Inferring and Explaining

63 hands of the doctors and students, something had to be done to stop this from happening in the future. Although nothing was known of the true nature of infectious disease, Semmelweis and his colleagues knew that chloride solutions had long been used to rid objects of the noxious odor of putrid materials. Semmelweis reasoned that a chloride solution would be the ideal substance to destroy the foul-smelling cadaver par- ticles. In the middle of May 1847, he ordered that a bowl of chlorine liquida , a dilute concentration of the disinfectant, be placed at the entrance to the First Division, and he insisted that every entering medi- cal attendant wash in it before touching a woman in labor. Small, stif brushes were kept nearby, to be used for cleaning under fngernails. 7 Although medical research was not his pri- mary goal at this juncture, it is quite natural to interpret Semmelweis’s actions as an inter- esting experiment designed to test his new hypothesis. One can imagine a bizarre and evil experiment that could have been used to test the cadaver particles theory. He might have ran- domly chosen twenty pregnant women and sep- arated them into two groups. To ten, he might have intentionally introduced cadaver particles into their bloodstreams; to the other ten, the “control,” he would have scrupulously allowed no entry of cadaver particles. He would have then waited to see if the ten he predicted would contract childbed fever did, while the other ten did not. Tankfully, this was not his experimen- tal procedure. He might, a little more sanely, have also conducted a similar experiment with laboratory animals, but again, his focus was on saving lives. His order of chlorinated lime, though, pro- duced some stunning new data: semmelweIs and ChIldBed feVer e 12 . Semmelweis ordered the chlorinated lime procedure in May of 1847. e 13 . By 1848, the death rate in the First Divi- sion from childbed fever had fallen to 1.2 per- cent, just a tick less than the Second Division, at 1.3 percent. Semmelweis’s Evidence Let us pause for a moment and use inference to the best explanation to assess the quality of Semmelweis’s evidence. Tere is a great deal of evidence to schematize. e 1 : e 13 t 0 Tere are also a number of rival explanations that had been discussed and partially tested. t 1 : t 5 When we now add t 0 to this list and rank order all them in terms of the best explanation, we would all agree, I trust, that t 0 is by far the best explanation and that Semmelweis’s evidence was quite overwhelming. I can imagine some of you seeing things dif- ferently. You are sophisticated about the true nature of infectious diseases such as childbed