A frequent commentary chronicling the creative and intellectual
excitement of discovery at Washington State University.

Brought to you by Washington State Magazine

As an Experiment, Leno at 10 Could Be Seen as a Success

Write “Conan,” “Leno,” and “failed experiment” in a search engine and you’ll get several thousand results. Apparently, there are a lot of people out there who think NBC was conducting an experiment when it moved Jay Leno from “The Tonight Show” to prime time, with Conan O’Brien taking his place at 11:35 p.m.

But was it really an experiment? And even if Leno’s prime-time appearances failed to bring profitable audiences, is that a failure in a scientific sense?

We asked Jerry Gough, who teaches the history of science at Washington State University. His response:

The problem here, I think, is that (according to the Oxford English Dictionary) there exist two quite distinct definitions of the word “experiment”, one of which applies to theory (“An action or operation undertaken in order to discover something unknown, to test a hypothesis, or establish or illustrate some known truth”) and the other of which applies to practice (a tentative procedure; a method, system of things, or course of action, adopted in uncertainty whether it will answer the purpose).

In the pure sciences (whose aim is simply to ascertain the truth), I am not sure that there can be such a thing as a failed experiment (as long as it is properly conducted—e.g., there was no hole in the filter paper, no one dropped the flask or added a destructive chemical by mistake).  Experiments are usually conducted to test pre-conceived hypotheses.  If the expected results do not come about, this does not mean the experiment has failed, but either that the hypothesis being tested needs correction or that the experiment did not really test the hypothesis in the way that the experimenter thought it did. In either case, something new is learned, and since learning is the goal of the operation, the experiment was in some sense “successful.”

NBC’s Jeff Zucker is no Antoine-Laurent de Lavoisier

An example of this was the case of great chemist Antoine-Laurent de Lavoisier who in March of 1775 collected the gas given off by heating the “red calx of mercury” (mercuric oxide).  He was probably expecting to get what was called “fixed air” (carbon dioxide).  However, when he applied the standard test for “fixed air” by mixing the gas with limewater (Ca(OH)2), he got a negative result (i.e., the limewater did not give off the chalky precipitate [CaCO3] that it normally does in the presence of “fixed air”).  However the experiment was not a failure.  Knowing that the gas was not what he had expected gave rise to further experiments that resulted in his successfully identifying oxygen as the cause of combustion.

In the applied sciences, the situation is not quite the same.  Let us suppose, for example, that the makers of the first atomic bomb had produced only a fizzled explosion.  The primary goal in this case was not to learn something but to produce a desired result. Scientifically they might have successfully learned something; practically, they would have failed, and I think most of the people involved with the operation would have described the trial as a failed experiment.

The NBC officials who moved Leno to his 10 o’clock spot were expecting to draw larger audiences than their competitors at a lesser cost.  Unless they were complete idiots, they had to have known that there was some uncertainty concerning the result of their actions and thus what they did was to some extent an experiment in the second sense of the word (the root meaning of “experiment,” by the way, is “trial”). One of the participants has claimed that the experiment was improperly conducted, because not enough time was allowed for the desirable results to come about.  But, of course, even if they believe this to be true, the directors at NBC cannot afford to continue the experiment, and so however successful it might have been in theory, it was in practical terms a failure.