Ben Allanach, guest blogger, is a Professor of Theoretical Physics at the University of Cambridge. He is grumpy about the way that public funds are being unnecessarily directed to scientific publishing houses. So I am offering this space to him to hear what he has to say about that...

Recently, I've had to arrange payment of thousands of pounds to research journals to ensure that anyone around the world can read my research paper, despite the facts that it is freely available online, that it is my own work, that other academics have reviewed (judged) my work for free, and that my computer and I have made the graphics and done the type-setting.

You may have heard of the recent drive to make researchers subscribe to the Open Access publishing model. Open Access cites some laudable aims: that research publications from publicly funded research should be freely available to anyone in the world. The trouble now is that researchers are being asked to pay charges made by the research publishing companies in order to make each paper `open access'. This money is paid from some funds sourced from the government which is ear marked for the process, and it's not cheap. I write four or more papers per year, and for the last couple of papers I've recently written, the charge was about 2000 pounds for each. Why I can't dodge this charge because my papers are all already freely available on the internet mystifies me, although the rules are in a state of flux and I hope that this will change.

Research publishers have a shifting terrain in which to conduct business. On the one hand, their job is made so much easier by the internet and computers. There is computer software which the authors use to do the type-setting and drawing of figures and equations. Also, there is also the internet to send and sell articles on, so papers' dissemination is much simpler. They pay a few researchers a small fee to be editors, who decide which other researchers will review each paper. These reviewers work for free, receiving the research papers by internet or email and deciding whether to publish or not, or whether corrections are necessary. Research journals get thousands of subscriptions from university libraries around the world. It's almost money for old rope, and there is a recent proliferation of journals as companies have realised this: I regularly get emails from new journals asking me to be an editor or inviting me to submit an article for a `special issue' of the journal.

Getting your paper into a highly thought-of journal is a stamp of approval. If you have a lot of these on your CV, it helps you get a good job or your next grant. It also sends a message to other people that the work is serious and hasn't been just quickly bashed out on your laptop on the tube. I am in favour of peer review for a couple of reasons: firstly, it is a sanity check, and can give the author valuable and detailed feedback on your work, sometimes improving it. Secondly, the psychological pressure of researchers needing to get their papers into reputable journals helps keep standards up.

Physicists, computer scientists, mathematicians, statisticians, quantitative biologists and quantitative financiers have been working with the web-site xxx.arXiv.org for 20 years now, and it works fantastically well. We post every paper on it and it is available from the following day for anyone to freely view and print. After a week, we wait for comments from the research community (usually via emails) and make changes accordingly. Then, we submit it to a research journal. If the journal's reviewer requires changes, we make these, and re-post the new version to the arXiv. The first thing I do when I start work every day is to read all of the titles and abstracts from the previous day's arXiv on particle physics. If any of them look interesting, I click and it's there on my screen, to be read and printed if I need to. When you post a paper on the arXiv, it receives a date and time stamp that banishes any arguments about which researchers did what when (successive versions also get date and time stamps, and as a reader you can choose which version to read). People use the arXiv for many different forms of research: from information on how many times others have referenced a paper to other systemic research about the field. One ingenious application used the arXiv to make a map of the subject by its papers and their references, for instance.

There is no need for us to be paying these Open Access charges to the publishers, effectively giving them tax-payers' money that I would like to see directed to the research itself. The versions of our papers on the arXiv aren't type-set exactly as in the the journal, and they haven't been through the journal's proof-reading process, but all of the essential information is there. For free.

So, why doesn't everyone use the arXiv? For example, medical, biological and chemical scientific research is not published there, nor is that from the humanities and other arts. Yes, there are some commercial interests: people who are funded by companies that don't want their competitors to see the results. But surely such funders are against publication in any form?

There are some worries from researchers that haven't had the advantage of using the arXiv about how it works. Researchers worry that publishers may come after you with law suits if you put it on there first. Or they worry that publishers would reject a paper if it had been placed on the arXiv. The most prestigious scientific research journals do allow postings to arXiv (although one may not be able to update them in response to the editorial process). Other journals have been accepting arXiv postings as a fact of life for decades. The power is in authors' hands; if a publishing house decided to disallow arXiv postings, we simply would not send our papers to that house: for users, the arXiv has become too much of an integral part of the scientific discourse.

Surely now is the time for each subject to organise its own world wide repository like the arxiv, and for researchers to begin posting there. This would ensure that the goals of the Open Access standard can be met without additional hoards of public money being spent.