JOSHUA LEDERBERG: Well my first important discovery came about through a rather unusual process. More than almost everything
else that I've ever done, it was theory driven rather than being data driven. It was theory driven in the sense that a
postulate had arisen out of the course of examination of the contemporary scene, namely Avery et al had shown transfer of
heritable characteristics in a bacterium. One was deeply motivated to want to know more about whether there were things like
genes in bacteria. The approach that speculatively arose and how to respond to it is: could one found out that there is a
possibility of genetic recombination, of crossing between bacterial cells? My work with Neurospora had given me the theoretical
basis -- one was able to put together a thought experiment that said, "if bacteria can be crossed, and if you start out
with two different nutritional mutants, and if they exchange with one another, they will form prototrophic genotypes, you
will be able to select for them, and define their occurrence even if they happen very very rarely by a rather simple procedure
-- plating the mixtures on minimal agar". So here was a case where the experimental design was worked out -- it was a
theoretical postulate: does genetic recombination occur? Does it not occur? And that was all worked out in advance of ever
doing the experiment.
Well somewhat to my surprise it worked very very promptly. I was rather fearful when the first positive results came in. I
was worried deeply that this was going to be an artifact, that my hopes would be dashed. I didn't want to get too excited
about it. I was scared. I knew it was a very important finding, but I didn't want to be out on a limb until I could be
absolutely certain and I didn't want to commit myself emotionally to the consequence until then. So it was a matter of
going back to the grindstone and repeating the experiment many times, doing it different ways and being sure it was a totally
reliable and reproducible result. Put in every control that I could think of to be sure it had controlled for possible artifacts.
Happily, with bacteria one can do these experiments, you can run two or one or two cycles a day in this kind of experimentation
so within a few weeks it was possible to get a total validation of that.
All that happened almost precisely 50 years ago. I arrived in New Haven [Connecticut] in March 1946 and by early June had
it completely taped down that crossing was taking place.