The History of AI

Théorie analytique des probabilités

Era 1 · The foundations (1763–1942) › Probability and statistics

Lineage
Data and learning
Kind
Publication

Who

What happened

Pierre-Simon Laplace publishes Théorie analytique des probabilités, the treatise that brings the probability theory of the day together into a single system.

Why it mattered

It formalizes and generalizes Bayes's result, which Laplace had reached independently in 1774, and combines it with least squares and generating functions into a general theory of probability and its applications.

The full story

Pierre-Simon Laplace, born in Normandy in 1749, was one of the great scientists of Napoleonic France, known above all for celestial mechanics. In 1774, at twenty-five, Laplace had independently arrived at the relation between prior and posterior probability that Bayes had set out, apparently without knowing Bayes's essay.

In 1812 Laplace gathered almost forty years of work into the Théorie analytique des probabilités. The first book deals with generating functions and with methods for approximating calculations involving very large numbers. The second defines probability and develops Bayes's rule, inverse probability, which reasons from effects to causes, along with the method of least squares, with applications such as mortality and life expectancy.

The first edition was dedicated to Napoleon; the dedication vanished from later ones. The 1814 edition added nearly 30 percent more material and a Philosophical Essay on Probabilities, based on Laplace's lectures of 1795 at the École Normale, which explained the theory without formulas, along with its uses in games of chance and in the natural and moral sciences. New supplements followed in 1820 and 1825.

For more than a century, the probability used to reason about causes was Laplace's, even though the rule ended up bearing Bayes's name.

Builds on

  • Bayes's essay (December 23, 1763)

    Independent convergence

    In 1774 Laplace arrives at the same relation between prior and posterior probability without knowing Bayes's essay

  • The method of least squares (1805)

    Inherited, with no proven direct link

    Laplace's treatise folds the method of least squares into Laplace's general theory of probability

Leads to

  • Maximum likelihood (April 19, 1922)

    Inherited, with no proven direct link

    Laplace and Gauss had used maximum likelihood in particular cases; Fisher turns it into a general method of estimation

  • Grundbegriffe der Wahrscheinlichkeitsrechnung (1933)

    Inherited, with no proven direct link

    The axioms replace Laplace's classical definition, favorable cases over possible cases, with a measure on a set of events

Sources

How to cite

“Théorie analytique des probabilités.” The History of AI, revised November 13, 2026. https://thehistoryofai.com/en/milestone/laplace-analytic-theory

This entry was last reviewed on November 13, 2026.

Open in the interactive timeline