Today, I recieved an email from a journal, asking me if I would review a paper. The paper in question is by, amoung others, Iddo Friedberg, and can be read on arXiv (n.d.a) I’ve known Iddo Friedberg for a while; he was an earlier user of my semantic similarity work (n.d.b) for protein function prediction (n.d.c) and was also the editor for our paper on realism in ontology development (n.d.d) I would have liked to review this paper, and I feel a little bad because I know these things are important for the careers of the scientists.

So, why did I decline? Well, nice and simple; the page charges are just too high. There is no real justification for this as it can be done much cheaper (n.d.e/) or so seems reasonable; more over, I think it is bad for science because it is one of the factors that cause authors to think very carefully, and often save up work for “a bigger publication”. This can delay publication for years after the work has happened. Scientists have to think carefully about their research, and their work; thinking about whether to publish now or later is one piece of baggage that we could do without (n.d.f)

The real irony of the situation though is that the peer-review for this paper is and has already happened. The paper concerns bias in Gene Ontology annotations of protein functions. Iddo posted his work to the various Gene Ontology mailing lists; unsurprisingly, the GO annotation team saw the paper, and Rachael Huntley responded. The academic debate has started, and is in full swing. Others may see and contribute. And, frankly, the quality of the discussion going on there, and the depth of the analysis is higher than I would have given. No journal has been involved; it happened because there is a mailing list which the scientists in question used.

The current peer-review system does not add value; my peers and the scientific debate that does this. And this can, and will happen, regardless of the journals; indeed, in this case, why don’t the journal editors just read the mailing list?

So why do scientists, including myself, continue to publish in this way? It can often be difficult particularly where there are no open access options available (n.d.g) We have to; it’s part, indeed, the main part of our assessment (n.d.h/?p=5388) As I have said before, this is now the only reason I publish in this way (n.d.i)

Having said this, I do have my doubts. I feel somewhat guilty toward Iddo Friedberg, for instance. There is also a degree of hypocrisy in this — I will still submit to journals (for my own sake, of course, but also for my PhD students); will people, perhaps, wish to not review my articles? What would happen if everybody thought like this (here, I can use the Yossarian defence: then I’d be a damn fool to think any different). If I set the bar at £200, then who will I review for? Well, I do review for conferences and workshops where I can. Still, I feel that this is not enough; people review my work, I should review theirs. So, I state here, that subject to some time constraints, I will happily review work that is posted either to the web in this form, or to sites such as arXiv. Reviews will be posted here, on this blog.

I have my doubts; but open access is not enough. Publication must get lighter, faster and much, much cheaper. I would welcome alternative courses of action.

Many thanks to the Simon Cockell and James Malone who peer-reviewed this post, and provided helpful comments. I am also grateful to Iddo Friedberg who gave me permission to use the story about his paper in this way. The opinions expressed here are, however, my own.


In response to feedback from Mike Taylor, it is worth pointing out that I do not review for paywall journals, and have not for quite a while.