Analysis Christmas 2024: Death is Just Around the Corner
Ethics of posthumous scholarly authorship in the sciences
David Nunan, Jeffrey K Aronson
BMJ 2024; 387 doi: https://doi.org/10.1136/bmj-2024-080830 (Published 16 December 2024)Cite this as: BMJ 2024;387:e080830
David Nunan and Jeff Aronson suggest that formal criteria are needed to determine eligibility for posthumous authorship and for dealing with associated ethical problems
Posthumous authorship poses problems in multiple ways in different disciplines. Max Brod, Franz Kafka’s literary executor, published The Trial, The Castle, and Amerika after Kafka’s death, despite Kafka’s instructions to burn his unpublished works. These works have enriched the literary corpus and have influenced many writers since, but the ethical problem is of concern.
Other notable literary cases include the posthumous publication of letters of Jane Austen, causing distress to members of her family, and Ted Hughes’s publication of Sylvia Plath’s poems in a disputed edition amid controversy about her suicide. In contrast, posthumous publication of J R R Tolkien’s papers by his son Christopher seems to have raised no difficulties.
However, despite interest in posthumous literary publication, little has been written about it in scholarly science. Calls for a clear policy1 have not resulted in uniform comprehensive guidelines, and occasional online discussions have shown that there is no consensus about how deceased colleagues should be credited in publications: some advocate coauthorship; others acknowledgment only.23 Nor are there clear guidelines on how an author’s death should be noted.
The International Committee of Medical Journal Editors (ICMJE) does not mention posthumous inclusion in its guidance on criteria for authorship of published papers.45 Nor do guidelines from the Committee on Publication Ethics (COPE) specifically mention posthumous authorship,6 although the committee has occasionally commented on specific cases7 and also reported a case highlighting the possibility of introducing errors that would otherwise have been corrected by the deceased author.8
Other important documents on research integrity have been published by the UK Research Integrity Office,9Universities UK,10 and a House of Commons Science and Technology Committee11; none, however, mentions posthumous authorship. Nor do various documents by publishers and editors dealing with publishing ethics.121314
Two reviews have examined advice given by learned societies (the American Meteorological Society, the American Ornithological Society, the American Physical Society, and Cochrane); journal publishers (BMJ Journals, Dove Press, Elsevier Science, Wiley, Nature Publishing, Oxford University Press, Springer Nature, Taylor and Francis, and Walter De Gruyter); and 16 individual journals.515 Several of those do not specifically address the question of deceased authors, and among those that do there are large variations in what they specify (box 1).5 Furthermore, there are conflicting views about whether coauthorship or an acknowledgment should be preferred. However, the guidelines discussed in these reviews have not dealt with other ethical concerns arising from posthumous authorship.
Box 1
Advice on posthumous authorship from 23 societies, publishers, and journals*
- Authorship according to criteria (indirectly implying, but not specifying, ICMJE): n=9
- Obelus (“death dagger”) required to mark a deceased author: n=4
- Requirement to mention the deceased in a footnote or contributor statement: n=13
- Corresponding author responsible: n=3
- Family consent required: n=5
- New contact to be appointed if required: n=1
- Editor in chief to arbitrate: n=1
- Deceased individuals should not be authors: n=1
- Acknowledgement an alternative to co-authorship: n=1
- *Analysis by Helgesson et al5
Extent of the problem
The exact number of publications that feature posthumous coauthors is not known; we estimate it to be at least 10 000 in the biomedical field alone. Jung and colleagues surveyed 2 601 457 peer reviewed biomedical publications during 1990-2020 and found 1439 deceased authors credited with 5477 posthumous publications; they found a 146-fold increase in the number of such authors since the year 2000.16 However, they underestimated the extent of the problem.17
First, they searched only in acknowledgment sections for declarations that coauthors who had died had been mentioned. But posthumous coauthors are often not mentioned in acknowledgments. For example, Jung and colleagues identified our late colleague Douglas G Altman as having 36 posthumous publications, but in fact he had 79 by the end of 2020, and we have now identified 103 (data available on request). Altman’s death is not always recorded in his posthumous publications.
Jung and colleagues also missed several posthumous biomedical coauthors altogether. These include David L Sackett (three posthumous publications listed in PubMed), Lisa M Schwartz (eight before the end of 2020), and Stanley J Korsmeyer (26). By restricting their search to biomedical authors listed in the Europe PubMed Central database, Jung and colleagues missed publications and scientists listed in other databases. For example, they did not mention the mathematician Paul Erdös, who died on 20 September 1996, leaving many problems in number theory unsolved and projects unfinished; he is credited with at least 73 posthumous items up to 2015 (https://drive.google.com/file/d/1kfOT-EI8DuKNwEbOQsxGgUrETuLHkYH_/view).18 Nor did they include the egregious case of the 1913 Nobel prize winning chemist Alfred Werner, who died in 1919 and was named as a coauthor of a paper published in 2001.19 We know of no extensive survey of non-biomedical fields.
Defining a posthumous publication
Although what constitutes authorship in scientific publications has been widely discussed, we have not found a satisfactory definition of a posthumous publication. It is not simply a publication that appears after a coauthor’s death, as our proposed definition shows (box 2).
Box 2
Proposed definition of a posthumous publication
A publication, whether in print or online, for which at least one author has died before the accredited date of publication of the version of record, the deceased author(s) being named in either the main list of authors or the list of members of a contributing group; this includes translations and updates of previously published lifetime or posthumous publications but excludes verbatim reprints and print versions of publications that were published online during the author’s lifetime.RETURN TO TEXT
Occasionally an author may be the sole author on a posthumous publication, generally when papers published during the author’s lifetime are republished as verbatim reprints—for example, in collections of their work—but occasionally after online publication. The definition takes note of such cases.
Ethical concerns and solutions
The few publications in which posthumous authorship has been discussed have mainly concentrated on the question of whether someone who has died should be credited with authorship or merely acknowledged for their contribution, principally highlighting the impossibility of obtaining posthumous consent to either authorship or acknowledgment.2 However, there are several concerns beside the question of coauthorship versus acknowledgment, which have not been widely aired (table 1).
Table 1
Ethical and practical problems raised by posthumous authorship, with proposed solutions
These observations raise questions about posthumous inclusion of authors’ names in publications. For example, posthumous authors may have contributed to the concept or design of a study (ICMJE criterion 1) but no more than that, or they may have taken part in drafting or revising the manuscript (criterion 2), or occasionally may have approved the final manuscript (criterion 3). The last ICMJE criterion clearly cannot be met (“agreement to be accountable for all aspects of the work”). Helgesson and colleagues5 made some suggestions for dealing with ICJME authorship criteria, and we build on these to propose a classification that authors could use to decide on posthumous authorship depending on the number of ICMJE criteria that the author met (box 3).
Box 3
Suggested classification for posthumous authorship
- Class 0: None of the ICMJE criteria is met. This does not merit authorship and perhaps not even acknowledgement
- Class 1: Only the first ICMJE criterion is met. This does not merit authorship, merely acknowledgement
- Class 2: The first two ICMJE criteria are met. This merits authorship
- Class 3: The first three ICMJE criteria are met. This merits authorship
We suggest that a footnoted obelus (†) should be appended to the deceased author’s name in the list of authors. We recognise that some journals prefer not to do this, but the death of an author, including the date of death, should at least be acknowledged in the notes about contributors. Contributions to such groups as data monitoring committees by a deceased individual should be acknowledged, where relevant.
Deceased authors of online versions of publications published before their deaths, should be credited with authorship when the publication appears in print posthumously. However, their death should be noted (eg, “since this paper was published online, X has died [date of death]”). In the case of a sole author, the publishers should seek a colleague of the deceased author to act as corresponding author and guarantor and to take responsibility for approving the print version.
Our proposals facilitate authorship, in the spirit of the ICMJE criteria, rather than denying it. There should generally be an assumption that the author would have agreed to be named in the list of authors. The extent of the author’s contribution should provide some validation of this assumption.
In the case of acknowledgments, it will not generally be known whether the individual would have agreed to the acknowledgment. Confidentiality is unlikely to be a problem when acknowledging the contribution of an individual who has died, but coauthors should be confident that the individual would not have objected to the acknowledgment. In such cases, discussions with members of the deceased author’s family or other associates may be helpful.
International consensus required
Posthumous coauthorship is not uncommon and is increasing in frequency worldwide. This suggests the need for internationally agreed uniform methods for addressing a choice between posthumous authorship or acknowledgment, for signalling that an author has died, and for dealing with the attendant ethical problems.
Our view on the question of authorship versus acknowledgment is determined by the extent to which the dead person fulfils the ICMJE criteria for authorship (box 3). Setting a time limit, of say 1 year, between the time of death and the first submission date of a publication would be a practical way of facilitating decisions. Both ICMJE and COPE should update their guidance to include posthumous authorship along the lines we suggest, and the approach would ideally be agreed by all publishing societies and journals, to ensure uniformity.
Key messages
- No comprehensive guidelines exist for posthumous authorship in scholarly scientific publications
- Death of a contributor raises several ethical questions, and some advocate posthumous acknowledgment rather than coauthorship
- Existing authorship criteria can be used to determine whether a deceased person should be included as an author or only acknowledged
- Requirements for noting an author’s death in publications should be standardised and included in established publishing guidelines