The cover page to Søren Kierkegaard's universi...
The cover page to Søren Kierkegaard’s university thesis. (Photo credit: Wikipedia)

I’m very honored to join the coterie of Scholarly Kitchen chefs.  I’m a regular reader and I’ve learned a lot from my fellow chefs.  But, it was only after I joined that I fully realized I was the only person writing from the humanities and social sciences, and the first from the point of view of a university press book program (though, of course, Joe Esposito writes insightfully about them all the time).   Because of this unique opportunity, I hope to use this blog to write largely to represent and explain issues key to the world of scholarly publishing I inhabit, while also making links, comparisons, and points of similarity and difference to chefs who occupy different kitchens.

There are a lot of practices and prescriptions for academic publishing that assume all publications are like STM ones, or all are basically alike.  Those mistaken assumptions have big consequences.  They can undermine well-intended but unsustainable models for open access or lead to ill-devised modes of evaluating scholars by employing impact factors or journal ranking in fields in which they make little sense.  I’ll try to bring out some of the differences apparent from where I stand, while attempting not to replicate such generalizations in the other direction.

One of the most marked differences is in the relation between research and publication.  At Duke University Press we publish several math journals, and, through Project Euclid, distribute many more.  My journals colleagues often remark that by the time an article is published in one of the math journals, everyone in the author’s subfield has seen it.  The publication is still an important last step for many reasons, including citation, but the reading has already occurred.

A few years ago, Phil Davis wrote a blog post about the delicate balance researchers face between secrecy and sharing. In a recent conversation, David Crotty noted how increasingly rare it is for biomedical researchers to present unpublished research at meetings, most likely out of fear of getting scooped (and how often one sees audience members taking pictures of talk slides or posters).

In either case, the research and the publication have a very close relationship.  In humanities and social science, the research itself can be more-or-less concluded five or ten years before the publication.  That alone is unremarkable, I suppose.  The kicker is that the research can be of minor importance as originally presented or published, compared to the influence of the eventual publication.  What gives the work its importance is the author’s argument, the smart and useful understanding they produce.  The argument or its interpretive, theoretical framework is usually the feature of a work that makes it widely read in and across fields. It can also be the last element to arrive.

I had a very graphic example of this a few years ago. A book I helped shape from a dissertation through several rounds of peer review and revision over three or four years was finally brought to our faculty board. They liked and approved the book but thought we needed a better title.  The title the author and I devised at that last instant became the central claim of the author’s revised introduction.  That idea is now what the book and author (now a full professor) are known for.  The research itself was available on Proquest and at the author’s university the whole time.  There is doubtless more raw information there.  But without the argument it is like raw film footage that hasn’t yet been edited to tell a story; of value to only a few researchers in the field.  The argument that accompanied the title is what made the book a hit, rather than something that went straight to cable.

Dissertations present much of their author’s research.  Cultural anthropologists report back on their fieldwork, describing what they found, analyzing it in the language of the field, offering preliminary insights, and reviewing existing scholarship and putting their work in that context.  The literary critic reads a set of texts, exploring each in turn, chapter by chapter.  Each scholar is demonstrating to their dissertation committee an ability to perform the core practices of the field.  The difference between this kind of object-centered interpretation and having an original argument that can organize and sustain a book is huge.

It’s a rare dissertation that I would send out to reviewers.  This isn’t only because of the dissertation features (like a literature review); it’s because the author is usually only at an early point in developing her or his own argument.  A major function of peer review in the humanities and social science, especially on these first books, is to encourage the author to develop and own an argument, to center it, to organize the rest of the work more clearly around it. Helpful peer reviewers walk the author through this process step-by-step, suggesting where a paragraph is needed or how different parts of the argument could be articulated together.

This process takes the publication further and further from the research itself. The research is cast in a new role.  It is not merely filling in a gap in knowledge (and of course many humanities objects — a work of Kant or set of novels — are already well known) or demonstrating the scholar’s ability to perform the work of the field.  The research provides its value as the grounds for making a scholarly argument in a publication that gains traction in and beyond the field.

It’s an odd but telling mistake when libraries make a budgetary decision not to buy books based on dissertations for just this reason.  What could have more distance from research to book than something that started as a dissertation? A senior scholar’s fifth book is far more likely to be presented in a recognizably nascent form in a journal article than an author’s first book would be found in a thesis.

Publication in the humanities and social sciences isn’t the reporting of research.  It’s the production of a compelling argument, based on a combination of research and interpretation. Those in a field may watch a scholar develop an argument in talks, but it generally only comes into being in writing, whether print or digital, journal article or book.  It’s not the recording of research but the forming of the argument that takes time and has lasting value.  That means we need different mechanisms for editorial review, process and publishing than the sciences might need.

I am not arguing that we must maintain the same forms of scholarship or scholarly publishing.  What I am arguing is that the differences between how we publish and why in different fields must be kept visible if we are to support those differences as we move to new forms.   For the humanities and interpretive social sciences we need to start with a realistic account of the value of argument and its relation to research.


32 Thoughts on "The Relationship Between Research and Publication, Or Why Libraries Should Buy More First Books Than Any Others"

Great piece – finally talk of arguments that are made by scholars rather than being given results of important questions! Some theoretical work in the natural sciences might fall into the interpretive category of publication too. In the UK-descended higher ed systems which have government-defined research metrics, one-size-fits-all evaluation dominated by STEM-disciplines is a major problem. Even where exceptions are made, this is not the same as thinking through the actual needs of humanities and social science fields. The fine arts is the limit case of this kind of publishing, and this situation now also needs to be thought in the academy.

Fascinating food for thought. This may relate to what I understand to be the fact that the article citation curve for the humanities and some social sciences peaks much later than for STM. This may be true for technology papers as well. Which in turn factors into the embargo time issue now before the US funding agencies and hence has great practical value. I will have to think about it, to formulate an argument.

it is certainly the case that an argument can take a long time to develop. In my own cognitive/social science work my understanding often came well after my research. My taxonomy of confusions for example was based on many years of small, specific analytical projects.

I’m not yet convinced that citation half-life is the right metric to be used in arguments for setting embargo periods. Is there a correlation between citations and subscriptions? I may read and download an article in the first week it is out, and then cite it 10 years later. Perhaps usage is the more important measurement in this area.

I would argue that what we need is a model not a metric. K’s post suggests the basis for such a model.

So glad to see your voice in the kitchen, Ken, a greatly needed corrective to views based solely on STM.

Interesting post, thanks. But I’m looking for some support for the alternate title: “Why Libraries Should Buy More First Books Than Any Others.” Towards the end I see an argument that it’s a mistake for libraries to exclude books based on dissertations — but I don’t see any explanation as to why most of our acquisitions ought to be first books.

I like this piece–a clear explanation of the work done between dissertation and book–but the library part of the title seems a last-minute toss-in to draw readers. Certainly it caught my eye, but the sloppiness is irritating–it doesn’t even represent Kwissoker’s argument about library acquisitions.

Why go on so about this? Titles, especially book titles, are a professional interest for me, as academic libraries increase their levels of demand-driven acquisitions. The purchase of a book depends, now more than ever before, on a successful match between the searcher’s words and the words in the book’s description, including the title.

So publishers, take note, catchy titles have their place, but accuracy and clarity get the job done.

thanks Rebecca. I think your observations about the perhaps inaccurate or late-added title do point to some intriguing issues.

While ‘core’ qualities of a work, scholarly or other, obviously matter — the argument, the research basis, originality, etc. — a significant part of a work’s reception and impact may be shaped by what you might call it’s garb. That is, titles, covers, conceptual frames, indexes and abstracting & indexing, now tweets or post titles, etc.

I’ve long mulled this point in the context of design and graphic design work, where “presentation” is perennially contrasted to “content,” or style to function, etc. I like the metaphor that a tool can’t be used unless it has form by which it can be recognized and grasped, and the ‘handle’ is usually inseperable from the ‘tool’ (e.g. hammer’s handle, pliers’ grips).

In the context of scholarly publishing, especially books, we might think about how the ‘tool’ (book) might more efficiently, effectively, & variously find its many handles, or paths to reception and relevance. For example, might a book be incrementally or nano-published, systematically, while being written and edited to discover its readership and resonance? I’m reminded of the recent development by Gawker Media’s Kinja platform to support users rewriting headlines and reframing stories (

Perhaps catchy, accurate, and clear, are all usable handles, and a good book finds many frames.

Tim McCormick
@tmccormick Palo Alto, California

I wouldn’t argue that most acquisitions should be first books. I’m just arguing that they shouldn’t be excluded if the grounds are libraries already having access to the same research. One could argue that in some disciplines like history or anthropology first books represent the longest period in the archive or field that a scholar ever gets. Senior scholars couldn’t take as much time on any given project. Of course, not all first books are based on the authors dissertation, though it is my experience that most start that way, and then move progressively away from the original form, through years of peer review and revision.

I wouldn’t argue that most acquisitions should be first books. I’m just arguing that they shouldn’t be excluded if the grounds are libraries already having access to the same research.

That is indeed a much more modest proposal than what your title says. I think it’s interesting, though, that so many publishers seem convinced that excluding dissertation-based books is a prevalent practice among libraries. Having worked as both an academic bookseller and a collection development office in multiple research libraries, my experience is that while most libraries are careful to exclude unrevised dissertations, and some might subject revised dissertations to additional scrutiny before making a selection decision, relatively few take a hard stance against books that have their origins in dissertations.

That said, budget times are increasingly hard for libraries. Given that most research libraries subscribe to dissertation databases that include the unrevised versions of most dissertation-based books, it’s possible that there will be a trend towards greater exclusion of them in the future as money gets tighter and tighter. Steeply-rising prices combined with relatively flat budgets inevitably means fewer acquisitions.

That’s exactly the argument my piece is attempting to contradict. Yes, libraries are having to make tough budget choices. Making those choices well requires smart criteria. My argument is that ‘dissertation-based’ is not a helpful or accurate criteria.

But I think you’re missing the main point of my comment, Ken, which is that I believe this issue is something of a red herring. In my experience, relatively few libraries actually exclude dissertation-based books. What I think would be very interesting, though, would be a study of those few libraries that do. Without knowing for certain what their reasoning is, I don’t think a publisher is in any position to pronounce judgement on it. One of the difficult things about being a library is that you don’t have the option of buying everything that deserves to be bought, so acquisition criteria will always–always–exclude worthy titles.

I made a similar argument in “Dissertations into Books?” (April 2007): The topic of whether libraries do or do not buy fewer revised dissertations than other types of monographs was debated, perhaps ad nauseum, by me and Rick Anderson on Liblicense. I have partial hard evidence reflecting the lower sales of revised dissertations in Latin American studies based on 20 years of sales at Penn State Press, which Rick disputes. I suspect Duke U.P. could produce similar evidence. So I agree with much of what Ken says here, except that I would caution again seeing the revisions as only being about interpretation and theory; in some fields, like comparative politics, the process of revision not infrequently also involves further field research, not only in the country initially studied as a case but in additional countries also, thus extending the empirical reach of the work as well.

The topic of whether libraries do or do not buy fewer revised dissertations than other types of monographs was debated, perhaps ad nauseum, by me and Rick Anderson on Liblicense.

This is not true. What Sandy and I debated was whether or not publishers should be expected to tell their library customers that a book is a revised dissertation, with Sandy arguing that they should not. (Gluttons for punishment can see the “Cost Savings” thread here.) At the end of that debate I said, in passing, that “The great majority of libraries do not exclude revised dissertations from approval coverage, and no library relies entirely on approval plans to select books for its collection” — but Sandy’s argument about those sales figures was with a bookseller, not with me.

I have partial hard evidence reflecting the lower sales of revised dissertations in Latin American studies based on 20 years of sales at Penn State Press, which Rick disputes.

No, let’s be clear. I can’t dispute the evidence because Sandy hasn’t offered to show it to me. He has only said that the evidence exists, and that it supports his position. It’s true that I’m skeptical–first, of the evidence itself, which I have not seen; second (assuming it does exist and says what Sandy claims it does), that it can responsibly be generalized to the marketplace beyond PSUP books on Latin American studies. This is one reason that I think a more rigorous study would be interesting.

The person who argued that publishers should not share information with libraries about the origin of books as dissertations was not I, but rather my former Penn State Press colleague, Tony Sanfilippo. I actually am happy to share the data with rRck, but he ended the conversation last time by saying that he didn’t want to continue it, so I didn’t send him the data. Do you want to see it now, Rick?

The person who argued that publishers should not share information with libraries about the origin of books as dissertations was not I…

It most certainly was you, as anyone who goes back to the list archive will see. (Here’s one relevant direct quote, from your posting of April 24: “If there is reason to believe that librarians are reluctant to buy revised dissertations, then indeed providing that information is a disservice to authors.”) Tony participated in making that argument as well.

I actually am happy to share the data with rRck, but he ended the conversation last time by saying that he didn’t want to continue it…

Well, let’s be more precise. I ended it by saying I was tired of my arguments being either ignored or deliberately misquoted. I would be very happy to see the data — although, as I said before, the data from one topical area of one publisher’s list isn’t going to tell us much about the larger market dynamic. (Much more suggestive of the broader dynamic is the data Michael Zeoli of YBP provided, indicating a negligible impact on UP publications generally when they are identified as revised dissertations.)

To be clear, I did indeed come to Tony’s defense and made the statement you quote later in the discussion, but my recollection is that the thread was started by a comment from Tony. Am I wrong?

I’ll send you the data. Perhaps we can join together and urge the AAUP to start collecting data on sales of revised dissertations. It would seem, from the new AHA embargo policy that Rebecca pointed out, that the AHA believes editors are not reviewing revised dissertations as readily as they once did, and that must be because they believe the online availability of dissertations must be affecting sales. So the AHA, at least, is buying my argument. If it is a faulty argument, as you claim, then it would seem to be important for us all to gather more data to test the hypothesis more rigorously.

To be clear, I did indeed come to Tony’s defense and made the statement you quote later in the discussion, but my recollection is that the thread was started by a comment from Tony. Am I wrong?

Yes. The discussion of the impact of “revised dissertation” began with your posting of 14 April, asserting that “many librarians have ceased including these titles in their approval plans.” After I responded to that posting, saying that your assertion was not borne out by my own experience at YBP administering approval plans for libraries, Tony jumped in.

… which means that i did not initiate the discussion about withholding information about the origins of revised dissertations from YPB. Tony did. That was the discussion i initially said i had not started.

It’s a fair comment on the title, Rick and Rebecca, since the piece was about a wide set of implications, including how scholars are evaluated by administrators and how one might imagine a plan for open access. Library choices was an example, not the central focus of the piece. It’s my first one. I’ll learn!

Ken, I hope my comments haven’t come across as too critical or harsh. I, too, am very happy to see someone from a UP joining us in the Kitchen and I look forward to further postings from you.

Ditto, Rick, with apologies for being harsh. Kwissoker, your piece is spot on, clearly written, and I’m glad you wrote it. My interest in working titles got the best of me, I’m afraid–the inevitable result of a career helping people find the information they need.

As someone working at a scholarly press that publishes primarily in the humanities, I’m glad to have a voice from “my” side of the fence; I look forward to more in the future.

I would, however, argue that the process of intensive peer review shaping and refocusing scholarship into novel interpretations might be better applied to second books rather than revised dissertations. Even after peer review and extensive revisions (a *bit* more than fixing the occasional typo, as some of our first-time authors expect!), dissertations still retain a bit of their “revised by committee” flavor, with odd side arguments or somewhat incongruous interpretive strategies that the author was forced to include by their dissertation advisors, but, in the process of writing and researching, somehow integrated them into their main argument to such a degree that they can’t be excised. To use a perhaps odd example, it’s like how Bodrum Castle in Turkey (formerly Halicarnassus) uses stones from an older building, the Mausoleum of Maussollos, in its walls; the white marble of the Mausoleum sticks out clearly, but is so integrated into the wall of the castle that, even supposing you’d want to remove it, you couldn’t.

Second books, however, are entirely the author’s, and were built completely out of her own interpretive stones. These tend to be much more coherent, more easily revised, and, in the end, more profound. Additionally, authors who have published previously tend to have a better idea of what is expected from them, are often more responsive to peer review, and are just plain easier to work with from our end. While there are exceptions (some students had good advisors who educated them on scholarly publishing and guided their dissertation to something more “booklike,” and students who graduated from German-speaking universities have to publish their dissertations without significant changes within a certain time span in order to receive their doctorates, making extensive revisions based on peer review practically impossible), I think second books tend to be much more insightful than first ones, but without having the ossified “book by numbers” flavor some works by senior scholars have.

It’s hazardous to make generalizations about any kind of book, but I can cite enough examples of first books based on dissertations that became classics in their field, such as Susan Okin’s “Women in Western Political Thought” and Peter Evans’s “Dependent Development,” to counter the characterization of dissertation-based books here. Sometimes, in my experience, the first book is the best book the scholar ever writes!

There’s also Wittgenstein’s “Tractatus Logico-Philosophicus” to add to your list, a classic not just of a particular field, but of philosophical literature in general. There are also innumerable works of striking originality by established scholars written late in their careers. Most rules have their exceptions, some of which can be prominent. Nevertheless, I’d contend that the majority of revised dissertations published as books still retain some of their dissertation flavor, to say nothing of book proposals based on dissertations.

Timely blog post from the American Historical Association: a proposed 6-year embargo on history dissertations, in order to protect the market for the book:

Welcome Ken. A great addition to the team, and look forward to your posts.

Comments are closed.