Epistemic Status: Not exhaustively researched, dates and particular events are selected based on my incomplete and biased memory. My point in this post also rest on me not having seen particular versions of the story of science in the wild as well as repeated encounters with people seeming uncomfortable with the more abstract components of science. YMMV.

When the story of science is relayed it is often in terms of the development of that body of knowledge. We celebrate milestones of understanding like Galileo and Heliocentrism, Newton and Universal Gravitation, and Darwin and Evolution by Natural Selection. 

An alternative story that is sometimes told is that of science as social activity and the role of power: The influence of funders politics and societal values on the direction of research and technology. The idea of progress vs stagnation. How the actions of individual scientists ripples outwards affecting generations. Science as part of the wider cultural fabric.

But I think it is worth reflecting on how science as a method has also changed over time. This development is fairly well documented but it is not presented as a narrative nearly as often in my experience. My guess is that the triumphs of such a story are too abstract to be as immediately compelling and were often unrecognized at the time as separate accomplishments from their associated advancements in the scientific body of knowledge. 

Presented here is a chronological list of the major conceptual milestones in the development of the scientific method. This is not a list of scientific discoveries or theories, this is a list of the conceptual tools that enabled such discoveries. Not all of these were recognized for what they were at the time, but have since become integral to the method of scientific investigation and inquiry. 

This list forms a sort of narrative. And like all stories much is left out. The hundred schools era in ancient China and the Islamic Golden age are fascinating periods of intellectual development that are valuable to consider. 

One of the insights that this list might provide, is that what counts as the scientific method has changed over time. This is suggestive that perhaps science has not reached its final form. Indeed, some more recent developments might be candidates for becoming part of ‘just how science is done.’ For example, interdisciplinary collaboration, open source knowledge sharing, and big data projects can rely on a foundation of consilience. 

Connor Leahy has made the suggestion that developing a robust theory of intelligence and AI safety might require what he calls science-2. Based off of the list above I think that might be more like science-20. Specifically Leahy is calling for a new form of predictive science to simulate how advanced AI systems will behave before they are invented. In light of all that has come before, I do not think this is very unreasonable.