Moving Beyond the Limitations of Science by Richard Seltzer

(excerpt from Lenses, a book-length collection of short essays, in search of a publisher)

Science progresses by testing educated guesses − hypotheses. But hypotheses depend on expectations based on previous knowledge and cultural bias.

We face the same limitation in everyday life. We filter what we see based on what we expect to see. We ignore anything seriously out of the range of our expectations. If we don't ask the right questions, we don't get the right answers. And as human beings, we have a limited range of hypotheses we are likely to consider plausible. Intuition and thinking-outside-the-box can expand that range, but not by much.

Today, computer simulation is used widely in conjunction with physical experiments to generate hypotheses and then test them. But such simulation typically stays within the range of human expectations.

To overcome that limitation, we need programs which generate hypotheses that are implausible and would not otherwise be considered; programs that come up with complex and improbable ideas and ways to test them. Such hypotheses could lead to experiments that record and help interpret data that would otherwise be ignored.

In the Middle Ages, the rule of thumb known as Occam's Razor was important in setting the stage for scientific advancement. "One should not increase, beyond what is necessary, the number of entities required to explain anything." That rule made practical sense because humans have limited time and limited brain power − focus your research on the most likely explanations. In today's vernacular KISS − "keep it simple stupid."

Now computers can deal with far more variables than humans can; and can calculate trees of causation far further; and hence can identify multiple explanations for the same event, all valid from different perspectives, and perhaps eacy leading to different long-term consequences. It is time to move beyond Occam's Razor, to expand the range of our research to deal with the complex, the unlikely, the redundant, and even the totally outlandish, admitting the possibility that truth might be messy rather than systematic and beautiful.