Beneath the root letter of a Tibetan syllable there may be a subscript, the btags, and four letters may take that position: ya, ra, la and wa. No others.
Each behaves differently, and none of them simply adds its own sound. Subjoined ya changes the letter above it, so that in modern Lhasa speech a labial with a ya beneath it is heard as a palatal. Subjoined ra likewise alters the consonant above it rather than being pronounced as an r. Subjoined la produces a value of its own that has little to do with either letter. Subjoined wa is generally inaudible.
The subscript is one of four positions in a syllable stack, and the positions are fixed: superscript above, root at the centre, subscript below, vowel sign above or below the whole, with a prefix before and a suffix or two after. A full stack can therefore run to six or seven written elements and is still one syllable, closed by a single tsheg.
Learning the four subscripts is not optional even for a reader who never writes, because they change what a stack sounds like and because they affect where a word is filed in a dictionary. The order of comparison in Tibetan lexicographic sorting runs root letter, then superscript, then subscript, then vowel, then suffix, so the subscript is a sorting key as well as a piece of pronunciation.