[Update: my book on The Edited Collection: Pasts, Present and Futures (inspired by this post) was published by CUP in 2020.]

Dorothy Bishop (@deevybee) last year argued that (at least in neuropsychology) ‘if you write a chapter for an edited book, you might as well write the paper and then bury it in a hole in the ground’. The issue is accessibility: (to paraphrase a little) most books aren’t available online as journals are, and no-one goes to libraries any more. (Read the post on Bishop Blog or as republished in the LSE Impact blog.)

Bishop admitted that things might be different in the humanities and social sciences, and something about her argument didn’t quite ring true with my own experience in history. Opinion on Twitter and amongst colleagues was divided: one eminent colleague had reached the point of refusing to contribute to edited volumes, so fast did they disappear from view; another thought that publishers were in collective flight from a format that had previously been fundamental. Others thought history one of the exceptions to an otherwise useful rule.

We are rather short of useful data on this. But my impression is that the format works in a different way to the (mostly online) journal. Granted, few of them are available digitally, and so no-one will find them by search. However, for as long as at least one article in the volume remains current, then readers will be picking the volume from a shelf; and so the other articles in theory at least remain visible – more so than in a journal issue. I’ve heard it often said that if a piece of work isn’t online, it may as well not exist at all; I think historians do still spend a good deal of time in libraries, picking books off shelves. I certainly do.

And then, as @tjowens pointed out, a coherent volume stands a good chance after a few years of becoming in effect a textbook, standing as a recent summary of the state of a particular field. I can certainly remember such volumes as an undergraduate; and my memory is that I read more of these than the weighty monographs listed alongside them. And although it isn’t properly recognised and rewarded, editing a text that influences a whole generation of younger minds should be an important part of what scholars do.

“But they’re not peer-reviewed!” Well, yes, if one accepts only one mode of peer review as legitimate – blind peer review, brokered by a journal. I would argue that some edited collections go through a different process, that is at least as creative of better work than the traditional system. Two years ago I was an invited plenary speaker at a tightly themed conference, leading to an edited volume. I responded to the theme as proposed; the paper was discussed at the conference, not least with one of the protagonists who happened to be living in retirement not far from the conference venue (the joys of contemporary history). A revised draft then went through two series of revisions with the two editors who organised the conference, influenced by an exchange of drafts between the contributors. It is now inproved far more than as a result of two or three vague paragraphs from a journal review. Does such a system place too much power in the hands of the editors ? Possibly; but it is at least open and transparent.

But so far these were only my impressions; and so I decided to create some data of my own. I looked at all the works that I myself have cited in the past six years: data from ten article-length pieces published since 2006, including two unpublished items at the copy edit stage. The field is the recent religious history of Britain, including writings on the sociology of religion, musicology and the history of the plastic arts and drama. (I’d be happy to expand on methodology if anyone is interested.)

Three interesting patterns came from the data.
(i) Citations of chapters in edited volumes formed a (to me) surprisingly high proportion of the whole, some 23%. (More later in another post on the humanities monograph and the invisibility of the journal article.)

(ii) These papers have a decent longevity. I looked at the time elapsed between the date of publication and the date at which I was making the final revisions to my own paper (ie. when I was actually citing it). Far from it being the case that a two or three year old paper is outdated, the median time was ten years.

(iii) I looked at the overall age profile of the volumes, the mean average of which was 14.9 years (to 2013); and there were few that I would not cite again if I were writing today.

All this would suggest that the edited volume continues to play a role for history, or at least for the kind of history that I write; and that Bishop’s observation doesn’t hold true. I should admit that my field is thinly documented –  several of my own pieces broke almost completely virgin soil – and so it may be that for scholarship on areas such as (say) Nazi Germany that are rather more densely overlaid with written work, the picture may be different.

Finally, what of the future ? The timescale of the data didn’t allow me to see whether we are indeed seeing the beginnings of a flight by authors from edited collections. Without data from publishers on the number of approaches they receive, that would be hard to establish empirically. However, data like mine would start to show that effect in a few years’ time. For now, rumours of the demise of the edited collection seem a little premature.