" firstly, what "really" attracted me to Indo-European, as well as to English, Polish, and Russian philology, wasn't the seductive variety of linguistic forms, or the infinitely picturesque accidents that fill the histories of words and dialects, but rather the fact that these obey lays that can be rigorously described, and that these laws, such as Grimm's Law in Germanic philology, or the principles of Slavic palatalization, which lie behind all those wonderful alveolar fricatives in Russia and the Auvergne, promised to submit the irresistible and etrnal movement of languages no longer to mere chance, but to something that closely resembled calculation;
- and that, secondly, and consequently, the noblest aspect of linguistics (and if I had been familiar with Trouetzkoy's phonology and with Jakobson, this conclusion would have been even more obvious) was its power of deduction -- but that there remained something even nobler, which was the terrain of pure deduction, in other words, mathematics. And that it is why I absolutely had to become a mathematician. "

― Jacques Roubaud