Tibetan Mark Tsa-Phru

It is a combining mark, not a word.

Tsa-phru is a flag-shaped part of Tibetan writing that can be attached to a consonant or written separately in an abbreviation. Its reading depends on the surrounding Tibetan letters, so by itself it is only a fragment.

Tibetan abbreviations appear in the earliest known written examples, from the mid-eighth to mid-ninth centuries. More specialized contractions developed after the Tibetan Empire fragmented, especially in scripts without a horizontal top stroke. An early-fourteenth-century manuscript of the Narthang Kangyur, a collection of Tibetan Buddhist scriptures, preserves many simple forms; woodblock editions from the fifteenth century used fewer in some cases. These contractions saved manuscript space, time, ink, and paper.

How does ༹ change a Tibetan sequence?

The consonant beside it determines what it contributes.

The flag-shaped stroke is part of the Tibetan consonants ཙ, ཚ, and ཛ. Written separately, ༹ can stand for one of four abbreviated syllables, conventionally romanized as tsa, tsha, dza, or za. In some spellings of foreign names or loanwords, ཕ༹ represents the f sound as in “fan,” while བ༹ represents its voiced partner, the v sound.

A 2003 Unicode proposal said this foreign-sound route was not used by Tibetan writers in the PRC and gave ཧྥ as an alternative spelling for the f sound. These are different conventions: read the whole sequence, not the floating mark alone.

Why does the mark render wrong after a vowel?

Normalization can move it away from the consonant it modifies.

In the encoded sequence, ༹ belongs immediately after its base consonant, but canonical normalization can reorder it after a following Tibetan vowel. A Unicode technical note in 2005 described this ordering problem; a 2006 report found that the reordered sequence could display incorrectly in Windows Vista with Uniscribe.

If the mark looks misplaced before a vowel, compare the original sequence with the normalized text in the font and text shaper you are using.