One More Epicycle

So on Friday we had Leslie Sage visit, who works for Nature (the massive science publishing giant) and is the point man for astronomy. I learned a fair bit about the way their process works, much of which surprised me, and left me thinking less of the journal than I’d thought beforehand, which wasn’t much. This is a bit of a problem, as Nature gets a fair amount of the most interesting science, and the boss and I were toying with aiming one of our ongoing projects in its direction.

Basically:

(1) Sage explained that Nature only publishes 7% of submitted papers, and roughly 12% of the astronomy-related submitted papers, because they’re looking for only the best papers. This isn’t true, though: they don’t publish the best papers. Not even close. The articles in astronomy in Nature are usually pretty terrible as papers. For example, in theory, it’s almost impossible to figure out what the authors have actually done from their description in Nature. The articles are so compressed that there’s little room for the details which are necessary to reproduce a simulation. It’s not the authors’ fault at all, or at least not most of the time; page budgets are very tightly controlled, often to the point of absurdity.

There are papers, and there are “Nature papers”. It’s not uncommon to be searching the literature, find out that a result you need to understand was published in Nature, and then find yourself desperately hoping they published something real on it somewhere useful.

Sage argued in Nature’s defence that they allow online “supplemental information”, but that’s more suited for data tables. To suggest putting necessary information in a supplement is merely another way of admitting that they don’t publish science papers but press releases/extended abstracts with graphs.

Nature isn’t trying to get the best papers, but the sexiest results. And this they often succeed at, but I don’t see why it’s in the community’s interest to have our most important results written up in a deliberately inferior format — “defective by design”, as they say.

(2) I’d always assumed that there was some team of notable astrophysicists who reviewed the submissions and decided if they were Nature-worthy results, independent of the refereeing process. It turns out that said team consists of Leslie Sage. (Plus cameo appearances by people he sometimes consults.) I’m sure he’s a decent enough guy, but I don’t like single-point-of-failure arrangements, and with all due respect to Dr Sage, “capturing Leslie Sage’s interest” doesn’t strike me as all that prestigious. Moreover, what if you manage to incur Dr Sage’s wrath? Then what?

In the end I suspect that we would all be better served simply by not sending our results to Nature, breaking the cycle of dependency. For various reasons people are taking a long, hard look at the way we do science publication at the moment, and it’s not at all clear how the landscape is going to look in ten years, but we can start by imagining one in which Nature doesn’t play a role.