Thread by @psforscher

For the past week or two, I’ve been mulling over this fascinating paper on the historical origins of open science. papers.ssrn.com/sol3/papers.cf…

I think it has a lot to teach meta-scientists. However, it’s long and abstruse. This is my attemp to make the content more accessible 🧵

David starts with the following question: why should scientists make their knowledge public?

Although public knowledge might accumulate more rapidly than private knowledge, the *institution* of public knowledge is pretty weird and rare, historically

In the medieval era, for example, the learned sought to keep their knowledge secret from “the unworthy”

People in this period had moral and economic reasons to maintain secrets.

Morally, there was worry about the negative effects of disclosing Nature’s Secrets to those who couldn’t handle them.

Economically, institutions like guilds could exploit secrets for private gain

Historically, it’s also unclear how open scientists would support themselves. Public disclosure of secrets prevents scientist from exploiting these secrets for money — so the money has to come from elsewhere

The money problem is particularly tricky because, if an open scientist is routinely and fully disclosing their secrets, people can “free ride” on the scientist’s labor — they can exploit the secret knowledge without giving anything in return

David argues that the answers to these problems lies, at least in part, in the motivations of European ruler-patrons.

These patrons were motivated to hire scientists for pragmatic reasons — scientists could help make better maps, build better weapons to conquer stuff, and so on

However, this pragmatic motivation doesn’t explain the rise of open science. Once a client-scientist has told everyone how to build better wepons, the scientist has thrown away their patron’s military advantage

Ruler-patrons has another reason to hire scientists, though “ornamental” reasons. Basically, having a great scientist in their employ made themselves look like magnificent badasses

Science produced to “ornament” a ruler is different from science produced to solve a practical problem in that it can only make the ruler look like a badass if everyone can see it

Ok, we’ve found a potential source of money for open scientists: wealthy ruler-patrons involved in pissing contests for prestige.

However, there are other ways to “ornament” rulers besides public disclosures of knowledge, like building applications that rely on the knowledge

Part of the answer comes from a problem faced by rulers: knowledge had become so technical that rulers had a hard time evaluating the expertise of their prospective clients.

Truly learned scientists therefore had an incentive to separate themselves from the pack

This led to a tradition of public boasts and challenges, all of which were aimed at gaining scientific credibility and prestige

This also led to an emphasis on “priority” of ideas, honor, and intellectual turf

Another way scientists gained credibility for themselves is through joining networks of correspondence among scientists. These networks eventually grew into scientific societies

These societies had strong institutional incentives to maintain their credibility. They therefore became somewhat conservative bodies bent on giving their members the imprimatur of excellence. They did so by bestowing the right of priority on specific scientists

David notes that even early on, many issues threatened to disrupt openness:

- Competition for prestige between scientists often led to incomplete disclosure
- Societies could be coopted by centralized political organs
- Priority concerns led to secret-keeping within societies

Why is all this interesting?

One reason — Even though this whole system of prestige and priority developed out of medieval institutions, we still see its imprints in modern science

I find David’s second reason especially compelling.

The history of open science reveals that the institutions that support it are fundamentally fragile. If we want to sustain institutions that promote good science, we need to continually re-create them through our own effort

I hope this thread piqued your interest. It’s a tough paper to slog through, but I think it’s worth the effort