Drunkard's Walk - Leonard Mlodinow [79]
The normal distribution describes the manner in which many phenomena vary around a central value that represents their most probable outcome; in his Essai philosophique sur les probabilités, Laplace argued that this new mathematics could be employed to assess legal testimony, predict marriage rates, calculate insurance premiums. But by the final edition of that work, Laplace was in his sixties, and so it fell to a younger man to develop his ideas. That man was Adolphe Quételet, born in Ghent, Flanders, on February 22, 1796.14
Quételet did not enter his studies spurred by a keen interest in the workings of society. His dissertation, which in 1819 earned him the first doctorate in science awarded by the new university in Ghent, was on the theory of conic sections, a topic in geometry. His interest then turned to astronomy, and around 1820 he became active in a movement to found a new observatory in Brussels, where he had taken a position. An ambitious man, Quételet apparently saw the observatory as a step toward establishing a scientific empire. It was an audacious move, not least because he knew relatively little about astronomy and virtually nothing about running an observatory. But he must have been persuasive, because not only did his observatory receive funding, but he personally received a grant to travel to Paris for several months to remedy the deficiencies in his knowledge. It proved a sound investment, for Quételet’s Royal Observatory of Belgium is still in existence today.
In Paris, Quételet was affected in his own way by the disorder of life, and it pulled him in a completely different direction. His romance with statistics began when he made the acquaintance of several great French mathematicians, including Laplace and Joseph Fourier, and studied statistics and probability with Fourier. In the end, though he learned how to run an observatory, he fell in love with a different pursuit, the idea of applying the mathematical tools of astronomy to social data.
When Quételet returned to Brussels, he began to collect and analyze demographic data, soon focusing on records of criminal activity that the French government began to publish in 1827. In Sur l’homme et le développement de ses facultés, a two-volume work he published in 1835, Quételet printed a table of annual murders reported in France from 1826 to 1831. The number of murders, he noted, was relatively constant, as was the proportion of murders committed each year with guns, swords, knives, canes, stones, instruments for cutting and stabbing, kicks and punches, strangulation, drowning, and fire.15 Quételet also analyzed mortality according to age, geography, season, and profession, as well as in hospitals and prisons. He studied statistics on drunkenness, insanity, and crime. And he discovered statistical regularities describing suicide by hanging in Paris and the number of marriages between sixty-something women and twenty-something men in Belgium.
Statisticians had conducted such studies before, but Quételet did something more with the data: he went beyond examining the average to scrutinizing the manner in which the data strayed from its average. Wherever he looked, Quételet found the normal distribution: in the propensities to crime, marriage, and suicide and in the height of American Indians and the chest measurements of Scottish soldiers (he came upon a sample of 5,738 chest measurements in an old issue of the Edinburgh Medical and Surgical Journal). In the height of 100,000 young Frenchmen called up for the draft he also found meaning in a deviation from the normal distribution. In that data, when the number of conscripts was plotted against