Sa: The Dental Sibilant and the Breath That Replaces It

Sa is the thirty-second consonant of the varnamala, the dental sibilant and the last of the three, made with the tongue at the teeth. It is the commonest of the sibilants and the one an English speaker produces without adjustment. Its behaviour at the ends of words is one of the more intricate parts of Sanskrit sandhi: a word-final dental sibilant is very often replaced by the visarga, a breath written as two dots to the right of the letter, and what it becomes before a following word depends on the sound that word begins with. Recovering the underlying sibilant from a visarga is a routine part of reading Sanskrit and one of the earliest things a learner is drilled on, because word-final sibilants are extremely common in the nominal endings.

The letter is also the source, by rule, of most occurrences of the retroflex sibilant, in the same way the dental nasal is the source of most retroflex nasals. Two of the three retroflex letters outside the stop classes are therefore derived rather than inherited, which says something about how much of written Sanskrit is the visible output of the language's own machinery. The letter also carries a large share of the language's grammatical information, since the sibilant appears in the endings of several noun cases and in the marks of two tenses. A reader who mishandles it at a word join therefore misidentifies the grammatical form as well as the sound, which is why the rules governing it are drilled early and at length.