Scott describes mētis and science as two distinct but complementary forms of knowledge with different methods and goals. If he were more familiar with practicing scientists, he would have thought that they are less distinct.

---

Prerequisites: I reference Seeing Like A State, especially the discussion of mētis. This post should be mostly self-contained, but if you want more context, see the section on mētis in my book review.

Originally Written: August 2021.

Confidence Level: This describes my involvement with science, but I can’t speak for science as whole.

---

---

‘Scientific’ Forestry

Scott opens Seeing Like A State with the story of how scientific foresters promoted uniform forests with trees in straight rows, even though this was bad for the biodiversity of the forest and the sustainability of lumber yields.

Something about this story struck me as odd. The scientists I know who spend the most time in forests are more likely to be digging through rotten logs looking for millipedes than wanting to rearrange the trees.

Scott mistakes science for scientism: people who claim and promote the legitimacy of science, although they are not scientists themselves. Although scientism usually tries to get the facts right, it often misses the general attitude and underlying philosophy of the field. This is a very easy mistake to make and is probably the default unless you regularly interact with practicing scientists.

Mētis vs Science

Scott does do more than criticize scientism. He also contrasts how science and mētis are supposed to work.

Mētis is the knowledge built through experience. It works best in situations which continually have small changes. If mētis is summarized, it is done as rules of thumb, which do not try to be consistent, and are usually embedded in a framework which looks thoroughly irrational.

Science, in contrast, tries to be repeatable, universal, and systematic. In order for a scientific result to be believable, it should be replicated, at least statistically. Scientists then try to summarize their results in the fewest, farthest reaching laws as possible. These laws are continually checked for consistency, and if they aren’t consistent, huge amounts of theoretical effort will try to reconcile them.

These ideals of science cause it to focus on certain types of problems and favor certain kinds of solutions. Situations with regular motion or where only one variable is important are often preferred over situations with random or chaotic motion or many variables. This tendency has become weaker as analytical tools has become more sophisticated.^[1]i.e. Scientists use fancier math to look at more complicated situations.

Scott sees science and mētis as distinct and complementary forms of knowledge. Ideally, problems should be approached both by a community and a scientific community, with significant but not complete overlap between the two. There should be continual collaboration to ensure that the strengths of both science and mētis are used to solve problems. Using mētis alone leads to irrational superstitions which work even when they’re wrong,^[2]Here is an interesting example. while using science alone leads to rational beliefs that don’t solve actual problems. When science alone is imposed on people by large institutions, it becomes imperial knowledge and leads to the failures of Legibilism described in Seeing Like A State.

Mētis as Science

Scott does correctly state some of the ideals of science, but he does not have the perspective of scientists ourselves.

Scientific research continually involves attempts things that no one has done before.^[3]Replicating others’ novel results is also important. Experimental equipment, much of which is custom-built or heavily modified, is pushed to its limits to investigate things no one has been able to do before. The results cannot be fully planned for because they are not known beforehand.

In these circumstances, rigid rules and formal methods are often not helpful. If we do not know what we are going to find, how can we preemptively decide the best way to find it? Instead, we mostly try things that have worked before elsewhere.

Scientists are trained through experience. Often after being in junior positions in multiple labs^[4]Working in multiple labs is important to avoid ‘academic incest’, or only being familiar with the traditions of a single group. does someone become the Principle Investigator of a major research project. Some philosophers of science do try to describe or proscribe a rational method for science. But philosophy of science has not been as successful of an endeavor as science^[5]At least for the hard sciences. ‘Hard’ meaning those not caught up in the Replication Crisis – which is the ones you’d expect. and many scientists are largely ignorant of it.^[6]I personally think that scientists should be more familiar with the philosophy and history of science than they are.

Science itself is a mētis-dominated field. There is lots of extremely local knowledge which is necessary to manage continually changing situations.

The conflict between science and mētis occurs when science is filtered through the ‘experts’ of large institutions. ‘Imperial’ knowledge is not only imperial from the perspective of the local farmer it is being forced upon. It is also a reduction of the original community which produced that knowledge. If you put actual practicing scientists in contact with local farmers, the results will be much better than what Scott described in Seeing Like A State.