The grammarian who named the layers
Kannada never descended from Sanskrit, but its formal register is saturated with it, and Kannada’s own grammarians sorted the layers in the thirteenth century. On tatsama, tadbhava, achchagannada, and the vowel that ends almost every borrowed noun.
Listen to a formal Kannada speech and, if you know Hindi, you will keep catching words: samaya, desha, samudra, sangeeta, pustaka. The natural conclusion is that Kannada must be related to Hindi. The conclusion is wrong, but the observation behind it is entirely right, and it deserves a real explanation rather than a wave of the hand.
Kannada supplies its own explanation, and did so long before European philology arrived to ask. Around the middle of the thirteenth century a grammarian named Kēśirāja wrote the Śabdamaṇidarpaṇa ಶಬ್ದಮಣಿದರ್ಪಣ, a grammar of Kannada composed in Kannada verse. It sorts the language’s vocabulary by origin, and three of its categories do all the work here. Tatsama ತತ್ಸಮ, literally “same as that,” is a Sanskrit word taken into Kannada essentially whole. Tadbhava ತದ್ಭವ, “born of that,” is a Sanskrit word reshaped by Kannada sounds on its way in. And the native Dravidian stock is dēśya, which Kannada also calls achchagannada ಅಚ್ಚಗನ್ನಡ, pure Kannada. The origin labels in this app are not imported jargon. They are Kannada’s own vocabulary for its own vocabulary.
The doublets are the proof, because they show one Sanskrit word arriving twice. Bhaashe ಭಾಷೆ and baase ಬಾಸೆ both go back to Sanskrit bhāṣā, language, but baase has lost the aspiration and the retroflex sh. Akshara ಅಕ್ಷರ and akkara ಅಕ್ಕರ, letter, split the same way, the cluster kṣ flattening to kk. Vidye ವಿದ್ಯೆ and bijje ಬಿಜ್ಜೆ, learning, part company even further. Mukha becomes moga ಮೊಗ, face; kaarya becomes kajja ಕಜ್ಜ, work. One arrival came through the book and kept its shape; the other came through the mouth and was worn to fit. Old Kannada poetry is full of the worn forms; modern prose usually reaches for the intact ones.
There is a second, quieter adaptation that catches a whole class of Sanskrit nouns on the way in, and it is the single best hook a Hindi speaker can learn. Sanskrit words ending in long -ā end in -e in Kannada. Maalaa becomes maale ಮಾಲೆ, a garland. Aashaa becomes aase ಆಸೆ, desire. Kathaa becomes kathe ಕಥೆ, a story. Shaalaa becomes shaale ಶಾಲೆ, a school. And bhaashaa भाषा becomes bhaashe ಭಾಷೆ, language, which is where this app got its name: bhaashegalu ಭಾಷೆಗಳು, languages, is bhaashe plus Kannada’s own plural. Hindi keeps the -a; Kannada turns it to -e; once you hear it, you hear it everywhere.
The rule comes with a trap, and this project walks into it deliberately so that readers do not. The -e ending on its own does not make a word tadbhava. Kannada’s own tradition treats it as the regular way the language inflects a Sanskrit ā-stem, not as reshaping. Bhaashe keeps its aspirated bh and its retroflex sh intact, so bhaashe is tatsama, a Sanskrit word taken whole and given a Kannada ending. Baase, which lost both, is the tadbhava. Bhashegalu classifies whichever form the entry actually uses, and the note points at the other when there is one.
Register is where all of this becomes audible. Everyday Kannada runs on achchagannada; formal Kannada reaches for tatsama. Flower is hoovu ಹೂವು at the market and pushpa ಪುಷ್ಪ in a hymn. Water is neeru ನೀರು at home and jala ಜಲ in an inscription. House is mane ಮನೆ and gruha ಗೃಹ. Eye is kannu ಕಣ್ಣು and netra ನೇತ್ರ. English does exactly the same thing with its own layers: kingly is native, royal came with the Normans, regal came straight from Latin. Stack hoovu against pushpa and you are hearing the same social physics.
Why did Kannada take on so much? Prestige, scholarship, and above all literature. The Kavirājamārga ಕವಿರಾಜಮಾರ್ಗ, a ninth-century work on poetics generally taken as the earliest surviving Kannada literary text, is already arguing about how Kannada should be written, and court, temple, and school kept the Sanskrit channel open for the thousand years since. The best evidence that the Sanskritized register was a real weight is that people pushed back against it: the vachana ವಚನ writers of the twelfth century chose deliberately plain spoken Kannada, and the choice was visible enough to be a statement.
One precision matters more than it looks, because the same two words mean different things on either side of this app. Hindi grammar also uses tatsama and tadbhava, but in Hindi a tadbhava word is one Hindi inherited from Sanskrit and wore down over centuries of ordinary speech. In Kannada a tadbhava word is one Kannada borrowed and reshaped on entry, because Kannada never descended from Sanskrit in the first place. Hindi diya is an inherited Sanskrit lamp; Kannada deepa ದೀಪ is a borrowed one. Same terms, two entirely different mechanisms, and confusing them is the fastest way to turn contact into imagined ancestry.
None of this makes Kannada an Indo-Aryan language, and the grammar is the proof. Borrowed nouns obey Kannada, not Sanskrit. Pustaka takes the Kannada plural and becomes pustakagalu ಪುಸ್ತಕಗಳು. Loans take Kannada case suffixes, sit in verb-final Kannada sentences, and submit to Kannada’s own gender system, which is semantic: men are masculine, women are feminine, and a table and a door are simply neuter because they are not people. Hindi assigns gender by convention and makes you memorize it. Kannada follows nature. English is a Germanic language wearing a very large Latin wardrobe; formal Kannada is a Dravidian language wearing a very large Sanskrit one. The wardrobe is real. So is the skeleton underneath, and the skeleton decides the family, as Unrelated languages, familiar words argues at length.
