Double the materials science, double the fun?

An interesting fact about Matter (and our associated titles at Cell Press) is that we do not have a precise number of articles to publish per issue. Sure, we have targets, but those are set on an annual basis. Currently, it's around twelve research articles per issue for Matter, plus or minus one or two depending on the month. There's a perpetual balance between the number of articles, quality, and scope, as well as editorial handling capacity. Some of the metrics—i.e., desk rejection rates—we have explicit control over, but if we sent more out for review (for example), editorial handling times would increase, peer reviewer requests would increase, and even more invitations would be sent to your inbox (if you are curious, our desk rejection rate for 2023 was around 85%). In terms of total number of articles published, the elephant in the room is clearly the effect it has on journal impact factor (JIF)—the more we publish, the higher our denominator, and thus the lower the JIF. Clearly, if all the works were highly cited, this would not be an issue. However, it's difficult to predict, so one strategy is to limit the number of published articles and maintain an established bar for quality. This "bar" is not only specific for each journal, it is also highly dynamic. When we launched Matter in 2019, we didn't want to be driven by trends and impact factor, so our editorial ethos was "if it's interesting and scientifically sound, send it out for review." This was a great approach at first as we hit our publication targets (and also achieved a healthy initial JIF, for what it's worth). We started building a good reputation in the community, and our submission volume grew. This is a good thing, right? When you only get 50 or 60 papers submitted a month, you can send out the top 20% for review, no problem. When that increases to over 100, then something has to give—either you raise the "bar" for review and increase the rejection rate, or you publish more articles. This is actually a widespread problem in publishing, as the volume of papers published every year is increasing. I quick-searched using SciVal (a research analytics tool for the Scopus dataset; both products developed by Elsevier, the parent company of Cell Press) and, as a rough example, the number of materials science publications grew from approximately 100,000 per annum circa 2013 to approximately 200,000 per annum in 2022—a 2× growth in 9 years (about 9% per year). Entire dissertations could be written about the overall effects this "publication inflation" has had on academia, from the emergence of OA journals to the over-valuing of metrics such as JIF and the strain on the peer review system, to name a few. Here, we're just focusing on the volume. Consider this simple case: a publisher such as Cell Press was accepting 1,000 articles a year across multiple journals (we do more than that, but keeping the numbers simple), and that represented the top 10% of 10,000 total articles. After a few years, the total number of articles increases to 20,000. If Cell Press wants to keep the top 10% of papers, then they have to publish more (either more papers per journal or more journal launches)—OR, if they want to maintain consistent volume, the bar will rise to the top 5% of articles. What has occurred in the industry in the past decade or so is a subtle (but consistent) raising of the bar of "high-impact" titles, with an associated launch of new journals (as well as so-called "megajournals") to handle the bulk. The launch of Matter was relatively successful not because we were "stealing" papers from our competitor journals but because there was a surplus of high-quality papers that could not be accommodated by the current offerings. Back to our humble journal. Again, we established an editorial policy of "if it's interesting science, send it out for review," with no fixed limit of papers per month. Once we established our bar, I kept telling my team we are remarkably consistent in terms of rejection and acceptance rates. I check the numbers monthly, but it's NOT a matter of 10 submissions come in, so we select the "best" of the lot to send OTR ("out-to-review" in editor parlance). It's a more organic process. Some weeks, we may send zero submissions out for review. Some weeks, we may send out six or seven. It all depends on what comes in. As our submissions grew, we kept a steady bar, and on average, we started sending out more papers for review each month. This led to an uptick in accepted papers. Now, as a publisher, you always want to be a little ahead in terms of accepted papers and scheduled issues (otherwise, if you have a slow month, an issue could be lacking content). So, when accepted in June, for example, your paper would be scheduled for the August or September issue, a couple of months ahead. Sometimes we may also have to skip months if there is a special issue. There are some journals in the industry that schedule MANY months ahead … as I'm writing this (January 2024), I found multiple articles on Scopus currently published with 2025 dates! The future is here! So, how do you reconcile a robust pipeline with issue lag? I found a potential solution thinking about the periodical magazines I used to subscribe to when I was in high school, i.e., Rolling Stone, Popular Science, and Entertainment Weekly. A few times a year, each of these magazines would release a "Special Double Issue." Twice the content! Same price! For such journals, a double issue was produced for one of two reasons. First, a single issue would be produced to "cover" multiple issues—i.e., if it was a weekly periodical, one issue would be produced for a two-week span. The other reason was to enable extra space for content, i.e., year-in-review-type issues (or, for example, People magazine's annual "Sexiest Man Alive"). Our use here is akin to the second reason—a mass of great articles we want to release as soon as possible to reset our pipeline and eliminate the acceptance-to-issue "lag" in our publishing timeline. With that motivation, I welcome you to our first annual Special Double Issue of Matter! Same quality with twice the materials science! This issue contains twenty-six research articles ranging from graphene quantum dots to perovskites, artificial muscles, locust mandibles, and a rice-based meat substitute. We even invited two covers (for the full, uncut images, see Figure 1). Double the science must be a good thing, right? More information disseminated more quickly to the community? Again, you can look at it as a very small sampling of academic publishing as a whole (more papers, more journals, more citations, more!). Perhaps we are being a little tongue-in-cheek, reflecting on the commoditization of academic papers, more akin to Popular Science than the original Philosophical Transactions of the Royal Society circa 1665. Perhaps. What else could we "double" in scientific research? Would the result be beneficial or disastrous? What if we doubled the number of reviewers from three to six? Six referees scrutinizing your work with six sets of comments to address … ugh. Horrible. Not only would the chances of rejection increase, but consensus would also be more difficult to attain. Moreover, the community is already pushed to the upper limits in terms of review requests and commitments. The one advantage would be that manuscripts, in theory, would be greatly improved at the behest of a battery of reviewers. In theory. Perhaps, to alleviate the workload, we offer double the time to review? For the reviewers, this would be good, but I believe it merely delays the work (reviews technically take hours, not weeks of effort … it's not the workload that's the issue, it's the scheduling!). Plus, the review process is already quite long for many journals, taking many, many months at times. At Matter, we aim for 100 days from submission to acceptance (including review and revisions) and currently average 106.6 days (at the time of writing this article). Additional delays are a disservice to science and contrary to science dissemination. What if we doubled the target and accepted amount of citations per paper? A current Matter manuscript has around 50 citations total. What if we upped that to 100? Clearly, if the number of citations were doubled, the effect would increase impact factors across fields, as well as author H-indices. At first glance, this might be good, but if everyone underwent the same increase, standards and expectations would simply increase to match (fun exercise: double your citations for every paper—it won't double your H-index, but it would double a journal's impact factor). It would be citation metric inflation—the relative value per citation would decrease. Perhaps we shouldn't strive to double any of these values. From a simplistic perspective, like any dynamic system, it is inevitable that all processes reach an equilibrium state. While publishing is undergoing consistent annual growth (due to external factors), there is some internal balance between number of articles, reviewers, time, and citations. We can get away with a double issue once a year, but if every journal produced double the content, then that could perturb the entire system. Double would become the new norm, and there would be a desire to double again (much like the stock market, for better or for worse). It's a strange industry. Perhaps the only room for explicit doubles in science is "double double" coffee (Canadian thing, IYKYK). Resources, however, are ultimately limited, and unrestricted growth will eventually be detrimental. Science cannot publish 100 million papers per year (an extreme exaggeration). In the Marvel Cinematic Universe, Thanos recognized this resource-to-growth limitation. His solution was to halve the requirements of the universe. Perhaps science publishing could also be halved … Wait, I just had an idea for another special issue …