| # | Vowel Sign (Matra) | Name | Example with क (ka) | Transliteration | |---|---|---|---|---| | 1 | (none) | अ (inherent) | क | ka | | 2 | ा | आ (aa) | का | kaa | | 3 | ि | इ (i) | कि | ki | | 4 | ी | ई (ee) | की | kee | | 5 | ु | उ (u) | कु | ku | | 6 | ू | ऊ (oo) | कू | koo | | 7 | ृ | ऋ (ri) | कृ | kri | | 8 | े | ए (e) | के | ke | | 9 | ै | ऐ (ai) | कै | kai | | 10 | ो | ओ (o) | को | ko | | 11 | ौ | औ (au) | कौ | kau | | 12 | ं | अनुस्वार (anusvara) | कं | kam |
The arrival of Unicode in the 1990s posed a challenge: how to encode a 396-cell matrix efficiently? Unicode solved this by encoding the base consonants (U+0915 for क) and vowel signs (U+093E for ा) as separate code points. Rendering engines (e.g., HarfBuzz, DirectWrite) dynamically combine them using Glyph Substitution tables.
Learning to read and write in any language requires internalizing a code. For languages using the Latin script, a child learns 26 discrete symbols (graphemes) and the arbitrary sounds (phonemes) they represent. However, for over 1.5 billion speakers of languages such as Hindi, Marathi, Nepali, Punjabi, and Gujarati, the foundational literacy unit is not the alphabet, but the Barakhadi .
