Introduction

There are many scientific and social benefits to interdisciplinary research. Such research, which spans disciplinary silos, is suitable for addressing broad questions and can potentially have impact on multiple fields. Furthermore, such research generates types of data that intentionally crosses disciplinary boundaries and can impact a larger swath of the research community. While there have been calls for more financial support for interdisciplinary research for many years1,2 funding agencies and universities have increasingly recognized these benefits in recent years and have encouraged the formation of interdisciplinary research teams. Indeed, some researchers3 claim that we are currently in a “golden age of social science” with interdisciplinary endeavors and opportunities, though they also indicate that there are hurdles that such work faces.

One hurdle is that the publication potential of interdisciplinary research may stifle the formation of interdisciplinary research teams. The co-authors that make up an interdisciplinary team typically have different home departments. These departments quite understandably place disproportionate weight on publications within their own disciplines. This matters greatly for tenure and career advancement and can create conflicts within teams when they decide which journals to pursue. Junior faculty or postdoctoral scholars may be dissuaded from pursuing interdisciplinary work until such time as tenure provides them job protection or achieving the rank of Full Professor is assured4. Junior researchers may avoid using data and methods from other disciplines with which senior department colleagues lack familiarity. Some high-potential research that could be conducted would not be undertaken, contributing to the trend of decreasing “disruptiveness” of science over time5. In this article, we propose and describe a simple solution to this problem: the cross-listing of journal publications in multiple journals who share the review process.

The problem

The presence of elite, interdisciplinary, general-interest journals that span multiple fields of science (such as Nature, Science, the Proceedings of the National Academic of Science, etc.) helps the right tail of very high-impact interdisciplinary research find a home that is recognized across disciplines. The interdisciplinary nature of these elite journals also means that they are organized to identify peer-reviewers who can span the research areas, data types, and methodology used within an article stemming from an interdisciplinary research effort. However, only a small fraction of interdisciplinary studies will ever find itself in those journals given their low acceptance rates and exacting standards. We are concerned here with the bulk of interdisciplinary research conducted, which makes a valuable contribution to science but is unlikely to achieve publication in elite outlets. For such articles, interdisciplinary teams must choose the discipline whose journals they target.

Targeting single discipline journals with interdisciplinary work leads to three potential hazards. The first is that a single-discipline journal will typically engage a set of peer-reviewers who are experts in the field of the target journal. Part of the reason for this is that researchers decline to referee for journals outside their discipline because they may consider writing a report for a journal that they would not be rewarded for publishing in to be a poor use of their time. For example, consider a research team made up of neuroscientists and economists who have a paper that explores a research question that bridges neuroscience and economics using both sophisticated fMRI data and behavioral economic decision-making data. If such a paper is sent to an economics journal, it would naturally most likely be peer-reviewed by a group of researchers who are principally evaluating the research from the perspective of economics. If the same paper were sent to a top neuroscience journal instead, it would most likely be peer-reviewed through the lens of neuroscience. While this is what one should properly expect from each journal’s perspective given its readership, it can distort the focus and emphasis of the paper away from the authors’ original intent and may leave half of the science insufficiently peer-reviewed. To be evaluated on its own terms, the content of interdisciplinary research requires peer-review from each of the disciplines represented in the scientific process. In general, neuroscience reviewers are not specialized in evaluating the nuances of the economics side of the science nor the fine details of the economic data just as the economics reviewers are not specialized in evaluating the neuroscience data and the nuances of the science on the neuroscience side. Upon publication in a single discipline journal, the research will likely be disproportionately visible to and cited by researchers from that discipline and its contribution to the other discipline may be lost.

The second problem concerns career incentives6. Sending research to single discipline journals means publishing outside of journals that are recognized, valued, and rewarded by the home departments of any co-authors outside of the journal’s field. This makes it difficult for those co-authors to clearly provide evidence of their contributions to their own field and to be appropriately rewarded for those contributions. Junior faculty at most universities are expected to develop a reputation within their own fields and even high-quality publications outside of their home fields do not help to demonstrate this reputation formation. Under the current academic system, this imperative clashes with the objective of advancing interdisciplinary research. Indeed, the Berkes study indicated that research survival rates of newly minted biomedical PhD students were dramatically shorter for heavily interdisciplinary researchers compared to modestly interdisciplinary researchers.

Third, if the eventual publication outlet of most interdisciplinary work is a single discipline journal, the co-authors from outside of the “primary” field may not be fully invested in the scientific process (design, data generation process, analysis, and communication of results), and instead simply focus upon the narrow slice of the project most related to their own field (a strategy that requires less of a research time investment). The opportunity cost for these non-primary-field co-authors researchers is work on other projects within their own fields and is thus relatively high. They may prefer to be involved in large groups rather than small groups which would naturally require deeper involvement by each co-author. This mentality may be partially responsible for the phenomenon of large interdisciplinary team sizes. Science as a whole, however, may suffer from this phenomenon as studies by small teams are suspected to generate more disruptive science and novel ideas that delve more deeply into topics while studies by large teams are thought to more fully develop pre-existing ideas7. Wu et al. explore this idea using a large data set of over 40 million published articles. They find that the citations for small teams’ work tend to be delayed while large teams receive their citations relatively quickly. These problems intensify the lack of incentive for junior researchers to engage in small interdisciplinary teams as the tenure clock disproportionately favors rapid citation accumulation.

There have been some attempts to address some of the problems we have outlined above. For instance, services that evaluate/rank journals, like Scopus, provide rankings of journals that have interdisciplinary appeal in multiple different field categories. This can enable Promotion and Tenure committees to glean the level at which journals sit with respect to different disciplines. Another example is Clarviate’s Journal Citation Reports relatively recent addition of their “Journal Citation Indicator” measure that incorporates a category level impact factor normalization that helps address citation norm differences between disciplines. This can also assist Promotion and Tenure committees in better evaluating publications outside their own fields. Editorial management systems that can suggest reviewers from a wide range of fields can also help improve the scientific evaluation of interdisciplinary papers that are submitted to single-field journals. These solutions, while useful, cannot solve the problems alone. For instance, it is impractical to evaluate every journal separately in every field, evaluation metrics cannot help journals in the peer-review of papers outside of their main field, etc. The following sections outline another solution that has the potential to broadly address the problems outlined above.

The cross-listing solution

A common practice at universities when advertising their courses to students, the cross-listing of courses across fields of study, can be adapted by journals and their publishers to help solve the problem. Courses that overlap in content from multiple disciplines are often cross-listed so that students in different majors can take the course simultaneously, while each student receives credit for taking a course within the department of their own major. Why not apply a similar concept to the publication process? Since the publishers of most scientific journals already host websites, receive submissions, and process payments for journals in many fields, they could also facilitate “cross-listed” journal publications from interdisciplinary research teams. The process could be relatively simple, as illustrated in Fig. 1:

  • Interdisciplinary teams could submit their papers jointly to two journals that are in different fields but that have the same parent publisher. In most cases this would be possible given the landscape of the academic publishing industry, which has several large players whose journals span many disciplines.

  • An editor at each journal would then be responsible for one half of the peer-review process.

  • Each editor would send the paper out to external referees within their discipline or could desk reject the paper. Authors would have the option to commit at the time of submission to withdrawal from both journals in response to a desk rejection from one of the journals. If one editor desk rejects, the editor of the other journal is notified that a desk rejection occurred. The paper may remain under review at the other journal at the discretion of that editor and the decision of the authors at the time of submission. We expect, however, that in most cases, editors will wait to see if the other editor desk rejects before sending the paper to reviewers.

  • After the two independent peer-review processes are completed, if both editors agree that the paper is publishable, the paper could be published as is conventionally done, but posted online under one DOI to which both journals separately link. All acceptances are conditional until a final decision is made at both journals. Revisions could go back to the referees at both journals at the discretion of the editors.

  • The data sets for the published paper could be published at easily accessible non-journal centric data repositories or published separately at journals like Nature’s Scientific Data that are available to researchers across disciplines.

  • If only one journal recommends that the paper be published (for instance if a paper is rejected or withdrawn from one of the journals) the handling editor at the remaining journal could be free to invite the authors to proceed with the paper as a non-cross-listed paper. That process could involve seeking further external reviews if the editor desires.

Fig. 1
figure 1

Flow of Activity Under Proposed Process (Abbreviations: J1 (Journal 1), J2 (Journal 2), DR (Desk Reject), Rev (Journal Decides to Review Paper), Refs (Referees)).

Full size image

Since it is now common for many journals to not have print editions at all (or have at least some online only papers), these cross-listed publications would not be appearing in multiple separate printed materials and thus not be confused as being two separate publications. In the event that both journals have print editions, cross-listed articles would appear in at most one print edition.

Benefits of cross-listing

This proposed solution addresses many concerns for the parties involved in the academic research and publication process. Some of the main advantages for the various stakeholders are:

For authors:

  • It connects authors of interdisciplinary work to the multiple audiences that the work is intended for.

  • Including multiple disciplinary journal homes in the evaluation process provides academic departments a clear way to evaluate publications for promotion and tenure purposes.

  • Cross-listing will help build a common language (a “lingua franca”3) between fields as researchers in different fields will increasingly be reading the same papers and exposed to the same scientific “jargon”.

  • Cross-listing will help expose disciplines to the data types and analysis strategies used effectively in different fields.

For editors:

  • Linking to multiple journals would be expected to increase the overall visibility, reach, and citations of research papers as the readership of each of the linked journals is likely to be very different.

  • Journals with interdisciplinary focus would be expected to get more citations and therefore have higher impact factors.

  • Editors can confine their attention to recruiting reviewers and interpreting referee reports within their own discipline. This allows the editors to concentrate their attention on areas that they know well and in which they can make more informed decisions.

For publishers:

  • Both journals get full credit for a citation to a cross-listed article and a boost to their respective impact factors.

  • For the article itself, a single DOI and a joint citation that lists both journals in a “Journal A/Journal B” style would help prevent double counting and would facilitate easy citation and tracking of its downloads.

  • The process does not require the creation of new journals or hiring of specialized interdisciplinary staff. It would require at most only relatively minor changes in journal infrastructure.

For the scientific community:

  • The dual review process strengthens the rigor of the scientific evaluation process through peer review in multiple fields rather than one.

  • The dual review process maintains the ability for the individual fields (via their journals) to independently vet research.

  • A cross-listed publication provides a simple and transparent signal to the research community that the paper is a quality contribution in each of the research fields. A standard single-discipline journal publication would continue to provide a signal to the research community that the paper is a quality contribution in the research field of that journal.

  • Cross-listed publications create an incentive to generate data that is useful across multiple disciplines.

  • Cross-linked publications facilitate the spread of common information and data between fields and help tear down the silo walls that discourage discussions between fields.

  • The increased willingness of researchers to conduct interdisciplinary research may help reverse the trend of decreasing “disruptiveness” of research over time5,8 and may increase the prevalence of small interdisciplinary research teams7.

Overall, the proposed solution of cross-listing journal publications creates a more incentive compatible process for authors, departments, journals, and publishers to evaluate and spread interdisciplinary research. Additionally, this solution may help increase research and data quality, improve the review process, and facilitate the increased participation of junior (non-tenured) researchers in interdisciplinary research endeavors in a relatively low-cost manner.

Challenges and possible solutions

The previous sections have outlined a new system to promote and facilitate interdisciplinary research through a system of cross-listed journal publications and discussed many of the potential benefits of the system for the research and academic community. While there are many benefits foreseen in this new process, any change in the current system of academic publishing is likely to run into challenges as well. We next outline some of these potential challenges and discuss how to move forward with those challenges in mind.

The most obvious source of challenges that may occur to people with previous experience in journal article publication is how a system like we propose would be operationalized within the editorial management systems that publishers use. The programs used for processing submissions, such as Editorial Manager or Scholar One, would have to be modified to allow joint submission. As we do not represent any specific publisher, we cannot speak to the fine details of those systems. However, our experience leads us to believe that the changes we propose are feasible since forwarding from one journal to another already is enabled with these software packages and workflows can be readily adjusted in these programs. The most important elements of the review process are already built into these systems (e.g. sharing of reviews, assignment of editors, identification and registration of reviewers, acceptance of payments, etc.) and would not be expected to need major changes. Large publishers have databases of reviewers that span many fields. Payments of submission fees and article processing charges would need to be addressed, but if cross-listing is occurring with the same publisher this should also be a minor problem. To avoid increasing submission and processing fees on the researcher side, we would simply recommend that the more expensive fee between the two journals be used by the publisher. This ensures that the publisher continues to receive no less payment for an article compared to publishing the paper as a single-journal article in the more expensive of the two journals. Authors would likewise not be paying any form of premium nor more than they would have if publishing an article in that same journal.

On the peer-review side, there may be concerns that this system may just intensify the strain placed upon peer-reviewers who are already largely unpaid and under-rewarded for the peer-reviews they currently do for journals. It is unclear what the new system will mean in terms of the number of submissions to different journals of a paper before eventual publication. If interdisciplinary teams currently struggle to find outlets for their papers, the new system may reduce the total number of submissions needed prior to publication. If the new system leaves that overall number of submissions unchanged then there may be an increase in the overall number of reviewers performing reviews. As such the total number of reviews and thus the amount of reviewer time may either increase or decrease. However, we see the proposed system improving the peer-review process and generating a better content/science-to-reviewer match and thus reducing the number of actual reviews.

Editors and reviewers tasked with evaluating interdisciplinary papers currently must judge work that is both within and outside of their own field. This can be difficult and stressful. Editorial systems already typically enable an editor to identify reviewers outside of their own field, but they typically do not have the expertise across many fields to have first-hand knowledge of who the appropriate reviewers are. This may result in mismatches and an inefficient reviewer identification process. It may also be unclear to a reviewer whether an editor wants them to focus on the entire content or only the content related to their particular field. Our proposed system could improve the situation by playing to the strengths of each reviewer and editor, since they are emboldened to concentrate their reviews on the content in their own field while they have the confidence that a different set of reviewers is evaluating the material from the other field. We would expect our proposed system to result in a decrease in the number of declined invitations since fewer reviewers would say that a paper is outside their area of expertise. This effect may reduce overall decision time.

There may be a related concern that review time may be increased due to a “weakest-link” problem. A paper submitted into a cross-listed review process must wait for reviewers to complete the review process from both journals before they can receive a revise and resubmit or an acceptance. That also means that the authors may wait a longer amount of time to discover that their paper was rejected. We see this as akin to the authors having submitted to the slower journal first. Authors already have a good sense of how long papers take at different journals (many journals advertise these turn around times) and it is reasonable to assume they could anticipate ahead of time which journal would be the slower of the two journals. Authors could then avoid journals they deem “too slow” to be worth the potential benefits of a cross-listed paper. Earlier, we discussed journals and authors pre-committing to what would occur upon the earliest desk-rejection from one of the journals, which should also help prevent extensive time spent in a futile review process.

We expect there will be a learning curve with the proposed system as it will likely take time for academics to adjust to seeing these types of cross-listed publications. It took time for the academic community to adjust to incorporating DOIs on CVs, but such inclusions are now generally common and unsurprising. Some article writers may struggle with citations in the “Journal A/Journal B” style we propose, but if the journals’ websites for the paper itself have the suggested citation we think this problem will be small. Adjustment will be helped by the papers being online only and thus these papers need not have the traditional “Volume/Issue/Page” citation elements of print articles.

We expect a similar learning curve for those evaluating academic researchers CVs for promotion and tenure or on hiring committees. There are already problems with these groups evaluating research from outside evaluators’ own fields, and there is no standard way such evaluations are done across disciplines nor across universities. We anticipate that this new system will help alleviate confusion more than it adds to confusion. Evaluators will be able to more readily ascertain that a contribution is interdisciplinary. They can more easily identify the quality of a study as they will be more likely to recognize one journal compared to only seeing the name of a journal outside their field. We also anticipate that article tracking systems (such as Google Scholar) and journal evaluation services (like Clarviate’s Journal Citation Reports) would adapt after a short period of time as well. Competition between such services may help push forward new ways to identify and aid evaluation of such papers as well. These would be expected to increase the value of cross-listed publications to the academic community in general.

Our proposal focuses upon cross-listed publications occurring between journals belonging to the same parent publisher (e.g. “intra-publisher” cross-listing). Some readers, though, might surmise that this type of cross-listing favors large publishing companies who publish across many fields and disadvantage society journals who self-publish (such as the American Economic Association in economics). While it is easiest for publishers to implement cross-listing within their own journal portfolio, there is no reason in principle why an independent journal could not enter into an agreement with a larger publisher to implement a cross-listing scheme. Similarly, it may be possible to have cross-listed publications across multiple large publishers as well. While we think such “inter-publisher” publications are feasible, we do see that there are more hurdles for these types of publications, as they would not share editorial systems, payment systems, etc. In principle at least, it is possible to find solutions to these types of hurdles. For instance, if two publishers wish to maintain DOIs linking directly to their own system, then both publishers could forgo the single DOI suggestion we propose and have a different DOI URL for each journal, with the separate URLs re-directing to a single base file and website. There would be negotiations for such inter-publisher cross-listed publications, but publishers are not unfamiliar with negotiations on other procedures. Such negotiations already must occur when journals switch publishers, a relatively common occurrence for the journals run by academic societies.

At first glance, one might anticipate that there may be some way to “game” or “cheat” the new system. There are a few strategic behaviors one may be worried about. What if authors submit via a cross-listing system with the intent of submitting to two journals at once and plan to withdraw their paper from one of the two journals as soon as they get a favorable report from the other journal? This type of scheming, we think, is unlikely to become a common occurrence. If journals have adopted the “conditional acceptance” only model we have proposed during the cross-listed review process and submissions are clearly joint, then there would be no guarantee the paper would be acceptable as a stand-alone paper by the journal making the first decision. In addition to the reputational damage to the author, there would be a general negative signal about the paper’s quality to the handling editor. If this problem was unexpectedly frequent, however, it could be reduced by decisions from the two journals being revealed simultaneously.

What if relatively low-ranking journals seek out interdisciplinary papers to boost their metrics (if both journals get full credit for citations)? While it is reasonable that lower-ranking journals might want to be paired with higher-ranking journals, we think the opportunity for the editor at either journal to desk-reject a paper helps eliminate this concern. Higher impact/elite journals would only want to cross-list with a lower-ranking journal if the paper was up to their own high standards and likely to generate an impact they are happy with. The fact that the paper was submitted to a lower-ranked journal in another field might be valuable negative information to an editor in judging its interdisciplinary quality. Anticipation of this stigma is a natural disincentive to authors and deters researchers from engaging in the strategy of submitting to one journal of high quality and one of low quality and thus also helps avoid the problem of low-ranking journals artificially boosting themselves by associating themselves with higher-ranking journals in other disciplines. If published, promotion committees would still have freedom to use either journal’s impact factor in their performance review (min or max) or a composite measure, such as the average of the impact factors of the two journals.

We do see a real possibility that some journals may heavily embrace this system and disproportionately target or seek interdisciplinary papers, and some of this behavior may be motivated by seeking more citations. We do not see this as a fundamental problem and thus are not inclined to encourage any form of divided or fractional impact calculation for journals publishing a cross-listed paper. If both journals participate in identifying a quality study and bear the cost of the peer-review, we advocate for them receiving credit for all impact coming from the paper. Some of the major impetus of the system is to help foster and reward interdisciplinary studies so that they will be better read, more scientifically evaluated, and their quality more easily discerned. This would help the contributions of interdisciplinary researchers to be measured and recognized. If some journals see value in interdisciplinary work and seek to encourage its submission, we would see it as a success of the new system. We would not, however, expect to see a general shift where all journals seek out such papers. Many journals in every discipline are focused upon discipline-specific topics and are not natural outlets for interdisciplinary work. Those journals would be unlikely to proceed with many cross-listed submissions unless the fit is particularly strong. We would expect that some fields and sub-fields and some journals would have relatively greater natural complementarities with other fields and thus would have more cross-listed publications. This would boost their impact, commensurate with the high value contribution of the cross-listed papers they publish. These journals will be naturally motivated to ensure that they are publishing high quality cross-listed papers, or they risk their reputation and metrics being harmed by low quality papers.

Conclusion

We proposed here a system of cross-listed journal publication to help facilitate and incentivize the conduct and publication of interdisciplinary research projects. Not only do we expect that such a system would facilitate the publication of such studies, but we also expect that the publication of interdisciplinary studies in multiple fields simultaneously would help disseminate the data, knowledge, and language from such science more rapidly and effectively across scientific disciplines. We see value and benefits that span across researchers, to editors, to journals, and to the scientific community as a whole. While there are many expected benefits, we do anticipate there will be some challenges such as those we described in the last section. However, we believe that these challenges can be overcome. We encourage the scientific community to see our proposal as an early blueprint of a modification of our current publication system, and we encourage a hearty discussion to improve upon our initial vision. Considering the large potential gains associated with cross-listed publication, we encourage academics to not let worries about inter-publisher challenges prevent changes at the intra-publisher level where there are many fewer challenges and the benefits are already large.