Sir

Your News story 'Open-access policy flourishes at NIH' (Nature 458, 690–691; 2009) raises the related question for publishers about the challenges inherent in providing the widest and most cost-effective access to quality scientific literature.

At the non-profit American Institute of Physics, we (like other publishers) are concerned about the inescapable realities of accomplishing this aim. The reality is that publishing a quality journal costs money. The trade-off is that those costs add value to a researcher's manuscript as it evolves from a draft submission to a final publication.

In our case, this transformation process requires peer review by tens of thousands of experts annually, as well as editing by postdoctoral physicists, refining of text and graphics, and use of increasingly sophisticated indexing and archiving measures. Government-mandated open-access policies will impose unintended negative consequences if they threaten the very business models that pay for quality publications.

Such mandates may not be necessary. We and other publishers are voluntarily developing copyright-friendly, fair-use policies that obviate the access issue. Harvard and the American Physical Society, for example, recently agreed on ways to facilitate authors' compliance with Harvard's new open-access policies when publishing in distinguished journals such as Physical Review, Physical Review Letters and Reviews of Modern Physics.

Even in the information age, maintaining the quality of the scientific literature costs money. How we can continue to improve access to scientific journals without compromising their quality is a question about which we should all be concerned.