Anonymous Author 1; Anonymous Author 2; Saidman, Lawrence J. MD
The 2 most common forms of scientific misconduct are fraud and plagiarism. In a 2011 editorial in this journal, Shafer1 describes a 5-level taxonomy of plagiarism ranging from the most serious: intellectual theft (the use without permission or attribution of another’s actual words, passing them off as one’s own) to least serious, text recycling or self-plagiarism (using one’s own words in more than 1 publication), which is not considered misconduct.1 While fraud including fabrication of results is difficult to detect and prevent, the routine application of plagiarism-detection software before review reduces the likelihood of intellectual theft to a vanishingly small fraction of articles published in peer-reviewed journals.
And voilà! The problem of plagiarism is solved, or so we thought. But we were mistaken.
In this article, we describe another form of intellectual theft, equally as deliberate as that occurring in peer-reviewed journals but involving multiauthored textbooks undergoing serial editions. We believe that this form of plagiarism represents a significant breach of the ethical foundation of academics.
Two authors of this article have been victimized by this practice and choose to remain anonymous because our goal is to inspire change rather than to focus on the specific unethical actions of individuals. Following descriptions of these experiences, we discuss the responsibilities of those at each stage of the publication process and propose preventative solutions for what we believe represents academic misconduct. Both Figures 1 and 2 are modified facsimiles of published textbook chapters. The highlighted portions were plagiarized from the previously published chapters written by our 2 anonymous authors. In rendering the images, we defined plagiarism as the use of verbatim text, figures, or tables from a previous publication without attribution to the author. The images have been intentionally altered beyond recognition and reviewed by the Editor-in-Chief of Anesthesia & Analgesia as well as the named author of this essay to ensure that the representation of the plagiarized portions is accurate.
TWO AUTHORS’ EXPERIENCES
Author #1 was invited to write a chapter for a new edition of a major textbook. Written with a colleague, the completed chapter consisted of more than 26,000 words of text, more than 40 figures, tables and summary boxes, and hundreds of references. Years later, the subsequent edition of the textbook was published. Despite being credited to 2 new authors, more than 85% of the text had been copied from the previous edition chapter written by Author #1 and a coauthor (Fig. 1). More than the first 100 references in the new edition chapter are identical to those in the original chapter and appear in the same order. The principal contribution of the new authors was the addition of a summary of information within the chapter. The only acknowledgement to the original is a footnote stating, “The [new] authors would like to acknowledge the significant contributions of [Author #1 and his or her coauthor].”
Author #2 was asked by a colleague to prepare a chapter for a new, multiauthored textbook devoted to a subspecialty of anesthesiology. The single-authored chapter consisted of more than 11,000 words of text, 10 to 15 original tables and figures, and 75 to 100 references. Years later, a second edition of the text was published containing a chapter authored by 2 other individuals. More than 80% of the text in this second edition chapter is verbatim material from the original chapter. The majority of the tables and figures had been created by Author #2 and published in the first edition as well (Fig. 2). More than 70% of the references are identical to those from the original chapter. All appear in the same order as those in the first edition. The only attribution is a footnote stating, “A portion of the material in this chapter was published previously as….” followed by the first edition chapter title and original author.
In both of these cases involving serial editions, there are no quotation marks surrounding verbatim passages or formal references to the original chapters. Neither Author #1 nor Author #2 was contacted by an author, editor, or publisher of the subsequent editions before publication. They only learned that their intellectual contributions had been copied by the revision authors when they read their own words in the subsequent edition.
STAKEHOLDERS AND THEIR RESPONSIBILITIES
To better understand how academicians and prominent medical textbook publishers become involved in what appears to represent intellectual theft, we describe the roles and ethical responsibilities of initial authors, authors of revised chapters, editors, publishers, and educational institutions. Our focus is on the ethical rather than legal issues pertinent to new authorship of chapters in revised textbooks.
Initial Chapter Authors
Initial chapter authors are those contributing an entirely new chapter for a textbook. An author or contributor agreement is signed and includes assigning copyright to the publisher or, in some cases, the editor. Contributor agreements typically specify that authors are responsible for all material they submit, for obtaining permission for material that is copyrighted by other publishers, and for providing appropriate attribution to previously published material written by other authors.
Chapter Revision Authors
Editors preparing a revision of an existing textbook may engage new authors to revise a chapter with or without prior discussion with the original authors. Revision authors have the same responsibility as initial chapter authors to ensure integrity of their manuscript including obtaining proper permissions and including complete attribution to previously published material. From a legal perspective, permission to republish material from a previous edition (assuming the same publisher) is rarely required because, unless otherwise specified, the publisher owns the copyright on the earlier edition. However, the fact that there is no legal requirement to obtain permission from the publisher does not grant the new authors the right to misrepresent the words of the original author as their own, and thus claim credit for another person’s scholarship.
Editors engage chapter authors. For a revised edition, editors decide whether to invite original authors to revise their chapters or to engage new authors. Authorship may be changed for several reasons, including editors’ expectations of writing quality and timeliness, relationship building, fostering individual careers or specific departments’ reputations, and enhancing the prestige of the new edition by engaging better recognized experts.
As with the original chapter authors and revision authors, editors are also responsible for ensuring integrity of the material to be published. These responsibilities include assessing whether authors have obtained appropriate permission for material under copyright. Editors must also assess whether plagiarism has occurred in the preparation of the material for the textbook. A key issue is whether the editors of the textbooks in which our plagiarized chapters were published realized that the revised material was overwhelmingly identical to that which had been previously published. We suggest 2 possibilities. (1) Editors failed to compare the original and revised versions and thus were not aware that the authors of the revision had taken credit for work that is predominantly the creation of others. (2) Editors were aware of the extent of duplication but did not believe that the ethical standards governing plagiarism of peer-reviewed articles also should apply to textbooks.
In one of our cases, an associate editor of the revised textbook was also a coauthor of the revised chapter and, therefore, was aware of the small fraction of original material. We also discovered a third anesthesia textbook chapter in which the editor of the textbook is the sole named author of a chapter that consists overwhelmingly (approximately 85%) of material from the previous edition. The contribution of the original authors is acknowledged with only a footnote stating that a portion of the material had been published under separate authorship. Similar to the experience of Authors #1 and #2, these authors were not contacted before the publication of the subsequent edition.
Publishers, Copyright, and Attribution
Copyright protection is included in Article I of the United States Constitution. For works created after 1978, copyright protection lasts for 70 years after the death of the last surviving author. When works are created for hire (as part of one’s job), protection is even longer—95 years from publication or 120 years from creation, whichever is shorter.2 Copyrighted material includes words, poems, songs, computer-generated images, artwork, charts, graphs, tables, and more. Contributor agreements for textbooks generally include assignment of copyright to the publisher, who earns income from copyrighted materials via direct sales, licensing, and permissions to others. When a different publisher owns the copyright on a work, permission and a formal citation are required when it is to be reproduced, even when the author is identical.
We have reviewed the public Web sites of 4 prominent publishers of anesthesia textbooks: Elsevier B.V.,3 Lippincott Williams & Wilkins,4 Springer,5 and Wiley-Blackwell.6 In addition, we have contacted those publishers with whom we have personal contact as authors and editors. All have easily accessed and very explicit instructions regarding when and how to obtain permission for material under copyright by another publisher and how to credit such work. Wiley clarifies on its Web site that permission must be obtained for more than 300 words of text in total. When permission is obtained, the acknowledgement of the original source is provided. However, if no permission is required (the copyright is already held by the publisher), in most cases, the guidelines and rules are less clear. Wiley outlines on its Web site instructions to authors the importance of giving credit to the original creator of work, regardless of whether permission is required. For others, unless permission was required, we were unable to discover written instructions regarding attribution.
In an effort to reduce costs and simplify publishing, the International Association of Scientific, Technical & Medical Publishers has issued guidelines that include the opportunity for members to automatically opt out of requiring formal permission from another member for limited amounts of material under copyright protection. For a single book chapter, this includes not more than 3 figures or 400 words.7 These guidelines do not abrogate the obligation of authors to provide proper attribution for material that is not their creative work.
How much of a book chapter can ethically be republished under new authorship without including the original writer as an author? We have not been able to find written publishers’ guidelines on this key point. This differs from permission or attribution, perhaps because it has not historically impacted the economic value of the texts for publishers. When contacted specifically on this question, a publisher of multiple anesthesia textbooks replied that their informal guideline was that when greater than 50% of the text of a chapter was to be used, the original author was retained as a coauthor and a contributor agreement signed. They went on to say that this was “more or less a courtesy” to the original author because copyright was already held. Fifty percent!
Educational institutions encourage publishing by faculty members via their promotions process, the often-called “pressure to publish.” Such academic activity by faculty members is viewed as an important component of the departmental and institutional scholarly and educational mission. Academic salaries support much of the time required to write chapters in textbooks. Therefore, universities are key stakeholders in the legitimate recognition of scholarly work supported predominantly by their respective institutions.
Plagiarism can be, and has been, career crippling or career ending. All universities have policies regarding academic scholarship and ethics including intellectual honesty. This typically includes an unambiguous definition of plagiarism as misrepresentation of the scholarship of another author as one’s own scholarship, accompanied by a clear statement that such misappropriation of scholarship represents academic misconduct. A particularly germane and explicit example of an educational institution addressing this issue can be found on the public Web site of Harvard Medical School,8 a stated list of recommendations for authors and editors of medical textbooks. They include both attribution and authorship: when a revision is planned, the editors are to communicate with all previous authors regarding their expected level of contribution and the planned attribution of their work in the subsequent edition, as well as a recommendation that if the previous contributor’s work accounts for a “significant identifiable portion” of the newer edition, that the previous author be retained as a contributor to the newer edition. They also state that “Editors of medical textbooks are responsible for maintaining academic standards of scholarship and proper attribution…” and encourage editors to reach an agreement with their publishers stating that academic standards of attribution will be followed.
CURRENT STANDARDS IN THE SCIENTIFIC COMMUNITY
Routine use of software to detect plagiarism is an effective tool to combat plagiarism at a time when the scientific community has shown keen interest in the problem. CrossCheck™ software from iThenticate® (iParadigms LLC, Oakland, CA) is used by more than 365 publishers, including Lippincott, Williams & Wilkins, the publisher of Anesthesia & Analgesia, and all others mentioned earlier. Their database includes nearly 80,000 global scientific, technical, and medical journals as well as chapters and books.9 Participating publishers share their content with iThenticate for use in screening for plagiarism before publication; however, not all textbooks are available in digital form and therefore may not be included in their database.
Editors of reputable journals have worked together to develop ethical standards, and organizations such as the Committee on Publication Ethics10 and the International Committee of Medical Journal Editors11 have issued ethical guidelines, requirements for authorship, and codes of conduct and best practices for journal editors and journal publishers. When plagiarism is detected after publication, the response varies according to the severity and frequency. Actions include retraction, watermarking of the manuscript as retracted in searchable databases, sanctions by journal editors against authors guilty of plagiarism, requests for letters of apology, and contacting the academic institutions affiliated with the authors found to have plagiarized.1,10 Other organizations with a focus on publication ethics in peer-reviewed journals include the World Association of Medical Editors12 and the Council on Science Editors.13 Their Web sites are informative and also include recommendations on authorship criteria. However, we were unable to find any materials that provide guidance on serial edition textbook chapters.
The United States Government’s firm stand against plagiarism is implemented by the Office of Research Integrity,14 a branch of the U.S. Department of Health and Human Services (HHS). The Public Health Service policies on research misconduct clearly state that plagiarism in any work funded via federal grants is prohibited. They define plagiarism as “the appropriation of another person’s ideas, processes, results, or words without giving appropriate credit.” Institutions are required to have policies in place prohibiting research misconduct including plagiarism and must investigate allegations to receive federal research grants. The HHS administrative response to plagiarism may include, but is not limited to sanctions, retraction, correction, letters of reprimand, suspension or termination of grants and contracts, and even recovery of public funds.15 The Office of Research Integrity maintains an active Web site that includes publication of research misconduct case summaries, blogs, newsletters, and a module on avoiding misconduct.14
Text recycling (also called self-plagiarism) is the reuse of specific language in multiple publications by the same author. It is not considered academic misconduct unless the originality of the work is misrepresented, and accepted limits on copying one’s own published words without specific quotes or reference are considerably less restrictive than those involving copying by others.16 In certain contexts, such as the Methods section of a scientific article, recycled text may exceed 30% to accommodate explanation of complex scientific methodology that is repeatedly used by investigators in separate studies.1 Otherwise, it appears that no more than 10% to 15% recycled text is considered acceptable without appropriate citation.17,18 We note that the degree of copying by others without specific quotes or reference that occurred in the cases we cite far exceeded even the norm for the recycling of one’s own words.
RECOMMENDATIONS—PREVENTION IS THE BEST MEDICINE
No definition of plagiarism places textbook chapters into a separate and privileged category or excludes either authors of textbook chapters from the unambiguous prohibitions against plagiarism of another person’s work or textbook editors from their responsibility to assure integrity of the text. We call upon authors, editors, and publishers to uphold the same ethical standards regarding plagiarism in textbooks as those currently enforced in peer-reviewed, scientific journals. If it is unacceptable to present as one’s own work a portion of a journal article copied from another author, then it is equally unacceptable to present as one’s own work a portion of a book chapter copied from another author. Regarding the examples we cite, the nonspecific footnotes acknowledging “significant contributions” or “a portion of the material in this chapter was previously published as…” are inadequate given the lack of communication with the original authors, absence of formal reference, and small fraction of new material by the revision authors. Plagiarism is plagiarism.
In addition, plagiarism and the associated misappropriated attribution of scholarship are more difficult to address post facto when they occur in a published textbook rather than in a peer-reviewed journal. Even if textbook publishers and purchasers become aware of ethical transgressions, books are unlikely to be removed from personal or library shelves. Most book chapters are not currently included in searchable databases such as the U.S. National Library of Medicine’s PubMed, and therefore once printed cannot be effectively deleted. Unless pursued in academic misconduct investigations, plagiarizing authors and their editors are likely to retain academic credit for the misappropriated work, while the original authors are denied due credit for the publication of their work in a subsequent edition. Publishers who own the copyright have little economic interest in correct attribution other than to minimize potential legal action. Therefore, prevention is the most potent corrective strategy.
Authors must take responsibility for the integrity of their work. If the original authors revise their work for the same publisher, they can do so without attribution. To avoid intellectual theft, if new authors are engaged (or editors become contributing authors of the new chapter), work must either be entirely original or those sections taken from the previous authors must be clearly acknowledged. When asked to write a chapter for a new edition of a textbook, in addition to providing proper attribution if previously published material is to be used, we recommend that the new author contact the author of the previous edition to determine the most appropriate way to proceed regarding authorship.
Regarding contributor agreements, we encourage chapter authors to request specific rights should their work be revised or incorporated into another edition without their participation as authors. For example, they may ask for a clause reserving the right to review the manuscript of any revision before publication, specifying the type of formal credit to be given if more than a certain number of words, figures, or tables are copied, or to be included as authors if more than a certain percentage of a future chapter is copied from their work. Despite owning the copyright, the publisher would not be able to rightfully ignore the original creator of the work during preparation of a future textbook if such clauses were included in a contributor agreement.
A textbook editor should be no less rigorous in preventing and detecting plagiarism before publication than is an Editor-in-Chief of a scientific journal. The editor of a textbook must make every reasonable effort to ensure that a chapter reflects the author’s own work and ideas and that portions previously published under separate authorship are properly identified and credited to the original author, regardless of whether or not the publisher holds the copyright. Editors are therefore advised to provide expected ethical standards in writing when engaging authors and require that authors declare that their submitted work is in compliance. We further propose that when engaging chapter authors, editors inform the authors that plagiarism-detection software will be applied to all submitted chapters. While software may not detect plagiarism when the original work has been paraphrased, translated to another language, or the plagiarism consists of stolen ideas, almost any detection software, including freeware, would detect the use of large blocks of verbatim text.19,20
As noted earlier, we recognize that for most textbooks, publishers have required authors and editors not only to assign their copyright to the publisher but to also give up any and all future rights to the work. In this situation, the publisher owning the copyright may republish any and all portions of an author’s work without obtaining the author’s permission.
Nevertheless, we maintain that publishers have an ethical responsibility to assure that the authors of a scholarly work receive proper attribution whenever the publisher republishes their scholarship, including its use in a subsequent textbook. Ultimately, prevention of the kind of plagiarism described in this essay is in the best interest of the industry and must be proactively supported. We urge publishers to develop clear and fair authorship and attribution policies (similar to those required when others hold the copyright) and incorporate them into contributor agreements for authors and editors. At a minimum, publishers should provide editors with appropriate plagiarism-detection software and require editors to attest that it has been applied to every contribution in the textbook. For software to be effective, publishers will need to share textbook content with plagiarism-detection databases (such as iThenticate), as is currently the case with scientific journal content.
Educational institutions should explicitly state that guidelines regarding misappropriation of written works apply equally to textbooks and book chapters, including serial edition revisions. We recommend that faculty members be informed of this in writing. Harvard Medical School’s guidelines provide an example of how policies on integrity in science may be specifically outlined for faculty members who are authors and editors of textbooks, including serial editions.8 Educational institutions are required by law to investigate allegations of misconduct, including plagiarism, if they involve federally funded scientific research. As noted previously, the definition of plagiarism used by the HHS Office of Research Integrity is not ambiguous and makes no exception for textbook chapters.
We urge authors, editors, and publishers to work together to ensure the integrity of all textbook content. Authors should attest that their manuscripts represent their own creative work and provide clear and appropriate attribution of material created by others. Educational institutions are also responsible for upholding their published guidelines with regard to plagiarism when it occurs in textbooks. Finally, before publication, the standard operating procedures for all textbooks should include efforts to prevent plagiarism and application of plagiarism-detection software to every contribution.
Textbooks count…It’s still plagiarism!
Dr. Lawrence J. Saidman is the Correspondence Editor for the Journal and the Editor-in-Chief for Anesthesia & Analgesia Case Reports. This manuscript was handled by Dr. Steven L. Shafer, Editor-in-Chief, and Dr. Saidman was not involved in any way with the editorial process or decision.
Name: Anonymous Author 1.
Attestation: Anonymous Author 1 has approved the final manuscript.
Name: Anonymous Author 2.
Attestation: Anonymous Author 2 has approved the final manuscript.
Name: Lawrence J. Saidman, MD.
Attestation: Lawrence J. Saidman has approved the final manuscript.