I was recently listening to a podcast with Sam Harris in which Nicholas Christakis used heliocentrism as an example of scientific truth. It’s a standard cliché in science communication. Copernicus and Galileo developed better instruments, collected better data, and followed the evidence impartially, concerned only with finding truth. Through their tireless empiricism and intellectual rigor, truth trimphed over religion, tradition, and supersition. This, we are supposed to believe, is how science works. However, while heliocentrism is a better model for understanding the orbital dynamics of the solar system, the sun is not the fixed center of the universe. Heliocentrism is also a really awful model for understanding the galaxy. It’s neither true nor the best model in any general sense, only the best model for understanding some processes within a particular context.

People who use this cliché would likely object that I’m missing the point. The point is that empirical data and intellectual rigor won the day. Science is an ongoing process of discovering the truth. That we have moved on from heliocentrism is a further illustration of the point, not a counterargument. I think this is fine so far as it goes, but too simplistic. Heliocentrism is given as an example of a straightforward and incontrovertible scientific truth, and it is not that. We shouldn’t flatten intellectual progress into a succession of contests between the true and the false. Neither geocentrism nor heliocentrism is true, and neither is false. They are different frames of reference. They both are, or at least can be, models that are equally compatible with the evidence but useful in different contexts. When navigating to the supermarket, it’s perfectly reasonable to treat your house as a fixed starting and ending point. Conceptualizing the earth’s rotation and its orbit around the sun is profoundly unhelpful if you’re trying to map the vegetation of a mountain range. Scientific progress is about figuring out which conceptual tools and mental models are best suited to helping us understand different phenomena. We should be more interested in holding multiple alternative models in mind, mapping them onto each other and navigating between them, than in pronouncing one of them Truth.