Bad incentives in science apply to your “solutions” too

This is really just a hasty screed copied over from facebook, which I thought I might keep in a slightly more permanent place.

It’s based on this opinion piece in Nature: Give researchers a lifetime word limit.

It starts with a good observation (though at this point it’s a fairly widely held belief):

The purpose of authorship has shifted. Once, its primary role was to share knowledge. Now it is to get a publication — ‘pubcoin’ if you will. Authorship has become a valuable commodity. And as with all valuable commodities, it is bought, sold, traded and stolen.

The proposed “solution” — a lifetime word limit per author, would, according to the author, have the following miraculous effects:

“Lifetime limits would create a natural incentive to do research that matters.”
“It would also lead to increased value being placed on concision and clarity, improving readability and efficiency.”
“Predatory publishers would vanish.”
“The task of evaluating candidates for jobs, advancement and prizes would become less scattershot.”

No. Have we learned nothing from communism? If you think that it’s a problem that something has been commodified, putting a hard cap on it is *not* a solution. I’ll get to the tremendous implementation difficulties later. For now, let’s see some ways your great idea would backfire due to the terrible incentives it creates.

Papers would become shorter and more obtuse, with more of the content shoved into appendices, “pre-prints”, personal websites, and anywhere else that doesn’t count.
Honourary authorship would become less attractive — but a shadow “authorship” would emerge, of (e.g.) grad students leaving off their supervisors, but still making it known as “their” work.
Many of these grad students, who aren’t going to get academic jobs anyway, would just be disposable husks used up for their word count.
Researchers would face new anxiety about whether their current topic was “important enough”.

I think the deeply unethical disposable-grad-student-husks should alone be enough to stop it. But perhaps that’s naive; it certainly doesn’t seem to have stopped our current system. The big point is this would not help science.

The implementation side is also, shall we say, nontrivial.

It’s technically easy (ish) to require all authors to have ORCIDs, and tally everyone’s word count. But that’s the only easy part.

You’d have to get all journals to agree to this system. Getting all journals to agree on anything is more or less impossible. Even if you did in your fantasy world, you’d have a new type of predatory publisher: those that faked compliance with this system.

And what happens with a submission that’s over someone’s lifetime word count? Is it auto-rejected? Or does that author just get dropped?

What happens when a faculty candidate is near their lifetime word limit? Are they automatically rejected?

What happens when a recognized authority is no longer allowed to publish — both for their own career, and for the field?

How is it phased in? Does it only apply to publications from this day forward, and hence disproportionately hit younger researchers? Or does it force a large cohort of senior academics into immediate retirement?

Who sets these word limits anyway? The author acknowledges “Some subjects and pursuits might be inherently wordier than others.”

But then he undermines that with “Exceptions might have to be made for experts such as statisticians and bioinformaticians whose skills are required on many papers — but perhaps this would boost the quality of collaborations. Perhaps researchers could apply for word bonuses for careful reproductions, cautious interpretations and meticulously described methods.” Noooooo! This just commodifies the bureaucratic filing of appeals.

Lifetime word limits do not help science, and it’s bad enough that I can’t see a way to make it work. It’s lucky that this can’t get off the ground. But for next time: please, please think about perverse incentives before proposing brilliant new solutions.