A Model Text Recycling Policy for Publishers

This viewpoint presents our new TRRP Policy on Text Recycling, 12 a model text recycling policy developed from our prior work.


A Model Text Recycling Policy for Publishers
an author-publisher agreement. 2 Thus, there is a need for clear and consistent guidelines on text recycling.
Publishers' policies on text recycling to date, however, have been unclear, as many scholars and journal editors have noted. As one part of their study of text recycling, Horbach and Halffman 3 investigated how often journal policies addressed text recycling. They found that, "[S]tatements on text recycling are rather uncommon in journals' policy guidelines," and that for those journals in which they identifi ed cases of text recycling in their corpus, these "almost uniformly lack statements on text recycling." Others have commented on the challenges faced by editors and authors resulting from incomplete and inconsistent guidelines. 2,[4][5][6] The Text Recycling Research Project (TRRP) defi nes text recycling as the reuse of textual material (prose, visuals, or equations) in a new document where (1) the material in the new document is identical to that of the source (or substantively equivalent in both form and content), (2) the material is not presented in the new document as a quotation (via quotation marks or block indentation), and (3) at least one author of the new document is also an author of the prior document. Under this defi nition, text recycling can be ethical or unethical, appropriate or inappropriate, depending on the details of each case. It may encompass any amount of text, from a single recognizable phrase to an entire manuscript, and it includes both verbatim replication and reused material that has been disguised via superfi cial alterations in appearance without changing its substance. Text recycling may or may not include citation of the source; whether a citation is appropriate for any instance of text recycling depends on both ethics and attribution practices in the fi eld. Like the Committee on Publication Ethics (COPE), we avoid the term "self-plagiarism" because of its inherently derogatory connotation.
Drawing on our research to date, 7 the TRRP has already produced a number of documents 8

Introduction
While scientists' new publications are generally expected to make substantive contributions distinct from their earlier papers, the close relationship among papers often requires authors to repeat some content. Such recycled material typically consists of methodological details but may also include background material such as defi nitions or exposition that describes prior research.
In many fi elds of science, "text recycling" (sometimes inaccurately called "self-plagiarism") is not an aberration but a common writing practice. 1 Deciding whether any instance of text recycling is ethical, legal, and appropriateand possibly even desirable-depends on factors such as the amount and nature of the recycled material as well as copyright laws and any limitations on reuse that are part of

The Challenges of Text Recycling
Researchers face a number of challenges when deciding whether to recycle text. The most signifi cant may be the inability to distinguish between different types of recycling and the consequent uncertainty about what is appropriate: Is recycling text from a grant proposal different-ethically or legally-from recycling material from a published article? Is it equally appropriate to recycle text from methodology sections and results sections?
Most researchers and editors seem to agree there are important differences between instances of recycling, but they may have diffi culty describing these differences in concrete terms. [13][14][15] This diffi culty is exacerbated by inconsistencies in the vocabulary used in discussions of text recycling. Professional organizations sometimes use the same terms to describe different recycling practices, making it diffi cult for authors to compare publishers' expectations. 2 Furthermore, publisher policies that address different types of text recycling are not always available in a single location. Instead, they are often sprinkled across multiple sections of policy documents or ignored altogether.
Publisher policies have also tended to ignore a crucial aspect of authorship in scientifi c settings. By defi nition, recycling involves reuse of "the author's own" material, but what should be considered "one's own" work? Scientists rarely write as solo authors, and as our research has shown, papers from research groups often have overlapping but not identical authors. In fact, in our analysis of pairs of papers produced under the same U.S. National Science Foundation grant, less than 7% had identical authors. 1 To date, no existing policies on text recycling have addressed this common situation.
Incomplete knowledge combined with inadequate guidance has left many editors in a diffi cult position: They recognize the absurdity and ineffi ciency of asking authors to reword recycled material merely to make it appear different, but they are also reluctant to leave recycled material in place for fear of violating vague ethical norms or copyright laws.
To address these challenges, the TRRP embarked on a series of studies to understand current beliefs and practices with regards to text recycling, the extent to which text recycling occurs in scientifi c publications, and the relevant legal issues of copyright and contract law. Our fi ndings include the following: 1. A majority of journal editors and editorial board members are willing to accept limited text recycling, particularly when the recycled material consists of methods or background material. 13,15 2. Editors are often uncertain as to whether text recycling infringes on copyright and sometimes direct authors to "reword" recycled text, masking the recycling by rearranging phrases and using synonyms. 13,15 Such rewording, however, does not satisfactorily resolve concerns about text recycling. 16 3. Both expert and novice researchers are confused about the ethics of text recycling, sometimes resulting in substantial disagreements about appropriate practice. 14 4. Text recycling is common across linked studies in a publication chain, often spanning multiple documents with varying authorial teams across several years. In spite of the negative connotations that have often been associated with text recycling, limited recycling is standard practice in much research writing. 1 In the course of this work, we recognized the need for terminology that could adequately distinguish between different types of recycling. The taxonomy we developed (see Moskovitz 17 ) is discussed below.

The TRRP Policy
The TRRP now offers to the scientifi c publishing community the fi rst comprehensive and research-based model text recycling policy, the TRRP Policy on Text Recycling. 12 This policy is intended to provide clear, straightforward guidance to authors in diverse publishing contexts. It has been thoroughly vetted by the TRRP Advisory Board, 18 whose members include offi cers of COPE and the Council of Science Editors (CSE) as well as representatives from for-profi t and nonprofi t publishers, government research agencies, and research integrity organizations.
The two major issues that apply to most instances of text recycling-authorship and transparency-are addressed fi rst. For authorship, the policy indicates that when any authors of the prior document are not authors of the new document, their permission should be sought when practical. For transparency, the policy states that when authors have included recycled material in a manuscript, that recycling should be disclosed during the submission process; editors can then provide guidance on whether it is appropriate and how authors should notify readers within the manuscript (if needed). These two policies alone will likely reduce much of the ambiguity and confusion caused by text recycling in the publication process.
The remainder of the policy is organized according to the type of text recycling using the TRRP terminology we developed. This terminology, as explained in more detail in our Understanding Text Recycling: A Guide for Editors, 11 is as follows: • developmental recycling: reusing material from one's unpublished documents • generative recycling: reusing portions of one's previously published documents in a new work that makes an original intellectual contribution • adaptive publication: republishing an entire document or its central part(s), modifi ed to fi t a new context (e.g., new audience, new genre) • duplicate publication: republishing a work having the same genre, content, and target audience as the previously published work The policy addresses recycling limits qualitatively rather than setting specifi c numeric thresholds such as word counts or percentages. Publishers can then establish their own quantitative limits for internal use if desired.
The last section alerts readers to recycling practices that will be journal or publisher specifi c. Because publishers' positions will differ as to the acceptability of recycling from preprints and conference proceedings or whether they publish translations, readers are directed to locate journalspecifi c policies for these cases.
Our policy should serve the needs of many publishers of original research, and we encourage adoption of the TRRP policy without modifi cation when possible. While every detail may not be precisely what any editorial board would prefer, we encourage careful consideration regarding how signifi cant any minor differences in preference might be in practice. We expect that many publishers will fi nd the greater benefi t in consistency-both for their authors and the workloads of their editors. That said, recognizing that the TRRP policy will not suit some publishers without modifi cation, we also offer an accompanying TRRP Guide to Developing Text Recycling Policies. 19 This guide maps out the issues that we feel every text recycling policy should address, explains what is at stake for each issue, and offers discussion questions to facilitate policy making.

Legal Issues and Their Resolution
The policy guidelines we offer here are intended to promote ethical text recycling practices based on disciplinary norms and fi ndings from our research. Though the policy does not explicitly address the legal aspects of text recycling, analyzing the relevant aspects of copyright and contracts has been one major dimension of our research, and these have proven to be quite complex and sometimes challenging to navigate. One complication is due to the wide variety of text recycling practices-from the clearly trivial, such as reusing a single clause in a description of methods, to the clearly problematic, such as surreptitiously republishing one's entire paper. Another complication is copyright law itself: Not only do copyright laws differ by country, but there are no laws or precedent cases in any jurisdiction that directly address text recycling in scholarly writing.
Additional complications come from the author-publisher agreements that authors usually sign in order to have their work published. Almost all such agreements include contractual language that impacts authors' text recycling rights-explicitly, implicitly, or both. To know whether recycling an author's previously-published material would be contractually allowed, editors would need to be familiar with each of the publishing agreements previously signed by the author(s) and be able to interpret the legal implications for recycling-clearly an unreasonable expectation.
Given the challenges in evaluating whether any given instance of text recycling would be legal, it is unsurprising that some editors take a risk-averse approach, directing authors to avoid text recycling altogether even in cases where the legal risks are actually negligible. In our research, we have been unable to locate even a single legal case brought to trial for text recycling in research papers, even though it has long been common practice in STEM research writing. And because the practice is so common, even those publishers with the resources to bring legal action would be reluctant to do so since they almost certainly have many similar instances of recycling within their own publications. Nevertheless, journals which adopt the TRRP policy may still have concerns about the legality of recycling in specifi c manuscripts; for those cases, editors can ask authors to follow the same process they use for obtaining permission for the reuse of fi gures, long prose passages, or other previously published materials.
While the practice of asking authors to obtain permissions is legally sound, it is more cumbersome than necessary for the majority of cases in which authors are reusing portions of their own previously published work ethically and responsibly. In our view, publishing agreements should explicitly allow authors to recycle from their published work in future publications when they do so within the bounds of ethical guidelines (such as the TRRP Best Practices for Researchers 9 ). Thus, we are currently formulating language that publishers can use in their publishing agreements to make the legal situation simpler and more transparent. For authors, this modifi cation would clarify in advance what they will be allowed to recycle; for editors, it would eliminate the work of managing a permissions process for the most common instances of text recycling.

Final Thoughts
The text recycling policy we announce here is the culmination of 5 years of focused work. While no single policy can be perfect, we fi rmly believe that widespread adoption (or, at the least, adaptation) of this policy would be a major step in addressing this thorny problem of publication ethics. We recognize that this will involve a nontrivial amount of work on the part of publishers. However, we believe that the long-term benefi ts-greater clarity for all stakeholders and reduced work and frustration for editors-will make the effort worthwhile.