The Purpose of Publication and Responsibilities for Sharing
THE TRADITION OF SCIENTIFIC PUBLICATION
The roots of scholarly scientific publishing can be traced to 1665, when Henry Oldenburg of the British Royal Society established the Philosophical Transactions of the Royal Society. Oldenburg was motivated, in part, by a desire to remove himself as diplomatic interlocutor between the dispersed, independent scientists of the time with whom he communicated individually. The aim of the new publication was to create a public record of original contributions to knowledge and to encourage scientists to “speak” directly to one another. By providing intellectual credit publicly for innovative claims in natural philosophy, the journal encouraged scientists to disclose knowledge that they might otherwise have kept secret.
The Philosophical Transactions of the Royal Society created a sense of competition among scientists to be the first to publish a new scientific finding, an incentive that is continued in modern scientific journals. If the journal is a prominent one, publication endows the author with an extra measure of prestige. In addition, as Cell editor Vivian Siegel and other workshop participants noted, publications also yield indirect rewards. For example, they affect a researcher’s job prospects and ability to be promoted or gain tenure. Publishing a scientific paper can result in fruitful new scientific collaborations, including financially profitable arrangements for authors in academe, as a result of commercial overtures for collaboration or consultancy.
Publishing also holds some risks for an author. Competitors might use results presented in a paper to advance their own research and “scoop” the original author in future publications. The careers of young scientists might be particularly vulnerable to having prospective research “picked off” by others. (However, if a researcher chooses not to publish his or her results or chooses to delay publication, someone else might publish the same findings first and receive the credit.) Another risk associated with publishing is that other researchers will use information presented in a paper to invalidate or question the author’s own findings, and publish conflicting results.
Are the benefits and risks of publishing any different for companies whose investigators publish than those for academic scientists? It was pointed out at the workshop that companies whose scientists publish their findings typically receive the intellectual credit, recognition, and prestige that come with such disclosure to the entire scientific community. Such nonfinancial benefits can translate into increased publicity and increased perceived value of a company to potential investors and business partners. They also strengthen the scientific reputation of companies in the eyes of potential collaborators. By encouraging others to use their methods and materials, companies can develop a net of researchers who are extolling and extending the value of the technology that the company has published. Moreover, companies that encourage their investigators to publish are attractive to employees or potential employees who wish to build and maintain their publication record, either in anticipation of someday returning to academe, as a vehicle for facilitating their participation in and recognition by their peer scientific community, or in buttressing their own career prospects within the company.
For a for-profit research entity, publication also carries financial risks. By revealing proprietary data or other trade secrets, publishing may harm a company’s competitiveness in the marketplace and thus endanger the return to investors. A competitor might use information disclosed in a scientific paper to develop a competing product or otherwise gain commercial advantage or to discredit the product claims of the company making the disclosure.
While companies whose scientists publish may worry about their competitive edge in the commercial market, researchers in academe worry about gaining a competitive edge in the rewards process and about getting their research grants renewed. Where academics are rewarded by priority, “fame,” and career advancement, companies whose investigators publish receive benefits in terms of visibility, public relations, and validation. Although there are different tradeoffs involved in publishing, in practice, researchers from these two worlds often have similar goals and are motivated by common incentives. Their common interests converge in the forum of scientific publication.
PUBLISHING AND COMMUNITY STANDARDS
By facilitating communication between individuals who had worked in isolation from one another, the Philosophical Transactions of the Royal Society also contributed to the development of a scientific community. As a result, modern journals do more than simply register the intellectual accomplishments of individual scientists; they record a collective body of knowledge. Journals are a centerpiece of the scientific enterprise and serve as a focal point for the description of scientific results. Journal articles supply information that helps scientists to develop new hypotheses, and they provide a foundation on which new scientific discoveries and inventions are built. As Eric Lander noted at the workshop, “science is fundamentally a cumulative enterprise. Each new discovery plays the role of one more brick in an edifice.” Authors cite previously published papers to make a case for their conclusions that is based on a combination of previously documented scientific evidence and the new information they have gathered. Scientific journals, many established by learned societies, provide a forum for a continuing dialogue of sorts, as authors discuss findings that add new pieces to others’ previously published results or announce alternative conclusions to those made by other authors or contradict them. Science moves forward in this way.
Because publication is central to the activity of the scientific community, and consequently, to scientific progress, principles and standards
that govern an author’s responsibilities related to publication have always been paramount. As the 1992 National Research Council report Responsible Science observed, “For centuries scientists have relied on each other, on the self correcting mechanisms intrinsic to the nature of science, and on the traditions of their community to safeguard the integrity of the research process. This approach has been successful largely because of the widespread acknowledgement that science cannot work otherwise, and also because high standards and reputation are important to scientists” (NRC, 1992).
Because standards related to publication are so important to the functioning of the community, calls for the publication system to adapt to the different risks of publication to scientists working in different circumstances are not easily implemented. Chapter 5 addresses specific arguments related to exceptions, but in general, applying a standard to some authors and not others weakens the incentive of distinction that has attracted scientists since Oldenburg’s day to publish publicly in a journal. When exceptions to the community standard are sought and granted, there is a danger that the value of publishing is diminished, not only for the author who requests an exception, but for the entire community. Moreover, if the same standard does not apply to all authors, then the community cannot assume that the quality of scientific papers and the information they purport to represent is reliable. That jeopardizes the integrity of the publication system.
That is not to say that publication-related community standards are insensitive to other important societal interests, such as protecting the identities of research subjects. Measures to protect that information do affect how data are reported and made available to other investigators; nevertheless, the community has striven to find ways to maximize the availability of relevant data without compromising privacy.
A current topic of discussion in the scientific community is the possibility that published information in the life sciences will be exploited by bioterrorists. It is too early to say where those discussions will lead, but current community standards abide by regulations on access to some research materials (for example, radioisotopes, explosives, controlled
substances, and pathogens) for good reason. If additional safeguards are found to be necessary in providing access to research data and materials, the community must make accommodations for them.
Similarly, the community complies with prohibitions imposed by some nations on the distribution of biological materials and organisms collected in those countries. Biological materials that are paleontological, archeological, or anthropological in nature, and sometimes samples of organisms, may by national law be required to be deposited in the country of origin, and even when material is allowed to be exported, there are often legal restrictions on its subsequent distribution and use. For example, the commercial use of such samples may be prohibited or restricted. Nonetheless, all such material is made fully available for study at the repository, and not normally under the control of the authors who published results derived from studying it. The details of the results of the original study, and images, DNA sequences, and other information derived from the specimens, are also made available.
The principles and standards of scientific publication are also consistent with society’s interest in the applications of scientific knowledge and their economic and other benefits. An author who publishes a paper is expected to share materials related to that publication to other scientists for research purposes, but that does not prevent an author from seeking intellectual property rights protection in order to realize the commercial value of those materials. To encourage the disclosure of scientific information, the patent system bestows inventors of a novel, nonobvious, and useful innovation with the right, for a limited time, to prevent others from making or using that innovation, unless licensed to do so. Scientific publication provides no such incentive, but to the contrary, encourages other scientists to use and integrate into new research those things described in a scientific publication. An author who publishes a scientific paper describing a patented process, for example, may have a legal right to prevent others from using it, but the scientific community holds the expectation that an author will make available a license to use that process for research. From a social perspective, the two systems are complementary: patenting fosters the commercialization of ideas;
scientific publication communicates the ideas that build the edifice of science. Scientific publications also influence the issuance of patent rights by defining the landscape of the “prior art” and “obviousness” criteria used in assessing the novelty of putative patent claims.
JOURNAL POLICIES AND COMMUNITY STANDARDS
Journals have their own policies that describe an author’s responsibilities related to publication and sharing publication-related data and materials. Publishers of journals include for-profit companies and not-for-profit enterprises, such as university presses, scientific societies, and associations, and each publisher is motivated by the intellectual objectives and fiduciary responsibilities of its own constituencies. Journal editors often compete for papers that increase the impact and standing of their journals in the scientific community and their mass media coverage. On occasion, journal editors have been willing to make exceptions to their usual policies on data sharing in return for the opportunity to publish a paper they believe will be of high impact in the scientific community and, increasingly, in the general public.
The extent to which journals state their policies for the sharing of materials and data is highly variable (Table 2–1). That variability and the diverse nature of journals might suggest that common principles and standards do not exist. But even the stated policies of journals do not capture what are generally recognized as accepted practices and expectations of the community. For example, most journals today explicitly require that authors provide enough detail about their materials and methods to allow a qualified reader to replicate all experimental procedures. A logical, often implicit, extension of that requirement is that authors must make available the data and materials needed for others to verify or refute the findings reported in a paper. Thus, for example, in a paper citing genetic results from one or a series of organisms, voucher specimens should be cited and deposited in an appropriate public repository where the identity of the organisms can be checked by subsequent workers (with the obvious exception of well-known and easily-available
Table 2-1. Policies of 56 Most Frequently Cited Life-Science and Medical Journals
strains). Insofar that scientific publication is central to the forward progress of the scientific community, it is presumed that an author must provide data and materials in a way that others can build on them. These widely held expectations are not necessarily incorporated in current journal policies.
THE PRINCIPLES OF PUBLICATION
At the workshop and in its deliberations, the committee attempted to distill the community’s most basic interests in the process of publication. It found that a majority of the scientific community held common ideas and values about publication and the role it plays in science, and that those ideas have guided the development of community standards that facilitate the use of scientific information and ensure its quality. Central to those ideas is a concept the committee called “the uniform
principle for sharing integral data and materials expeditiously (UPSIDE),” as follows:
Community standards for sharing publication-related data and materials should flow from the general principle that the fundamental purpose of publication of scientific information is to move science forward. More specifically, the act of publishing is a quid pro quo in which authors receive credit and acknowledgment in exchange for disclosure of their scientific findings. An author’s obligation is not only to release data and materials to enable others to verify or replicate published findings (as journals already implicitly or explicitly require) but also to provide them in a form on which other scientists can build with further research. All members of the scientific community— whether working in academia, government, or commercial enterprise—share responsibility for upholding community standards as equal participants in the publication system, and all should be equally able to derive benefits from it.
Along with UPSIDE, five additional principles guide the development and implementation of community standards. Chapters 3 and 4 discuss those principles and the nuances of how they are embodied in examples of community standards for sharing data, software, and materials. New community standards are likely to evolve as science itself changes, but the principles remain a fundamental underpinning of the their development. The principles motivate the creation of standards that maximize the value of scientific findings to the community, because this has proved to be the way that science progresses most rapidly. In addition to the principles of publication, Chapters 3 and 4 include the Committee’s recommendations for increasing the effectiveness of community standards for sharing data and materials.