The Batak script has five vowels, six including the Schwa, but there are only three separate characters for the vowels [a], [i] and [u]. There are no ina ni surat characters to render syllable initial /e/, /ə/, or /o/.
How then do we write them if they appear in initial position of a word or if they follow a vowel?
We have to do this by using ᯀ and combine it with either /e/, /ə/, or /o/.
For example the words eda ‘sister in law’ and olo ‘yes’ must be written ᯀᯩᯑ and ᯀᯬᯞᯬ. The Schwa [ə] is an important sound in Karo and Pakpak: enggo ‘already’ is written ᯁᯧᯰᯎᯭ.
We can also apply that to /i/ or /u/! Whether we use ᯤ and ᯥ is entirely optional! Sai can be written either ᯘᯤ or ᯘᯀᯪ and even ᯘᯛᯪ. Udan ‘rain’ can be written ᯀᯮᯑᯉ᯲, and uli ‘beautiful, good’can be written ᯀᯮᯞᯪ. There is not even consistency between writers. It is quite common to find one word written once with ᯤ and then again in the same manuscript with ᯀᯮ!
So why do we have separate characters for [i] and [u], but not for [e] and [o]? There is no conclusive answer, but what we know is that words with initial [i] and [u] are about twice as frequent as words with initial [e] and [o]. In the Kerinci script, which is geographically the next nearest surat script, there is not even a diacritic for [e] and [o]! The Kerinci language has six vowels, but only four can be written! The diacritic for /i/ is also used to render /e/, and the diacritic /u/ is also used for /o/.