Google hits for "Pascal-Fermat correspondence": 3,230. For "Fermat-Pascal correspondence": 206.
Does anyone have any idea why? In particular it seems like one would sort either on importance (but which of these two is more important?) or alphabetically (but that gives the wrong result).
And given two individuals X and Y, how can we predict whether "X-Y correspondence" or "Y-X correspondence" is more common? I'm sure there are no hard and fast rules here but there must at least be some trends, and would expect a situation similar to how gender is assigned to words in languages, foreign to me, where words have gender.