Numbers were a fetish, numbers for their own sake. What could be done with them? They were supposed to be a guide to legislation. There was the nascent idea of statistical law, but hardly any statistical inference. Yes, one could conclude that the French are more prone to suicide than the English. Yes, Guerry could invent (almost without knowing it) rank-order statistics to argue that improved education does not counter crime rate. But hardly anyone sensed that a new style of reasoning was in the making.

Medicine is a good example. The statistics of Paris were full of tables reporting the great hospitals. Would not these batteries of numbers lead at once to tests of treatment and cures? Not at all. When the numbers were used, it was more out of professional jealousy than in a quest for objective knowledge.

The first use of statistical data to evaluate treatment was, it appears, in connection with the charismatic and polemical F.-J.-V. Broussais, whose belief in a ‘stay-alive’ organ I mentioned briefly in chapter 8. He will figure in chapter 19 because of the way that Comte transferred his physiological conception of the ‘normal state’ to society and politics.