Cargo Cult Science

by Richard Feynman Read the Original

Richard Feynman discusses "Cargo Cult Science" - following scientific methods but missing a sense of integrity and honesty - and how it can lead to false conclusions and public deception. He emphasizes the importance of repeating experiments and of reporting all results honestly regardless of outcome. He gives examples of successful experiments, including a rat experiment by Young in 1937 and failed parapsychology experiments, to illustrate the importance of scientific integrity.

Richard Feynman explains how the scientific method was developed to separate fact from fiction. He then shares his investigations into mysticism, hallucinations, Esalen, reflexology, extrasensory perception, and PSI phenomena, showing how people believe in many "wonderful things". He ultimately concludes that it's not a scientific world. “Then a method was discovered for separating the ideas—which was to try one to see if it worked, and if it didn’t work, to eliminate it.” We can separate good and bad science by testing experiments and eliminating those that don't work. This is an important piece of the scientific process and helps to ensure honest results.

We often believe in things that don't work, such as ineffective reading methods, criminal treatments, and psychotherapy. This is "Cargo Cult Science" - following all the forms of scientific investigation but missing something essential, like airplanes not landing. “So I was unable to investigate that phenomenon.” Feynman found that his investigations into paranormal phenomena often yielded no results, emphasizing the need for repeatable experiments and unbiased reporting.

Cargo Cult Science is missing a sense of scientific integrity, a principle of utter honesty and care not to fool oneself. This includes reporting everything that could potentially make an experiment invalid, and acknowledging facts that disagree with one's theory. Without this, planes won't land, no matter the difficulty of the subject. “So I found things that even more people believe, such as that we have some knowledge of how to educate. There are big schools of reading methods and mathematics methods, and so forth, but if you notice, you’ll see the reading scores keep going down—or hardly going up—in spite of the fact that we continually use these same people to improve the methods. There’s a witch doctor remedy that doesn’t work.” Even if something is widely believed to be true, it may not necessarily be effective. When something fails to produce results, it might be time to question it, and search for a more effective solution.

Scientists must have utter integrity to avoid self-deception and should always publish results honestly regardless of outcome. Failing to do so can lead to false conclusions and deceive the public. “It is not something simple like telling them how to improve the shapes of the earphones.” There is a need for deeper understanding and knowledge, often beyond what is obvious, in order to make meaningful change.

Scientists must publish their results to be seen as giving scientific advice. Poor science is often characterized by not repeating experiments. At NAL, the need for new results is prioritized over scientific integrity. An example of a successful psychological experiment was done by Young in 1937, which showed rats learn patterns. “You must do the best you can—if you know anything at all wrong, or possibly wrong—to explain it.” It is important to be honest about any weaknesses or uncertainties in one’s research in order to ensure scientific integrity.

A rat experiment illustrated the importance of using scientific integrity when conducting research, as the rats were able to distinguish between doors despite efforts to fool them. Parapsychology experiments also illustrate the importance of integrity, as they fail to produce repeatable results. The speaker encourages the audience to maintain their integrity and to not succumb to pressure. “If we only publish results of a certain kind, we can make the argument look good. We must publish both kinds of result.” Honesty in research is essential. We must not be selective in what results we show, or else our argument becomes biased and unreliable.