marian by means of his writings. The condition of Tamil (or any other living language) one thousand years ago was not what it had been a thousand years still earlier. And its grammar, which is essentially an empirical or inductive science, necessarily varies with the conditions of that language. In any language, literature always precedes grammar ; and this fundamental principle was not unknown to the early Tamil grammarians, who have explained it in unmistakable terms thus :—
இலக்கி யத்தினின் றெடுபடு மிலக்கணம். -Agat.
இலக்கியங் கண்டதற் கிலக்கண மியம்பல்.-Nan.
(Literature yields the grammar ; grammar follows the literature.)
They have also recognised the principle of change in a living language, and provided for popular acceptation of innovations.
கடிசொ லில்லைக் காலத்துப்படினே.-Tol.
(Usage sanctifies any new word.)
பழையன கழிதலும் புதியன புகுதலும்
வழுவல கால வகையி னானே.—Nan.
(The order of things is for the old to give place to the new.)
Thus the statement of Sivagnanaswami that, தொல்லாசிரியர் வழக்கே வழக்கு. பிற்காலத்து வேறு படவழங்கப்படுமாயின் அவ்வழக்கு இலக்கணத்தோடு பொருந்தாதென விலக்குக, is not only unscientific, but also an obstinate clinging to that old superstition which believed the ancients to have discovered all wisdom.
