Nature abhors a vacuum: the schools taught it on Aristotle's authority, and held that empty space cannot exist. In Rouen a sickly young man decided the question could be settled by a measurement. Working from Torricelli's mercury tube, he had his brother-in-law Florin Périer carry a barometer up the Puy-de-Dôme in September 1648 — he was too ill to make the climb himself — and the mercury fell as the mountain rose. Air has weight. The schoolmen were wrong, and an experiment said so. Educated Europe was meanwhile being told that certainty had one shape and that shape was geometry, Descartes having made deduction from clear first principles the model for anything worth knowing.
He was the century's great demonstrator: an essay on conic sections at sixteen; a working mechanical calculator, built to spare his father the arithmetic of tax assessment and patented in 1649; and in 1654, an exchange of letters with Pierre de Fermat about how to divide the stakes of a gambling game broken off before the end, out of which came mathematical expectation, the first machinery for reasoning about what cannot be known.
Then, on 23 November 1654, from about half past ten at night until half past midnight, something happened to Pascal. He wrote it down on a scrap of paper, sewed the paper into the lining of his coat, and moved it from garment to garment for the rest of his life; a servant found it eight years later while preparing his body. It begins with the word for fire, and it sets the God of Abraham, Isaac and Jacob against the God of the philosophers and the scholars.
That distinction is the argument of everything he wrote afterwards. Reason cannot deliver the thing that matters most — and the claim carries because of who is making it, a man who had just spent a decade showing exactly how much reason can deliver.
So he turned his own instrument on the question. The wager, the most famous fragment he left, is not a proof that God exists, and he says plainly that it is not one. It is the calculus of expectation aimed at a case where the evidence has run out. You are already staking your life on an answer; weigh a finite loss against an infinite gain and see what the arithmetic recommends. Readers have been arguing with it since 1670, when the abbé de Villars answered that it reduces faith to self-interest — the objection Voltaire and most of the Enlightenment repeated, and one still made.
Taking sides cost him: after the Sorbonne expelled Antoine Arnauld for heresy, Pascal published eighteen anonymous letters between January 1656 and March 1657 in defence of the Jansenists of Port-Royal. The Provincial Letters made moral theology funny, sold across France, and were condemned in Rome and put on the Index, where they went on selling. In 1661 the crown required every French religious to sign a formulary condemning Jansenist doctrine. Pascal refused, argued in writing that signing meant condemning Augustine, and broke with Arnauld for yielding. His sister Jacqueline, a nun at Port-Royal, signed under protest and died that October.
He died on 19 August 1662, aged thirty-nine, five months after the five-sol carriages he had designed began running the first scheduled public transport in Paris. The great defence of Christianity he had been building was a heap of loose paper. Port-Royal published a tidied, softened version in 1670; editors have disagreed about the order of the fragments ever since, and every edition on sale is one editor's arrangement of them.