We are publishing many fine papers across the whole of the physical sciences, but here I want to focus on those we do not publish. Our current rejection rate is approaching 80 per cent. This is partly a result of the journal's popularity: we receive many more submissions than our page budget allows, so there are some good papers we simply do not have space for. But our main reasons for rejection are different: papers are too specialized for our readers, or report advances that are incremental rather than fundamental.
For a paper to be publishable it must report work that is correct and new. But these obvious conditions are not sufficient, as this absurd example demonstrates: 3921×4723=18 518 883. This result is surely correct, and has almost certainly never been published: it is ‘new’. To be publishable in Proc. R. Soc. A, much more is required. In addition to the criteria listed on the journal's home page (‘high-quality, original, fundamental articles of interest to a wide range of scientists’), I like to see an element of surprise. And in connection with ‘new’, I ask authors to delete the strident phrase ‘for the first time’: would they be submitting a paper if it were ‘for the second time’?
Perhaps unsurprisingly for such a prestigious journal, we receive a steady stream of papers proposing a new theory of the universe, or a replacement for Einstein's relativity, or an ‘obvious’ proof of the Riemann hypothesis. Such papers usually fail several tests that any new idea must satisfy: demonstrable consistency with everything existing theory can explain, quantitative explanations of what existing theory fails to explain and/or unambiguous prediction of something not yet observed. Rejecting them takes much of our Editorial Board members' time, because, although we are under no obligation to publish any article that we receive, we recognize the seriousness of all authors, including those who are mistaken. Authors of rejected papers sometimes complain that their discoveries are being suppressed, but they are now free to publish on the Internet (if they do, it could constitute prior publication, disqualifying their papers from Proc. R. Soc. A).
To avoid misunderstanding, I declare that we welcome speculative papers that meet our quality standards. My own view is that all our current scientific concepts are probably wrong, but in a very specific sense: they will eventually be discovered to be special cases of much more general theoretical schemes (just as quantum mechanics superseded classical, by revealing it as a limiting case). I would be delighted if the paper that sparks the next scientific revolution were published in Proc. R. Soc. A, but such an advance will have to surmount a very high barrier indeed.
- This journal is © 2011 The Royal Society