A brief case for wild research

Call a discipline cumulative if researchers in that discipline depend on the prior work of other researchers in two senses:

1. They need it for their own work to proceed.

2. They cannot easily check the work they rely on.

On this basis we might rank disciplines by cumulativity. Here’s a rough ranking:

Biology

Chemistry

Physics

Psychology

Cultural anthropology

History

Mathematics (these days at least- checking proofs is so hard now)

Economics

Sociology

Literary studies

Philosophy

In the non-cumulative disciplines, the negative reason wild research is to be encouraged is that there’s vastly less harm in being wrong. You’re unlikely to lead anyone astray, and if you do lead people astray, it’s probably not because you didn’t cross your t’s and dot your i’s in the setup of your research- anyone can check your arguments. If your stuff becomes influential, it will be gone over with a fine tooth comb.

The positive reason wild research is to be encouraged is that, in my experience, only a small handful of research in these disciplines non-cumulative disciplines has an ongoing effect decades after it’s published, and that work is often wild- brave and speculative. Small bits of research don’t become building blocks in the growth of knowledge, they’re just forgotten. The stuff that lasts really does something. Thus wild research is the best way to do anything that, in the long run, will matter.

In philosophy, there’s a common view that we’re a pathological discipline because we don’t do our homework before jumping into fields and problems. Certainly philosophers could always do a bit more reading, but I’d rather ten articles with howlers in them get published than one really good idea miss out.

.