small added mark called the tsa phru is dza, the voiced affricate of the fifth line, in the same way that ca becomes tsa and cha becomes tsha. The family resemblance on the page between ja and dza is therefore not accidental, and once a reader has seen the rule the three affricates cost nothing extra to learn.

In practice ja is met most often at the head of very ordinary words and, in a rather different setting, in the transcription of Indic names, where it stands for the Sanskrit letter of the same value. That second use puts it in the company of the letters that had to be pressed into service when Tibetan became a language of Buddhist translation, and it is one of several places where the alphabet has to do work its designers sized it for deliberately.

Within a century or two of its creation the Tibetan script had a task it was not shaped for: rendering Sanskrit exactly, in the names, mantras and technical terms of a translated Buddhist canon. Approximation would not do, because a mantra depends on its sounds and a proper name has to be recoverable. So a second apparatus was built beside the thirty letters.

Three devices carry most of it. The retroflex consonants of Sanskrit, which Tibetan does not have, are written as mirror images of the dental letters, so that a reversed ta writes the retroflex, a reversed na writes the retroflex nasal, and so on. The voiced aspirates, likewise absent from Tibetan, are written as the plain voiced letter with a subjoined ha, so that ga with ha beneath it writes the Sanskrit gha. Long vowels, which Tibetan does not distinguish, are written by subjoining the small a beneath the letter, or by doubling the vowel sign.

None of this entered the recitation. A Tibetan child still learns thirty letters and four vowels, and the reversed forms and the rest are taught afterwards as a specialist extension for a specialist purpose.

The result is a script that can hold two languages at different levels of precision, and a page of Tibetan Buddhist writing in which the ordinary text and the transliterated Sanskrit are visibly different in texture.