You Need to Read Thomas Kuhn.
I’m trying to learn more about the history of science. My hope is to find patterns that lead scientists to great work. I also want reassurance that doing science outside of modern institutions can be done.
I’ve started reading Thomas Kuhn’s ‘The Structure of Scientific Revolutions’. Thomas Kuhn was a physicist, and then worked on the history of science. The book is his crowning achievement, and has helped me build my own classification for the different kinds of science I do.
Kuhn says that scientific fields start off as pre-paradigmatic. This is when there are lots of scattered ideas about the object of study, with no big agreement. If you read about the history of electricity before the 1800s you realise that no-one agreed on anything!
Then, a dominant set of ideas called a paradigm emerges. For example, the theory of electrons.
After that, we have two ‘phases’ that alternate. The first phase is ‘normal science’, where scientists seek to articulate, find, and close gaps in this current paradigm. This is what the majority of scientists work on, constantly testing and tweaking the paradigm to make sure it fits experiment. It involves things like puzzle solving and engineering instrumentation. Our current paradigm for how electrons behave is quantum mechanics and quantum field theory, where scientists test agreement with theory in everything from materials to light to chemistry.
The second phase is ‘revolutionary science’ in which a new idea challenges the current paradigm, and soon becomes the new paradigm itself. A paradigm shift in quantum physics would look something like the rejection of Schrodinger’s equation in favour of something radically different.
To be honest, I don’t like these terms because it implies that ‘revolutionary science’ is cooler, but in my eyes both are crucial!
A lot of the stuff I do in chemistry feels like what Kuhn would describe as ‘normal science’. Computational chemistry seems to have a clear paradigm, which is to find approximations to Schrodinger’s equations that allow a computer to solve it quickly. There are different flavours of this (Born-Oppenheimer, excitons, DFT, SCF, …), but this it what it all shakes out to.
On the other hand, biosecurity as a field feels pre-paradigmatic. We still have huge ongoing debate about how humanity can best defend against pandemics without significant economic damage. We still don’t have very strong frameworks on quality standards for air in buildings, like we do for drugs. It’s all very messy.
I generally think doing different things is a great way to hedge your bets about what works. Some work should be messy, other work should be more like puzzles with a solid end goal. I like using this classification to help me spread out what I’m working on!