Kannada wasn't ‘born' from Tamil. The truth is much more interesting

Indian Express

30-05-2025

Latin had its roots in the Aeolic dialect of Greek. Or rather, that's what some scholars in antiquity, chiefly Greeks as one might expect, believed. Modern historical linguistics shows this notion, known as Aeolism, to be false; the two languages merely share a common ancestor, Proto-Indo-European. In India, the 15th-century Manipravalam grammar, Lilatilakam, hewed a little closer to the truth in its speculation: The author groups the languages of the Kerala, Pandya and Chola regions under the name 'Dramida', but rejects the inclusion of the 'Karnata' and 'Andhra' tongues as they were too distant from the language of the 'Tamil Veda', that is, the Tiruvaymozhi. He does admit, however, that others would include the latter two languages in the category; who these others were, and what exactly they thought, is lost to time. Such passing references offer us tantalising glimpses of a vanished world of Brahminical scholarship, of historical-linguistic imaginations that by far antedated Robert Caldwell's 'discovery' of the Dravidian language family.
Let's set this against another piece of historical evidence: A mediaeval prashasti, or praise-poem, from Venadu, a kingdom that controlled parts of southern Kerala and Tamil Nadu, which later became Travancore. Dating to the early 1100s, it praises a king of Venadu for his defeat of a Pandya lord, Rajasimha, who is described as 'Tamil', implying that his rival was in some sense not Tamil. Perhaps it's a question of Tamil versus Malayalam, as the former had split off from Early Middle Tamil a few centuries prior. However, this particular poem is not in Old Malayalam but in standard literary Tamil. Malayalis would also continue to call even their own increasingly distinct language 'Tamil' for centuries to come. It's a reminder that questions of language on the one hand, and ethnic identity on the other, have long been complex matters in 'Dravidian' South India.
Amid the brouhaha over actor Kamal Haasan's statement that Kannada 'was born out of Tamil', it's important to set a few things straight. Although there are still yawning chasms in scholarship on the Dravidian family, and more work is urgently needed, the taxonomy of at least the literary South Dravidian languages is broadly established. The split between Kannada and Tamil is older than any of the surviving literary or epigraphic evidence. Their common ancestor, itself a descendant of Proto-South-Dravidian, is conventionally termed Proto-Tamil-Kannada, but we don't know what its own speakers called it. Both branched off from this unattested language; it is meaningless to say that Kannada came from Tamil or Tamil from Kannada. In general, it's also meaningless to say that any natural language (with some exceptions such as sign languages) is 'older' than any other, as ancestral forms of both would have been spoken at any given point in time.
The surviving corpus of Tamil literature is likely older than its Kannada counterpart, dating to the last few centuries BCE or first few centuries CE, despite revisionist attempts to cast the Sangam poems as an elaborate faux-historical forgery created at the court of the early mediaeval Pandyas; such claims fly in the face of all logic and plausibility. That still doesn't say anything about the age of the languages themselves. Another claim is that Tamil is more conservative than related languages, preserving older forms, but counterexamples are easy to find (even ignoring the advantage Tamil has by dint of being attested in earlier forms):
The Tamil word for 'ear' is cevi, whereas Kannada has kivi, preserving the older Dravidian *k (reconstructed form).
That said, several mysteries remain in the classification of the Dravidian languages, the best known being the debate over the origins of Brahui, spoken largely in Pakistan and to an extent in Afghanistan and Iran. The question is whether Brahui arrived in this region in mediaeval times or it descended from a language that was present there thousands of years ago, when Dravidian languages were far more widely spoken. The answer could, perhaps, tie into the origins of the Dravidian family itself, and whether it was spoken in the Indus Valley Civilisation. The classification of the common ancestor, Proto-Dravidian, is another open question, with theories such as the one positing a connection to Elamite, spoken in Iran before it was Iran, having found little support.
The controversy over Haasan's comments reveals the minefield of competing ethno-linguistic claims that any appeal to Dravidian brotherhood must navigate to be successful. To rise above it would require a politics of regional aspiration and counter-hegemony that is shorn of chauvinism and claims of historical priority, leaving the linguistics to the linguists.
[e-mail]