Sir

Raghavendra Gadagkar (Nature 453, 450; 2008) argues that the open-access 'pay to publish and read for free' model leads to a disadvantage for scientists in developing countries. I disagree. Gadagkar correctly states: “page charges may be waived for authors who cannot afford to pay.” He then adds: “a model that depends on payment by authors can afford only a few such waivers.” This is not necessarily true: for example, some open-access journals provide discounts to particular institutions.

I would prefer to see what little money is available to a developing country spent on helping to publish their scientists' papers rather than financing publishing houses based in First World countries. At present, open-access publication may be hard for those in the developing world to afford, but in the long run it will be advantageous, offering them free access to educational and academic resources.

Most important, the future of open access probably does not lie in journal publishing models. The huge success of online literature databases such as arXiv (http://arxiv.org), free to publish and access, is significant. Such databases currently host mostly non-peer-reviewed preprints, and so are of little value for career building. But academic organizations throughout the world could, if they wished, build an equivalent archive of peer-reviewed papers.

I also disagree with Gadagkar's view: “If I must choose between publishing or reading, I would choose to publish". No one can expect to do serious science without access to the current academic literature.

Although many subscription journals are free to access online in developing countries through the HINARI, AGORA and OARE initiatives of the United Nations, the principle remains that if you cannot afford to read, you automatically cannot afford to publish. Perhaps Gadagkar will agree next time he is denied access to a fundamental paper for his research because his institution does not subscribe to it.