Chitrakāvya चित्रकाव्य

The thesis

Reading classical Indian poetry as a computational object.

Chitrakāvya — literally picture-poem — is a tradition of Sanskrit verse that survives only when read in two media at once: as sound, and as shape. The poet writes a śloka whose syllables, when laid out on a grid, form a wheel, a lotus, a drum, a sword, or a chessboard, and whose readings trace paths through that figure: along petals, around spokes, in palindromic symmetry, or along the L-jumps of a knight. The constraints are severe and the meanings are precise. The same artefact is a poem, a diagram, and an algorithm — three faces of one composition.

This project — originally hosted by the Computational Sciences Laboratory at IIIT-Bangalore — collects these objects, restores their broken visualisations, and treats them as exhibits in a longer history of computation as form. The site you are reading is a fresh build of that archive, with new illustrations, working interactive demos, and references repaired against Wikipedia, peer-reviewed journals, and open-source code.

स्थितः समः सरः मर्मा रामा रासः समः सर: ।
सरः समः सरामारा मार्मरः समः स्थितः ॥ — Sarvatobhadra: each line a palindrome; the whole stanza a fixed point under reflection.

Enter the project

Five doors into the work.

A short orientation

Three families of artefact.

The claim

A new computing paradigm — older than its name.

We tend to read the history of computing as a march of mechanisms — abacus, clockwork, valve, transistor, GPU. But computation is not only the manipulation of state; it is the act of specifying a structure precisely enough that someone else can reproduce it. By that older definition, a chitrakāvya verse is a program: an exact recipe for a shape, a path, and a meaning, encoded in a substrate (Sanskrit) chosen because its grammar permits the recipe to compile.

This site treats the corpus as exactly that — runnable specifications. Each artefact is presented with three things: the verse itself, a reconstructed visualisation, and a pointer to the modern algorithmic problem it encodes (Hamiltonian path, magic square, palindrome, lipogram, constrained composition). The intention is research, not romance: where a primary source is contested, we say so, and we link the strongest secondary scholarship we can find.