Revise and resubmit

A couple of years ago, I wrote a post about an article of mine that had just come out, and its very long road to publication. At the end of it I blithely wrote that “The next article I publish should be out rather quicker—though the gestation time has been almost as long.”

Detail of front cover of Humanity, volume 10, number 1 (2019) 1This was both correct and incorrect. The next article I published did come out much more quickly, but it wasn’t the one I was talking about. ‘Humans and animals in a refugee camp: Baquba, Iraq, 1918-1920’ was submitted to the Journal of Refugee Studies in November 2017, went through peer review and revision that winter, and was published online in May 2018 (and in print this week). But the article I was actually referring to, on the history of humanitarian evacuations, didn’t come out until the end of April 2019. So I was wrong about it being published more quickly, but I was right about the long gestation time: I’d started drafting it, in early April 2013, based on research done over the previous couple of years, and I first gave versions of it as a seminar paper in November 2013 and January 2014.

This isn’t an unusually slow turnaround time for publishing an academic journal article in history. One thing that added to the delay is simply that the journal it came out in has a crowded publication pipeline: by the time I got my final acceptance in early summer 2017, the contents for all of the issues through to the end of 2018 had already been set. But what I want to write about here is peer review, another routine part of the academic publication process.

Peer review often gets a bad press: just google ‘Reviewer 2’ and you’ll see what I mean. (Perhaps you already belong to the public Facebook group ‘Reviewer 2 must be stopped!’, which at time of writing has 19,602 members.) No-one likes a rejection; no-one greets a critical review with glad eye and open heart; and I doubt that many people relish revising an article that they’d hoped was off their desk forever. I certainly didn’t. Between 2013 and 2017 this article was drafted, submitted, rejected, redrafted, submitted to a different journal, reviewed, accepted ‘with revisions’, revised, edited sharply down, and ‘finally’ accepted. That final acceptance made me very happy, not just because it meant the article was going to be published, but also because by then I was—to say the least—ready to see the back of it.

But I’m also absolutely certain that the published version is much better than the version of it that I first submitted. Even in a humanities discipline like history, which tends to presume (/idealize) single-authored work, getting a piece of work to publication is a much more collaborative process than the single name beneath the title implies. Often, we read other people’s work as if they’d simply sat down one morning and dashed off a print-ready piece to a response of “Grt thx will publish!” from the editors of a leading journal—while we bash our heads against multiple drafts, and nervously await reviewer 2’s stiletto. But any piece of work that’s been published in a reputable journal should have gone through a peer review process, and been improved by it. Plenty of good or excellent published work has been rejected at some point along the way. So this post tries to demystify the process, and show how much effort by other people is being alluded to when a footnote thanks “two anonymous reviewers for their thoughtful critical comments”. It’s aimed especially at people who are writing their doctoral dissertations and getting feedback on draft work. As supervisors, I’m not sure how well we explain the fact that our own work gets the red-ink treatment from peer reviewers all the time, and it’s not simply a hazing ritual: it makes our work better. There are plenty of blog posts out there discussing what happens when peer review goes wrong, as it often does. This one is about what happens when it goes right.

The first draft

Algae and seaweed trailing from the bottom of a small boat recently lifted out of the water.
Approximate length of main text : length of footnotes ratio (first draft)

In its first draft, my article told a story: how French forces, as they withdrew from Cilicia in what is now Turkey in November-December 1921, grudgingly evacuated many thousands of Armenian refugees. I’d done tons of archival research, and the story spread across the eastern Mediterranean and beyond, so writing the article took a while. I also needed to keep in mind three overlapping but distinct audiences: historians of the Middle East; historians of refugees and displacement; and historians of humanitarian action. This meant that the text grew footnotes the way a ship’s hull grows barnacles and seaweed the longer it spends in the water: in some cases, bibliographical mini-essays summarizing the literature on, say, the history of Cilicia up to the first world war for the benefit of readers with no prior knowledge of the region, or the history of first world war–era population displacements across Eurasia for historians of the Middle East. (The footnotes were also for my benefit, of course, like a pair of stabilizer wheels on an unconfident child’s bike.) Still, I was pretty happy with it, and friends and colleagues working in several different areas read the draft and offered broadly positive feedback and helpful suggestions. So eventually I felt I’d told the story well enough, and in the early spring of 2016 I sent it off. I thought I might as well try for a biggie: the American Historical Review, perhaps the leading journal in the discipline. I mean—it was an interesting story, right?

The rejection (well: revise and resubmit)

As the most prominent English-language journal in the discipline, the AHR has a larger full-time editorial staff than most. This means they have a somewhat different review process than most other journals I’ve encountered (in history, area studies, and refugee studies). Before the editor sends submissions out for peer review, they’re read by a member of the editorial staff, who writes a report recommending whether they should be sent out or not. Given the volume of submissions, this reduces the burden on (unpaid) peer reviewers, who don’t have to spend time reading and commenting on articles that clearly aren’t suitable, or aren’t good enough. The articles that are sent to peer review therefore get through it more quickly, while people whose submissions are not sent for review find out promptly rather than having to wait months for a rejection.

I was in the second group. No more than a couple of weeks after my submission was acknowledged, an email from the AHR editor arrived, with two attachments. One was a letter informing me, quite kindly, that my article would not be sent to peer review. The other was the reader’s report explaining why not. (With the permission of the editor and the anonymous reader, you can download it here [PDF].)

The letter was a ‘revise and resubmit’ rather than an outright rejection: my article wouldn’t be sent out for peer review at this stage, but a revised version would be considered. The editor suggested what the main revisions should be, and said that “the fix required is rather simple… and can probably be done simply by moving a few paragraphs around”. The report, which ran to nearly a thousand words, critically and incisively summarized my article, outlining why it wasn’t ready for peer review at the AHR yet. Publishable in a specialist journal, maybe, but “the AHR’s readership expects the kind of conceptual or analytical innovation that can transfer across fields of historical inquiry”.

Specialist readers might be intrigued in the regional history or the humanitarian history, but the vast majority of the AHR’s distinctive comprehensive readership spanning all fields of historical inquiry would most likely have felt alienated and deterred from reading very far into the piece.

If the letter was the good cop, the report was the bad cop.

A ‘revise and resubmit’ is always discouraging, and on reflection I was more discouraged by this one than I should have been. But I can understand why. When this email arrived, I hadn’t published anything other than book reviews since my own book came out four years earlier: a worrying gap. A few months earlier I’d submitted another hefty article to a different big-name journal, which would eventually publish it—but I didn’t know that yet, as I was still waiting for their reviewers’ response. These were also my first submissions to general history journals rather than more specialist Middle East studies journals. So I was, frankly, doubting my ability to write publishable work in history. (I’m now in a history department, not a Middle East studies department: there were institutional as well as intellectual reasons why I wanted to publish in non-specialist journals.) It was also the back end of a busy spring semester, and I was feeling as fraught and teary as usual at that stage of the teaching year. So I set the whole thing aside until the Easter break, by which time a positive response for the other article had allayed my catastrophizing. But it had also requested some minor revisions, which—since they would definitely lead to a publication—took priority as the spring turned to summer.

The revised draft

It wasn’t until later that summer that I returned to the article about humanitarian evacuations, and read through the AHR reader’s report again. Bad cop? More like correct cop. The gist of it was that this draft told the story of the Cilician evacuation at length, but I’d only really explained why I was telling the story at the end, and hadn’t offered nearly enough to convince a non-specialist audience of historians why this story should matter to them. Rereading the draft, I could see how right they were. I’d stayed too close to the archival material, giving a forensic but descriptive reconstruction of the historical events without analyzing them or making an argument for their larger significance. The report challenged me to do better:

[C]an the author front-load in the submission a decisive argument that bespeaks an original and innovative contribution to the historiography on humanitarianism, beyond simply the addition of the factor of evacuations earlier than expected? […C]an the author draw out more from the subject and the case study?

The first draft had said I wanted to use this story to establish humanitarian evacuations as an object of historical enquiry. Great. But it didn’t say why, or do much more than present an early example. Reframing it for a revised draft meant introducing the policy and practice of humanitarian evacuations, sketching out their history, and reviewing the available scholarly literature—limited, and mostly in the social sciences—to say why historians should study humanitarian evacuations. It meant stating more clearly the significance of humanitarian evacuations. And it meant more rigorously using ‘my’ case study to set out a comparative framework for understanding other evacuations, or failures to evacuate.

Illustration of an iceberg showing how much of it is below the surface.
Approximate main text : footnotes ratio (second draft)

All of this took a while. Friends helped, providing suggestions and references: the footnotes proliferated still further. But the AHR report had given me a clear idea of what needed to be done, and once I’d worked out how to do it I made steady progress. By late 2016 I had a revised draft, and a friend who works for Médecins sans frontières had read it and confirmed that it did a better job than the first of explaining the significance of the topic and providing directions for future research. I sent it off—but not to the AHR. This was partly out of insecurity: I still wasn’t sure I’d done enough to make the article relevant to a discipline-wide audience of historians. But having revised the article, and ’embedded’ it in a wider literature on humanitarian evacuations (and evacuees) in a number of different disciplines, I was pretty sure that I’d made the story relevant to historians and other researchers who are interested in humanitarianism. So I submitted it to Humanity, a newish journal dedicated to the interdisciplinary study of human rights, humanitarianism, and development. Because that other article of mine was by now in press with a mainstream history journal, I no longer minded going for a more specialist journal for this one.

Then the wait began, again.

‘Accepted with revisions’

It wasn’t too long. In January 2017 I got an email from the editor with good news: Humanity would like to publish the article—subject to some revisions. One of the peer reviewers had been on the fence, while the other had been enthusiastic but also pointed out that there was some existing literature that I ought to have cited. The editorial collective had liked it. So the editor suggested particular areas for revision based on what the reviewers had said.

This was all ‘good practice’ in peer review. If reviewers’ reports are contradictory, or suggest varying degrees of revision, the author of an article should be able to get a steer from the editor as to which points they should prioritize in revising the article. In this case, I didn’t need to ask: the editor made it clear in his email how I should approach the revisions. And, while being asked to go and revise the article again was a bit discouraging—we do get sick of things we’ve been working on forever!—this was quite different. I knew the article would be published if I made the revisions. I also basically agreed with all of the reviewers’ suggestions.

What do I mean by that? The editor suggested “reframing the intro… so that it is less around when the origins [of humanitarian evacuation] were and more around what the political logics were”: this was pushing me a bit further in the direction I’d already been travelling. I’d reframed the article for the revised draft, but I’d still spent more time than I needed to discussing when things happened and not enough saying why they happened then. The lukewarm reviewer pointed out that it wasn’t good enough to say that my sources didn’t allow me to reconstruct refugees’ own experiences, and they were right. This was a key methodological problem that I’d tried to grapple with (it’s something historians of refugees run into frequently). I couldn’t fix it—at least, not without extensive further research in several different countries requiring languages that I don’t speak—but the comment pushed me to make a more explicit argument about what the absence of ‘refugee voices’ in the French archives tells us. The more enthusiastic reviewer had several helpful suggestions for readings (more footnotes!), including one book that had, shamingly, been lying unread on my table throughout the research and writing of the article, and picked out several points where I needed to provide more information about certain people and organizations or more support for a point of argument. So this revision was much quicker than the more comprehensive rethink and reframing I’d had to do between the first and second draft. It still took a few months, because I had an exceptionally busy spring and summer in 2017. But in early June I sent off the revised version, along with a document explaining how I’d incorporated the reviewers’ suggestions—or why I’d decided not to.

Image of spiral galaxy M51: a bright centre with great swirling arms
Main text : footnotes ratio, third draft

The editor was happy with the revisions. There was just one problem: the article was now a monster, with the main text coming in at just under 10,000 words but the footnotes taking the total closer to 18,000. (Most journals ask for articles in the range of 7–10,000 words, notes included.) Could I please get it down to a more publishable length—say, 14,000 words?

To my own surprise, I could, and pretty quickly: this savage-sounding edit—over a fifth of the total length—took no more than a morning. The successive rounds of comments and revisions had helped me clarify the argument, so I could now see which details were essential and which were interesting (to me) but unnecessary. The immense scaffolding of footnotes had been keeping hold, at each stage, of yet more information that might come in handy for a revision: this could all go. It had also introduced several sets of readings to different potential audiences, but now I knew which audience the article would be read by and could trim accordingly. And that was the version that got the ‘final acceptance’ email, and is now published. It’s still pretty long, but it’s no longer a monster. So now other people can read it, decide what they find useful about it, and explain in their own work what I got wrong.


This, believe it or not, is a somewhat condensed summary of the peer review process. I haven’t done justice to the contribution of five friends who read drafts of the article and sent extensive comments; I haven’t explained how helpful the ‘reverse outline‘ technique was when I revised the draft. But I hope it shows how two rounds of peer review, one unsuccessful and one successful, helped to make a better article—and gives an idea of just how much work other people put into that. (The comments I got from friends, colleagues, and reviewers probably total about the length of the article’s main text.) I’m very grateful to them.


Quotes from correspondence with the editors and from readers’ reports are given with permission: thanks to the current editors of the AHR and Humanity for helping to arrange this.

Click images for source.