Same grandparent, different scripts

Devanagari and Telugu script are both grandchildren of Brahmi, which is why the alphabet charts rhyme row by row while the languages stay strangers. Also: what పాత and పాట know that paata cannot say.

Here is a puzzle this project keeps handing people. Hindi and Telugu are unrelated languages; that case is made in Two unrelated languages, a thousand shared words. Yet set the two alphabet charts side by side and they match row for row: క ఖ గ ఘ lines up with क ख ग घ, ka kha ga gha, the same sounds in the same order, organised by the same logic. If the languages are strangers, why do their alphabets look like family?

Because they are family. Scripts have their own genealogy, separate from the languages they write, and both of these scripts descend from Brahmi, the script of Ashoka's inscriptions in the third century BCE. Devanagari comes down the northern line of Brahmi's descendants. The Telugu script comes down the southern line, through a rounded southern letterform that Telugu and Kannada shared for centuries; the two are usually treated as one Telugu-Kannada script until they drift visibly apart, a parting scholars place somewhere between the thirteenth and fifteenth centuries. The southern line travelled further still: through the Pallava script it seeded writing systems across Southeast Asia. The grandparent has many grandchildren.

Descent from Brahmi is why the two scripts agree on their deep design. Both are abugidas: every consonant letter carries an automatic a, so క is ka and क is ka, and you add a mark to change the vowel or to silence it, the halant in Devanagari, the pollu in Telugu. Both order their letters by the Sanskrit grammarians' phonetic grid: vowels first, then consonants row by row according to where the tongue makes them, from the back of the throat forward to the lips. That chart is not a coincidence, and it is not a borrowing between Hindi and Telugu. It is a shared inheritance from the same source, preserved on both sides for two thousand years.

A script, though, is clothing, not anatomy. English and Finnish are unrelated languages that both wear the Latin alphabet, and nobody concludes from the shared letters that the languages are kin. The same separation runs straight through the subcontinent: Brahmi's descendants write Indo-Aryan languages like Hindi and Marathi, and Dravidian languages like Telugu, Kannada, and Tamil. The script family tree records who adapted letterforms from whom. It says nothing about the ancestry of the words those letters spell.

Why is the correspondence so complete, right down to letters Telugu barely needs? Because both scripts kept the full Sanskrit inventory, and both languages write Sanskrit. Native Dravidian words have no aspirated consonants, yet the Telugu script carries a full set of aspirate letters. They earn their keep writing the borrowed layers: the ఖ of ఖర్చు kharchu and the భ of భోజనం bhojanam are spelling sounds that arrived with loanwords, one Perso-Arabic and one Sanskrit. Each script is sized for the most demanding register of its language, and on both sides that means sized for Sanskrit.

The fit runs the other way too: each script marks exactly what its own language needs. Telugu contrasts short e and o against long ones, a distinction Hindi simply does not have, so the Telugu script has separate letters, ఎ against ఏ and ఒ against ఓ, while Devanagari never developed everyday letters for that split. Same workshop, different tailoring. And this is where romanization, however convenient, starts dropping information on the floor.

Take two entries from this dictionary. పాత means old; it is a native Dravidian word, standing opposite Hindi puraana, which continues the Sanskrit word behind the Puranas. పాట means song; it is also native Dravidian, with close cousins in Tamil and Kannada, standing opposite Hindi gaana. In the everyday texting roman this app uses for search and display, both words come out as paata. The scripts never confuse them for a moment: పాత ends in త, a dental t made at the teeth, and పాట ends in ట, a retroflex t made with the tongue curled back, exactly the cut Devanagari makes between त and ट. The dictionary's reading guide writes the pair as paata and paaTa, spending a capital letter to carry what lowercase ascii cannot.

So the resemblance between the charts is real, and it means something, just not what it seems to mean. The letters line up because the scripts are cousins. The words they spell do not, because the languages are strangers. Bhashalu shows the script, the texting roman, and the reading guide together on every Telugu entry for precisely this reason: the roman gets you typing, and the script tells you the truth.