I quite like this definition of science from the UK’s science council. We all kind of know what science is, but in postmodern times, this is not enough and battles are fought over the meanings of words under the assumption that this is how to win political arguments.
Science can refer to a process — the scientific method — and the knowledge obtained through that process. It consists of building models of the world. Note that they are only models and rarely attempt to fully and accurately describe the world. At some point, there is usually some hand waving and an appeal to another realm of science or perhaps one yet to be discovered.
The models have to be testable to be useful. They need to make predictions about experiments or observations that have not yet been conducted. We then conduct those experiments or observations and if the model’s predictions are correct, it survives. This survival is only ever provisional until a better or more complete model comes along.
Science is great. I am one of its biggest fans. But it is not without problems. In psychology, we suffer from a replication crisis where junk models persist due to dodgy methods. There are arguments about tests of statistical significance and the meaning of effect sizes. There is the file-drawer problem, where studies showing no effect are less likely to be published because they are seen as less interesting. But this means that models do not get properly tested.
Keep reading with a 7-day free trial
Subscribe to Filling The Pail to keep reading this post and get 7 days of free access to the full post archives.