Tuesday, September 4, 2012

Publishing, rejection and peer review in philosophy

“Peer review is the worst form of appraising academic work, except all the others that have been tried from time to time.”

Today I want to offer just a couple of reflections on peer review and publication, in the context of philosophy and the humanities. Hopeful they will be of some help or comfort to those just setting out on the tumultuous journey that is academic publication, or – like myself – still wrestling with its slings and arrows.

The first thing to keep in mind – as my bastardization on Churchill’s line on democracy above is meant to capture – is that peer review is by no means a perfect system. To be sure, it has substantial merits. What better way to judge academic work than to send it to experts in the field? Stripped of the name of its author, the experts judge it only by the cogency of its argument, the originality of the thinking, and the plausibility of its premises. The experts make their determination – hopefully offering thoughts for improvement – and the good papers are published and the poor ones passed over (usually to be reworked and submitted elsewhere). For the most part, the system works and is resilient enough to the occasional less-than-ideal behaviour on the part of individuals within it.

Indeed, one can feel a real dignity and virtue in the system when it works well. Reviewers and editors routinely decide to publish papers with whose conclusions they profoundly disagree; but they nevertheless accept the significance and the cogency of the argument and decide it is worth publishing on that basis. So too, the anonymity of the process really can create an equal playing field. Papers written by professors from Ivy League universities are passed over, and articles penned by PhD students from some backwater are selected. I may not have been to Cambridge, but I can (and have) published in its journal of ethics.

So that’s the good news. Of course it doesn’t always work like that. Sometimes, journal editors will reject papers immediately, before sending them out for review. And in such cases usually they will know the name(s) of the author, meaning the process is no longer blinded (processes can be put in place to fix this, of course). I hazard in such cases editors are more likely to give the benefit of the doubt to well-known professors than unknown students. There is no reason to believe editors are more judicious than reviewers in this respect, so it's hard to see why they shouldn't be blinded.

So too, reviewers can be biased. They can evaluate the argument on the basis of whether they approve of its conclusions, or not. Now I think this is often not as large a problem as it might seem, as there are institutional factors that press against it. Effectively, philosophers like (need!) to argue with each other, and to do that requires finding intelligent people who disagree with them. So a good argument against their favourite theory is not something they want to suppress – but rather to publish and then respond to. But bias can rear its head at certain points – both the Sokal hoax and the ‘climategate’ scandal carried the charge that reviewers in those cases focused on whether they agreed with the conclusion, rather than the cogency of the evidence and argument itself.

But the major problem with peer-review is simply this: journal referees are often time-poor, and there are few institutional reward-mechanisms to ensure they carry out their task diligently and judiciously. If a referee makes a laughably wrong decision in rejecting a paper (and the world is littered with stories of what-came-to-be landmark (even Nobel-prize winning) articles that were passed over with brusque rejections) then there is usually no professional fallout from their doing so. Indeed, the institution that employs them will almost never know what happened. This means that every time a referee gives a fair and careful appraisal of a manuscript it is almost always an act of virtue. They did it because they themselves wanted to do the task well, not because there was a system of accountability by which they might be judged and found wanting.

Even if there are no issues of bias or brusqueness, it still remains a grim fact that top journals have (and must have in order to function) very high rejection rates – often well above 90%. This means that reviewers need to be pretty critical; it’s not enough for them to think the manuscript is okay. To be successful the manuscript has to convince two separate referees (and often the editor as well) that this paper simply cannot be passed over; it is brilliant and a perfect fit for the journal. Unless you are a scintillating genius, this means that if you are submitting manuscripts to top journals, you will get rejections.

If you are anything like me, you will get a lot of rejections.

So how to survive, in this world?

The first thing is, of course, not to take rejections too seriously. If you’re in this game, you will get rejections. You will get lots of them. The mere fact that you have had a long series of rejections on a particular manuscript does not mean that the paper is not worthwhile. I have met colleagues who are surprised when I tell them one of my manuscripts has been rejected seven times and I am still revising it and sending it out. (This is not merely my own personal stubbornness; one of the smartest and most well-published early-career philosophers I know told me just the other day that she had hit seven rejections with a manuscript and was still sending it out.) To be sure, in some cases it may be that the paper just isn’t good enough – if all the reviewers tend to be saying the same thing, then changes are in order, at least. But you can just be unlucky. If you believe in the paper and its significance, then you need to stick with it. All of my best publications were rejected at least once by some journal or other. My favourite paper, and the one I feel to be my best contribution to political theory – Two Concepts of Property (The Philosophical Forum, 2011) – had no less than six rejections at various venues before it was accepted. (Though, to be fair to the reviewers that had rejected it, it was heavily revised as it went along, especially in the light of some sympathetic but trenchant critiques put forward by two excellent referees from Social Theory and Practice.)

The point is more general. Just as rejections can make you doubt a paper, they can make you doubt yourself. I think it is fair to say that I have a pretty robust self-belief about my capacities as a philosopher. But when you are hit, again and again, with rejections from multiple sources, about multiple papers, it is easy to start thinking you are getting things wrong in some deep way. But the fact is that if the journals you are submitting to only publish about one in twenty submissions they receive, then even if you are producing high-quality work, multiple rejections will be the order of the day.

With this in mind, it is important to keep in mind that – in publishing like in so many other pursuits – when you are doing well you will tend to chalk it up to your merits, and when you are doing poorly you will put it down to luck. We attribute successes to things in our control, and failures to things outside our control. This is just a part of how human beings think, but it can be a persistent bias. If you (or one of your peers) gets a string of acceptances, then you/they are probably doing a lot right. But you/they are probably also lucky. Around 2009, during my PhD, I received three journal acceptances in a short period of time. Frankly, at that point I thought journal publication was the easiest thing in the world, and that referees were clearly able to recognize manifest genius (i.e. me) when they saw it. Sometime in 2011, I began to realise how wrong I had been. The manuscripts had been good, yes, but I had been lucky as well. And having that lucky start didn’t stand me in good stead when I began to receive rejection after rejection for my next set of manuscripts. Over time it all evens out, but that can mean long periods of rejection after rejection (people are often surprised at how many times a fair coin can turn up heads, if you toss it long enough). That can be pretty tough to handle, especially if you are early in your career, when each publication really counts, and you need them now - not later.

My next point is that there is never a comment from a referee that is not worth thinking seriously about. I mean that. Never ever. So what if they just skimmed your argument? So what if they clearly don’t know the literature as well as you do? So what if they hold to a shoddy interpretation of Hume or Foucault or whoever? It still matters what they thought. When your manuscript finally does get published, it will be read by people who just skim it, who don’t know the literature, and have dubious understandings of Hume and Foucault. You need to consider those readers as well – are there some small changes you can make so as to make your main point sensible to these readers? Could you make the abstract clearer – so that even the laziest scholar will not misunderstand what you are aiming to do in the paper? The point here isn’t to try and please everyone – that’s impossible. The point, rather, is never to write off a reviewer’s comments merely because you are sure they are an idiot. Often, they will still have said something that is worth thinking about seriously. It usually will cost you between two and six months to get those few comments; value them accordingly.

(The point about the abstract is worth emphasis too. I often fall into the trap of getting the paper itself very well-polished, and then realise I need the abstract only when I am about to submit, and just whip something up quickly. This is a mistake. The abstract is crucial. Every word in it matters. It will be the first thing editors and reviewers read – and the first thing scholars generally read when your article is published. You have to get it exactly right.)

Next point: always take the revise-and-resubmit process really seriously. A ‘revise and resubmit’ result should be what you are aiming for when you submit a manuscript. Most good journals will almost never give a straightforward acceptance. And a ‘revise and resubmit’ is usually an opportunity to really improve your paper. The comments you have been given are provided by someone who is clearly sympathetic to what you are doing (otherwise it would have been rejected). This is often the single best source of constructive criticism a scholar can access. Referees and editors will usually not require that you have made every change requested. What they will require, at absolute minimum, is that you have thought very carefully and sympathetically about their comments, and how they might benefit the paper. You don’t need just to do this – you need to demonstrate unequivocally to them that you have done it. (I’ll often provide a document detailing all the changes and my reasons, even if this is not requested.)

One final point: always always always be polite when dealing with journals, even if you feel you have been hard done by. There is nothing to be gained by telling the editor how incompetent one of her referees is, or that she should at least have sent it out for review, or whatever. Editing a journal is hard and often thankless work. The editor cannot second-guess her referees, or her own judgment, or her task becomes impossible. Moreover, venting your frustrations on the editor is a waste of emotional energy that would be better channelled into thinking seriously about whether and how to rework the manuscript. 

Take the criticism to your typewriter, as the old writing adage goes.


Anonymous said...

First, there are few institutional rewards for running a journal. And especially in the humanities, there are no research budgets to support the use of research assistants to do the basic secretarial work associated with editing (and there is quite a bit of that). Accordingly, there is an undersupply of journals to prospective authors. The result is that editors, especially of good journals, have far more publishable papers than they can actually publish and so are happy to eliminate prospective papers for any reason at all, even pretty poor ones.

Second, the fact that the reviewers are generally not rewarded leads to there not really being enough of them. Further the fact that the process is anonymous to the author encourages sloppy, self-interested, and often cruel reviews. The editor is the only person who will attach a name to the review. If the author complains to the editor? Go back to point one. The editor has very little motivation to care or to do anything about it. They have too many papers anyway.

This anonymity parallels the problems with social media and ‘trolls’ engaging in vitriolic abuse and hiding behind a username. There has been a bit in the news about this recently following the attempted suicide of a blogger under the sustained attacks of nameless commentator unwilling to speak publically. I’ve seen it argued that free speech only works when it is not anonymous: Martin Luther signed his own name to the 95 thesis and all that.

Third, the major problem here is you are obviously without talent. This is obviously the best explanation for your continued rejection: you’re a talentless moron. (It’s called an ‘argument to best explanation’ – have you ever heard of it??) But who would expect more from a communist, three-hugging, self-proclaimed ‘philosopher’? Worse, you’re ugly and bald, and I’ve got it on good authority that you smell. You’d probably have dreadlocks if you weren’t so bald. And flees. Why don’t you spend less time complaining about the ‘unfairness’ of the ‘system’ and more time working? Listen to you: The world is not fair!! I deserve success handed to me on a platter. Boo hoo hoo. Next you’re going to be complaining that it’s all the fault of white men with no friends and an inferiority complex. Grow up and get a real job!

Hugh Breakey said...

Blinded reviewers as the academic version of social media trolls! This anonymity issue is clearly more worrisome than I had thought...

William Ferguson said...

Clearly there are scarcity of resource issues at work, namely quality reviewers and journal spots.

Ways of dealing with those might be to name the reviewers alongside the published paper thereby providing some merit to those reviewers that help to find/refine papers that are deemed quality. Similar meritocracies have worked well in the IT world on sites like StackOverflow.

On the journal front, it seems unfair/uncouth/unintelligent to be rejecting good papers just because there isn't the room for them. I understand that this is a financial decision, after all producing a journal requires funds and the more articles the greater the expense. But digital publishing has reduced the cost in a number of areas and it may be that it would be worthwhile splitting the journals into two. The first being the current top end papers, and the rest being those that are good enough quality to publish but are deemed by the editors to not be star quality.

But my comments are all from the sideline as I don't ever expect to have anything of my own academically published.

Hugh Breakey said...

Hi Bill,

Yes; I think there is probably a fair amount to be learned off the way IT developed its systems of merit and credit. They seem in many respects much more sophisticated and judicious than we have in academia.

I do have sympathy with the journals needing to constrain their size though. Committing to publishing a journal is different from publishing an encyclopaedia (as is deciding to read one). And if the big journals (especially the non-specialized ones) really did publish everything decent they were sent, they probably would soon become encyclopaedias!