Office of the Dean of University Libraries Carnegie Mellon 4909 Frew St., Pittsburgh, PA 15213 May 27, 2006 Elias A. Zerhouni, M.D., Director National Institutes of Health 9000 Rockville Pike Bethesda, MD 20892 RE: NIH Request for Comments Dear Dr. Zerhouni, On behalf of Carnegie Mellon and the roughly 4,000 faculty and staff we represent, I am writing in response to your Request for Information posted to the Federal Register on March 31, 2008. 1. Do you have recommendations for alternative implementation approaches to those already reflected in the NIH Public Access Policy? We have no recommendations for alternative implementation approaches at this time. 2. Do you have recommendations for monitoring and ensuring compliance with the NIH Public Access Policy? The best way to ensure the submission of articles and the inclusion of PubMed Central reference numbers in citations is to delay or prevent the distribution of funds. When investigators see that funding will be impacted, they will comply. The recent addition to the NIH FAQ that provides information about actions the NIH may take when investigators and institutions fail to take steps to ensure compliance with the NIH Public Access Policy is very helpful in this regard. Concerns about investigators submitting articles to PubMed Central that are not covered by the NIH Public Access Policy can be addressed through education. Concerns about investigators retaining the necessary rights to submit to PubMed Central articles that are covered by the NIH Public Access Policy can be addressed through education and by developing partnerships with more publishers so that the publisher will submit final published articles on behalf of the authors. In the absence of publisher partnerships, NIH must trust the institutions and investigators they fund to comply with the Public Access Policy – just as publishers must trust the authors they publish to comply with the terms of the copyright agreement. Institutions are working to educate faculty about their copyrights and responsibilities, including the NIH Public Access Policy and publisher policies regarding open access. Some publishers have suggested that NIH should closely monitor submissions to PubMed Central to prevent copyright infringement. This is not feasible because only the author and the publisher know what rights were transferred and what rights were retained. For this reason only copyright owners can legally claim copyright infringement. If copyright infringement occurs in PubMed Central, the copyright owner should notify NIH and the author(s) and the infringing work should be removed from PubMed Central. Expecting a government agency to monitor and ensure that funded authors comply with publisher copyright transfer agreements introduces a new approach to copyright law enforcement that would have implications far beyond the implementation of the NIH Public Access Policy. Serious consideration of this approach would require a much broader discussion. 3. In addition to the information already posted at http://publicaccess.nih.gov/communications.htm, what additional information, training or communications related to the NIH Public Access Policy would be helpful to you? The NIH Public Access Policy is profoundly important. Public access to the results of NIH-funded research will accelerate scientific discovery and give the American public vital information. The pattern that the NIH has established for gathering input from stakeholders and addressing their concerns is to be praised. The policy has just been implemented. Experience will indicate where clarification or modification is needed. For example, initial public comments revealed inconsistencies in the language used on the NIH website to describe procedures. These inconsistencies were a source of confusion. We are pleased that NIH responded promptly and made the language consistent. Over time, as researchers and their assistants work with the procedural information and tools provided by the NIH, other problems might surface that need to be addressed. Periodic requests for public comment will ensure that the dialog with stakeholders and the fine tuning of tools, training and documentation continue. 4. Do you have other comments related to the NIH Public Access Policy? Respect for copyright In the academic environment, the original copyright owners of written works are typically the authors who conducted the research. They are free to choose what rights, if any, they want to transfer and to whom. Concerns that publisher copyrights are being infringed by the mandatory NIH Public Access Policy are unfounded. Publishers have no copyrights unless or until authors grant them rights. Furthermore publishers have only the rights that authors grant to them. Publishers must respect author copyrights and the legal authority of authors to manage their copyrights as they see fit. Publisher approval or disapproval of how authors choose to manage their copyrights is irrelevant under the law. The NIH Public Access Policy does not introduce a new limitation or exception in copyright law, but rather constrains the contract terms to which funded authors can agree. This constraint does not constitute a compulsory license because authors as copyright owners are not legally required to permit others to use their work without permission. What they are legally required to do – if and only if they accept NIH funding – is to retain the rights necessary to deposit their work in PubMed Central. The NIH is not the copyright owner of peer-reviewed work based on the research it funds. To comply with the legislative mandate that the NIH Public Access Policy be implemented “in a manner consistent with copyright law,” NIH must look to the original copyright owners (the authors) to retain the necessary rights to comply with the policy. NIH has no authority to manage or to negotiate authors’ rights. Researchers are free to seek funding from the NIH, in which case they must comply with the NIH Public Access Policy, or to seek funding elsewhere. Publishers are free to allow authors to retain the necessary rights to comply with the NIH policy or to refuse to publish the results of NIH-funded research. Requiring authors to manage their copyrights and requiring publishers to respect author copyrights are positive developments in keeping with the letter and spirit of U.S. copyright law. Author – publisher relationships Concerns that the NIH Public Access Policy places authors in conflict with publishers suggest that the traditional practice, whereby authors routinely transferred all copyrights to their publishers, has become so entrenched that some publishers feel entitled to the transfer of all copyrights and question or resent the authors’ freedom to choose otherwise. There is no legal basis for this entitlement or questioning. The recommendation offered by some publishers that the NIH simply pay publishers for the articles rather than requiring authors or their institutions to negotiate copyright suggests that economic concerns might be driving publisher interest in retaining all copyrights. Concerns that the NIH Public Access Policy will influence the selection of publishers are real. However the claim that this influencing is unethical masks an underlying discrepancy in operating principles presumed to be allowable for different stakeholders. Publishers understandably want to manage the copyrights granted to them by authors. All copyright owners should want to manage their copyrights. Why, then, is it problematic for authors to manage their copyrights? Under the publish or perish paradigm, authors unreflectively transferred all of their copyrights to the publisher. This paradigm is crumbling, being replaced by a publish and perish paradigm because restricted access and the cost of scholarly journals are shrinking readership. The impact advantage of providing open access to scholarly work and new metrics for evaluating the quality and influence of scholarly work are changing relationships. Authors are beginning to reflect on their copyrights and to decide what rights they want or need to retain. The NIH Public Access Policy might accelerate a trend that has already begun, but the policy did not initiate the trend. Rather, the policy results from the trend. If indeed the list of PubMed Central (PMC) journal partners available on the NIH website encourages authors to publish in these journals (for the convenience of having the publisher submit their work in compliance with the NIH policy), publishers are free to become PMC partners. If indeed the NIH public access policy leads authors to favor journals with access policies consistent with the NIH policy, that will be beneficial. Publishers are free to change their journal policies. Competitive advantage is a key factor in the free enterprise system. Business models that lose their competitive advantage gain nothing by bucking the trends. I’m reminded of the story of the fellow who continued to manufacture buggy whips long after the horse and buggy were replaced by the automobile. His lamentations did not bring back the horse and buggy. Publishers are understandably concerned about retaining the commercial value of their copyrights. The allowable twelve-month embargo period will protect their interests. NIH-funded researchers work in disciplines where access to the most up-to-date information is essential to their work and continued funding. They cannot wait until the embargo period is over to see the newest research results. Furthermore reviews, commentaries, and research not funded by the NIH will not be available in PMC. Subscriptions and interlibrary loan will continue to be the primary means of prompt access to this information. The integrity and usefulness of the scientific record Some publishers have claimed that the NIH Public Access Policy will harm the integrity and usefulness of the scientific record because multiple versions of a given article will confuse readers and because final peer-reviewed manuscripts in PubMed Central might contain errors. These conditions predate and are irrelevant to the NIH policy. Multiple versions of many articles have been available on the Internet for many years (e.g., technical reports, conference papers, pre-prints, post-prints). Furthermore even published journal articles can contain errors; errata are issued later. Certainly scholars and medical professionals know that the published article is the scientific record. The many citizens who responded to the NIH’s call for comments, eager to have free access to research findings, will no doubt raise and discuss what they read with medical professionals. NIH archive and public access Many publishers have recommended replacing public access to peer-reviewed manuscripts of NIH-funded work in PubMed Central (PMC) with a link from PMC to the final published version on the publisher’s website. However, doing this would not accomplish one of the primary goals of the NIH, which is to create a searchable archive of the work it funds. An alternative recommendation that acknowledges the importance of an NIH archive suggests that PubMed Central become a dark archive used only by NIH and NLM for administrative purposes. Other users would access the published articles on publisher websites. The dangers of dark archives are well known. For example, a single dark archive offers a very low assurance level and requires elaborate systems to ensure that bit integrity is maintained. Currently the only efficient, affordable way to ensure the integrity and usefulness of a digital collection is to allow its use. The dangers of relying on publisher websites for open access to NIH-funded articles are also well known. Publishers can go out of business or fail to maintain their archive. Furthermore, they are free to change their access policies. Thus the only way for NIH to ensure that the research it funds is publicly accessible and preserved over time is to maintain an active, replicated archive. Thank you for inviting and considering our comments. Sincerely, Gloriana St. Clair, Ph.D. Dean of University Libraries