An obsession with publishing in three high-prestige journals is ruining careers and undermining good science.

Cell. Nature. Science. Three short words – but like another worshipped trio, their influence belies the simplicity of their names. These three journals exert a pathological and mesmeric hold on the entire biomedical research body, and it is high time that that spell was broken.

First off, and let’s get this in the clear so that there are no misunderstandings, it should be stressed that Nature and Science are terrible, terrible organs for the publication of scientific research. Egregiously so. They have clung to an outdated model of publishing where 1-4 figures was considered sufficient for a research paper and have smugly and complacently stuck to that model in contempt of all modern developments.

Their defenders will point to the apparently wonderful concession of allowing extended supplemental data, but this does not change – and not by one iota – the fundamental flaw with their entire format when it comes to presenting scientific data. The times have changed. Most research papers require far more than four figures to satisfy readers, and the result has been the frankly farcical inflation of supplementary figures appended to a paper. In fact, the amount of supplemental data now exceeds the amount of “actual” data accompanying the text of the article. As a means of conveying scientific work, it is poor. Embarrassingly so.

Cell, at least, is a decent vehicle for communicating work. The articles are a decent length, figures are of dimensions that do not require 400x magnification in order to discern details, and the suffocating sense of compression is absent. It should be shunned nonetheless. It is the blameless bad guy. Nature and Science too are not without some redeeming merit – both feature outstanding magazine sections that are probably the best source of science news, and worth reading on a weekly basis.

So why the anger? Why the obsession? Why should two bad research journals and one good one produce such a strength of feeling?

Simple. Because if you’re published in any one of those three, you’re more likely to get a job interview. In fact, it has become a pernicious and self-fulfilling truth that a first-author paper in one of those journals is a ticket to an independent position for early-career scientists. (The reason? A lot of open-call positions are swamped by 200+ applications and the way they’re whittled down to a shortlist of interview candidates is by discarding all the ones that don’t have a first-author Cell/Nature/Science paper. It’s never official policy of course.)

Therefore, the act of publishing in these three journals brings career benefits that are not accrued by publishing the same story elsewhere. Again, let’s reiterate that – it is actually better to publish your work in a form that is virtually unintelligible even to specialists in your own field, simply because of the cachet that is attached to these journals’ names. It is better to deliberately mutilate your own data, squashing it and dismembering it, and throwing the pieces onto a huge pile marked “supplemental”, than it is to present it to best effect.

More worryingly, the disproportionate impact that comes from publishing in Cell/Nature/Science has predictably been an incentive of the worst possible kind when it comes to unscientific behaviour, and not surprisingly the three articles have higher levels of retraction than the average. You might argue that the scientific record is self-correcting and that truth will win out, but that’s scant consolation to all the young scientists who have left academia because they couldn’t even get a job interview and a chance to prove themselves.

How have we ended up here? Why do we collude in such an awful system by discarding job applications from people that don’t have a Cell/Nature/Science paper? Because we currently don’t have an effective way of assessing the quality of work done outside our own narrow scientific niche. The recent proliferation of metrics is an attempt to do this, but seems a misguided one – trying to condense a person and all that they might offer into one or several numbers is no improvement on the current slur of glancing at their publication list and reading only the journal names. And it is a monstrous, monstrous insult to any scientist to suggest that the quality of their work can be determined by their choice of journal alone.

Naturally, it’s of obvious value to Cell/Nature/Science to maintain the hegemony that they currently wield over the biomedical publishing landscape, so the chance of change coming from that quarter is minimal. The arrival of the journal eLife, with support from on high courtesy of the Max Planck Society, the Wellcome Trust, and the Howard Hughes Medical Institute, is to be welcomed but is not a long-term solution. It does at least provide a viable riposte to the “sour grapes” argument – that those who complain about Cell/Nature/Science are the people who couldn’t get published there even if they tried – but in the long run it may simply end up joining the club. Real change must come from elsewhere.

The first and most important step is that those who are in a position to hire young scientists must reject the Cell/Nature/Science criterion for shortlisting candidates (how many tenured professors would have passed that same criterion at the equivalent point in their own careers?). There does need to be a way of assessing and shortlisting candidates – some kind of triage – but the “journal choice” method is not it. Change too must come from the young scientists themselves. By continuing to collude in a system that so arbitrarily judges their work and their capabilities, they are collaborating in their own belittlement. The Salem witch trials, another instance of a community gripped by a pathological mania, came to an end only when more and more of the accused refused to cooperate. Young scientists must be brave and do likewise.

Reader, be my witness. I will never submit a research paper to Cell/Nature/Science.