Science

Science Is the Study of Predictability

A four-word definition of science that none of the textbook definitions quite manage.

By Bernard Weiss 1 min read

I asked an AI platform to cite various definitions of science. It listed six, varying in length from 11 to 24 words. None of them use the word “predictability” or any of its derivatives.

Everybody has ideas about the universe and how it came about. We live in a democratic society where everybody is supposed to be equal. So we think that all ideas are equal. But they are not. Some ideas are good and some ideas are bad. Good ideas are better than bad ones.

What do we mean by good ideas? Most of us define good ideas as our own ideas; bad ideas are everyone else’s. But there is a more objective definition: good ideas are ideas that work; bad ideas are ideas that don’t work. The physicist Richard Feynman described science as the study of ideas that work. I should like to extend that explanation by adding that an idea the works is one that has demonstrably predictable consequences. Science is the study of predictability.

To see if an idea works, we have to test it through experimentation. An hypothesis should have predictable consequences, and these can be tested. If a prediction is not confirmed, then the hypothesis is modified, and the new hypothesis is tested. That is how science progresses. Theories are not laws written in stone. They are not like religious dogma. If they don’t pass a reality check, they are replaced by ones that do.


This is an excerpt. The full essay appears in One-Sided Conversations: Musings of a Scientist.

Join the conversation

Your email address will not be published. Comments appear after approval.

An occasional letter

Hear about new essays and the book.

When there is a new piece on the site, or news about the book, you will get a short note. Not often, and never anything else.

Occasional news and new essays only. No spam; unsubscribe any time. Privacy policy