Particle physicists have a weird relationship to journals. We publish all our results for free on a website called the arXiv, and when we need to read a paper that’s the first place we look. But we still submit our work to journals, because we need some way to vouch that we’re doing good work. Explicit numbers (h-index, impact factor) are falling out of favor, but we still need to demonstrate that we get published in good journals, that we do enough work, and that work has an impact on others. We need it to get jobs, to get grants to fund research at those jobs, and to get future jobs for the students and postdocs we hire with those grants. Our employers need it to justify their own funding, to summarize their progress so governments and administrators can decide who gets what.

This can create weird tensions. When people love a topic, they want to talk about it with each other. They want to say all sorts of things, big and small, to contribute new ideas and correct others and move things forward. But as professional physicists, we also have to publish papers. We can publish some “notes”, little statements on the arXiv that we don’t plan to make into a paper, but we don’t really get “credit” for those. So in practice, we try to force anything we want to say into a paper-sized chunk.

That wouldn’t be a problem if paper-sized chunks were flexible, and you can see when journals historically tried to make them that way. Some journals publish “letters”, short pieces a few pages long, to contrast with their usual papers that can run from twenty to a few hundred pages. These “letters” tend to be viewed as prestigious, though, so they end up being judged on roughly the same standards as the normal papers, if not more strictly.

What standards are those? For each journal, you can find an official list. The Journal of High-Energy Physics, for example, instructs reviewers to look for “high scientific
quality, originality and relevance”. That rules out papers that just reproduce old results, but otherwise is frustratingly vague. What constitutes high scientific quality? Relevant to whom?

In practice, reviewers use a much fuzzier criterion: is this “paper-like”? Does this look like other things that get published, or not?

Each field will assess that differently. It’s a criterion of familiarity, of whether a paper looks like what people in the field generally publish. In my field, one rule of thumb is that a paper must contain a significant calculation.

A “significant calculation” is still quite fuzzy, but the idea is to make sure that a paper requires some amount of actual work. Someone has to do something challenging, and the work shouldn’t be half-done: as much as feasible, they should finish, and calculate something new. Ideally, this should be something that nobody had calculated before, but if the perspective is new enough it can be something old. It should “look hard”, though.

That’s a fine way to judge whether someone is working hard, which is something we sometimes want to judge. But since we’re incentivized to make everything into a paper, this means that every time we want to say something, we want to accompany it with some “significant calculation”, some concrete time-consuming work. This can happen even if we want to say something that’s quite direct and simple, a fact that can be quickly justified but nonetheless has been ignored by the field. If we don’t want it to be “just” an un-credited note, we have to find some way to turn it into a “significant calculation”. We do extra work, sometimes pointless work, in order to make something “paper-sized”.

I like to think about what academia would be like without the need to fill out a career. The model I keep imagining is that of a web forum or a blogging platform. There would be the big projects, the in-depth guides and effortposts. But there would also be shorter contributions, people building off each other, comments on longer pieces and quick alerts pinned to the top of the page. We’d have a shared record of knowledge, where everyone contributes what they want to whatever level of detail they want.

I think math is a bit closer to this ideal. Despite their higher standards for review, checking the logic of every paper to make sure it makes sense to publish, math papers can sometimes be very short, or on apparently trivial things. Physics doesn’t quite work this way, and I suspect part of it is our funding sources. If you’re mostly paid to teach, like many mathematicians, your research is more flexible. If you’re paid to research, like many physicists, then people want to make sure your research is productive, and that tends to cram it into measurable boxes.

In today’s world, I don’t think physics can shift cultures that drastically. Even as we build new structures to rival the journals, the career incentives remain. Physics couldn’t become math unless it shed most of the world’s physicists.

In the long run, though…well, we may one day find ourselves in a world where we don’t have to work all our days to keep each other alive. And if we do, hopefully we’ll change how scientists publish.