Mathematical Conversations

Grothendieck has proposed, under the name of "motives" a kind of Galois theory for non algebraic numbers. A mystery of the so-called "standard model" in high energy physics is the occurrence of about twenty numerical constants, independent of the choice of any unit, and whose determination rests solely on data-fitting. I propose the beginning of an explanation by introducing a new kind of symmetry group acting on these constants. In a way this symmetry group should be an extension of the so-called "renormalisation group" a la Connes-Kreimer.