258 THE EVOLUTION OF SCIEXTIFIC THEORIES 



tative data come in question. Pasteur and Tyndall maintain the 

 nonoccurrence of spontaneous generation. Pouchet and Bastian main- 

 tain that it does occur. All agree that nutrient media boiled for a long 

 time will remain sterile in contact with sterile air, and that media 

 boiled only a short time may in these conditions develop an abundant 

 microbial life. Pasteur and his cohorts argued that in the second case 

 the medium simply had not been sterilized, as it could have been by 

 longer boiling. Pouchet and his partisans argued that in the first case 

 the "vegetative force" had been destroyed by long (or repeated) 

 boiling, so that the medium was no longer capable of engendering 

 life. No experiment then performable could resolve this issue, for 

 Pasteur could no more prove the presence of spores in the second 

 case than Bastian could prove the presence of a vegetative force in 

 the first. 



During the first half of die 20th century we find an analogous pro- 

 tracted dispute over whether the crystalline viruses are or are not 

 "alive." No crucial test was feasible simply because these viruses 

 could be grown only in culture media containing living cells. One 

 could then equally well maintain either that viruses are alive but 

 require this special culture medium or, alternatively, dead but capa- 

 ble of commandeering the metabolic machinery of living cells for the 

 proliferation of virus material. Once again there is no dispute about 

 facts, perhaps resolv^able by experiment, but dispute only about in- 

 terpretations of facts, not readily so resohable. Here, as in many 

 other such cases, the dispute is ultimately resolved through the dis- 

 sipation of a once-sharp dichotomy. Still other cases may ultimately 

 be settled with the contrivance of more searching experiments. But 

 what may happen itltimatehj is completely irrelevant to the fact that 

 in situations of deadlock no crucial experiment can help us reach a de- 

 cision now. 



The ramified chain of reasoning. When well accredited ideas are 

 challenged by the result of a purportedly "crucial" experiment, our 

 first impulse is always to question the fact as such. May it reflect no 

 more than illusion or hallucination, as displayed in reports of flying 

 saucers? May it reflect human error— observational error in reading 

 the scales of a telescope, or calculational error in computing the 

 orbit of some satellite? May it reflect some kind of instrumental error 

 —defective graduation of the telescope scales, or some uneven settle- 

 ment of the foundation for that telescope? Or may it reflect some 



