Our editorial method

How this project works

Thirukkural has been translated more than almost any text in human history, and there are already dozens of apps and sites. What makes this one different is not features — it is a set of commitments we hold to on every single line. Here is what they mean in practice.

Every source is public domain — and named

The original Tamil is over two thousand years old and belongs to everyone. For translations and commentary we use only public-domain works we can name and date: G.U. Pope (1886), V.V.S. Aiyar (1916), and Parimelazhagar's 13th-century commentary. We never publish a translation or commentary we cannot freely attribute.

Three translations, side by side

Every Kural shows three English translations at once — never one presented as 'the' translation. Seeing three skilled translators reach for different words is the point: it shows that translating a 2,000-year-old verse is an act of interpretation, not a lookup.

AI drafts. Humans approve. Never the reverse

Modern explanations and translations are first drafted with AI, then read, edited, and approved by a named Tamil scholar before any reader sees them. The scholar's name and the date of approval are recorded and shown. Nothing AI-written reaches the public site without a human signing their name to it.

We never invent Thiruvalluvar's life

We know almost nothing reliable about Thiruvalluvar as a person. So our narratives describe the human situation a verse names — never what he 'must have witnessed.' Every narrative carries the line: we make no biographical claims about Thiruvalluvar; the narrative illuminates the situation the verse describes.

Classical commentary, shown as written

Parimelazhagar's commentary appears verbatim, in his own 13th-century Tamil — we do not modernise or rewrite his words. Where we add a plain-English explanation or a plain modern-Tamil paraphrase to make it accessible, it sits beside the original, clearly marked as an AI draft pending scholar review.

Provenance on every line

At the bottom of every Kural is a short, expandable note showing exactly where each piece came from — the source and licence of each translation, the classical commentator, and, for the modern narrative, which AI drafted it, when, and which scholar reviewed and approved it. If any of that is missing, our build refuses to publish the page. That is the chain of trust you can check for yourself.

What we deliberately don't do

— No advertising, no tracking, no selling of data.
— No engagement-hacking — wisdom literature is not a game.
— No chatbot pretending to be Valluvar.
— No hidden sources: every line stays traceable to a named, public-domain origin.