A new code is no weapon against excessive self-citation

When does self-citation tip from being a useful way to avoid repetition into a means to artificially inflate your academic status?

That is an issue which has exercised research ethicists such as myself for many years. And having spent the past three decades devising, promoting and advising on (at national and international levels) standards that hold as globally as possible, I pay close attention to any initiative that might see scientific misconduct identified and roundly punished.

So the prospect that Switzerland’s new national code of conduct will officially sanction scholars for “unjustified” self-citation – or claiming authorship despite contributing little to a project – raised my interest, as did suggestions that it could lead to similar initiatives elsewhere.

It is not a bad code. Indeed, it is based on the All European Academies (Allea) principles of reliability, honesty, respect and accountability, so it would be hard to go wrong. Sadly, however, the policy’s small print suggests that relatively few scholars will be dragged across the coals for citation manipulation or gift authorship infraction.

That is because, like the UK’s concordat for research integrity, the Swiss code leaves the responsibility for enforcement to “the competent bodies of the institutions and funding organisations concerned”. But universities and funders rarely have a strong incentive to call out misconduct by their own researchers. Doing so risks creating administration burdens, incurring legal costs and endangering corporate reputation. If research misconduct claims were handed to a Dutch-style independent regulator given hefty powers to name and shame, I’d be more optimistic.

Yet there is a further problem. Because if there are no easy measures for misbehaviours, how can they be reprimanded? And in the case of self-citation, the Swiss code offers little guidance on when it becomes “unjustified”.

Nor are academics’ instincts likely to aid enforcement. Judging by the chatter on scientific blogs, many regard self-citation as OK when it avoids having to plagiarise yourself and helps establish your present work’s continuity with previous studies. One contributor to a recent discourse mentioned having cited themselves to illustrate a point, for instance.

In other words, the consensus is that there is nothing inherently wrong with self-citation as long as you don’t overdo it and the references are generally relevant. But what is overdoing it? What percentage of citations being self-citations would be unjustified?

I doubt everyone could agree that a breach has been committed in all but the most egregious of cases – such as that of Sundarapandian Vaidyanathan, a computer scientist at India’s Vel Tech university who received 94 per cent of his citations from himself or his co-authors up to 2017. And that, of course, only further disinclines institutions to take action except when they are extreme and more publicly visible.

At least publishers and journal editors are in a position to make a judgement concerning what might be acceptable in their own publications. And the Committee on Publication Ethics (Cope) has produced some fairly definitive statements over the years about how journals should approach the issue. It recommends, for instance, that journals “consider policies about appropriate levels of self-citation for authors and reviewers” and “provide education for editors about appropriate times and ways to request citations to the editors’ or journal’s publications”.

What interests me is how questions of research ethics keep going around and around – and how we keep reinventing the wheel trying to address them. The answer is that particular definitions fail to make any impact with researchers because the threat of enforcement just isn’t there. All the work that has been done on research integrity, much of it funded by the European Commission via its Responsible Research and Innovation projects, requires regulatory action at the national level, based on global standard setting. But there is little to show for all this work in either legislation or effective sanctions.

Regarding self-citation, the most effective solution would be to stop assessing academics using such an unreliable measure as citation volume. Even those who enjoy high h-indexes can do so for the strangest of reasons; for instance, I have read others citing me for things that I either did not write or, in my view, were not relevant to their argument. But there is little I can do about such aberrations unless I write to the journal about a matter that might seem quite trivial in the great order of things. And when I was in a full-time post, even an ethics expert like me would have been unlikely to kick up a fuss about it with my superiors.

Personally, I try very hard only to cite myself to avoid repeating something dealt with extensively elsewhere – and, specifically, by me. As with all matters of research integrity, self-citation comes down to personal responsibility and authors’ sense of honour. But in a high-pressure environment full of shades of grey, that is often not enough.

Ron Iphofen is an independent research consultant and leads the UK Academy of Social Sciences’ team in the European Commission’s PRO-RES project.