For more than 15 years, books available only in paper form have fought a losing battle with digitally-available articles in academic journals – the publishing equivalent of horse cavalry repeatedly charging barbed wire defences with machine guns. As their usefulness and effectiveness waned, so the intellectual status of books in the social sciences declined strongly. In the first of a two-part blog post, Patrick Dunleavy traces the declining role of books that reached a nadir in 2010. Part 2 of the argument explores the second coming of books in digital forms.
Three years ago, drawing up my reading lists for the new academic year, I realized that I had almost stopped setting books altogether, in favour of journal articles. The reasons were simple. University reading lists are now generally held on some form of electronic ‘learning management system’ (LMS), such as Moodle (now claiming 58 million student users) or Blackboard. If I include journal articles on my Moodle reading list, students have instant one-click access to a free electronic copy (via LSE’s library). They can download PDFs, and keep them permanently in full text form beyond the seminar week, using the electronic article for later essay writing and revising for exams. In addition, the whole class can access and read the same materials simultaneously. And I can add journal articles right up to the last minute in digital on-demand form.
By contrast, if I cite a paper book (and they were almost all paper-only three years ago), none of this is feasible. I would have to ask students to spend hours chasing around the library in search of elusive paper copies of books. Even for key references, they would have to wait to access one of the necessarily limited number copies in our teaching collection. Certainly the whole of six seminar groups for my courses could never access the same books at the same time. We would do our best with electronic study packs, but they could only cover one chapter per book for copyright reasons and they took our Library months to prepare. You could not have any last-minute thoughts on what to put in the e-packs.
As in so many other industries and areas of our social life then, the competition of ancient, pre-digital forms of products with modern digital forms was inherently uneven from the outset. And for a long time it only got worse. The outcome was foreseeable and known to everyone in the publishing industry and in academia decades ago. But conservative book publishers were determined to follow the music industry down the path of digital denial for as long as they possibly could. Partly this reflected some genuine intellectual property rights (IPR) anxieties, but partly also an entrenched commitment to existing ways of doing things.
Academics were also in digital denial in a big way, especially in the humanities and social sciences. Here books remained a core medium of scholarly communication. And often in these areas referencing practices have been so poor for so long that very many journal articles are not cited by anyone – especially in the humanities. Hence academics were mad keen to hang onto books, as the only things that got (lightly) cited.
And of course academia and publishing are both full of book fetishists – people who genuinely love reading books, objectify them, love bookshops, love the heft and feel and smell of books, lap up articles about how anonymous little grey Kindles can never compete and so on. (Disclosure time – my house and my LSE study are both packed with thousands of the blighters too.)
The declining prestige of books
So, just as horse cavalry continued to be funded by armies and nostalgic politicians long into the era when barbed wire and machine guns made them completely useless, the illusory position that paper books could still compete in a digital world was maintained. With what results? I am a great believer in the importance of teaching in conditioning and colouring the whole of academic life, research included. In the social sciences, unlike the STEM subjects, very few academics are funded just to do research. Full-time researchers are far fewer in the social science (10 per cent of staff, as opposed to 35 per cent in the STEM subjects), and often dependent on vulnerable ‘soft’ budgets.
The social sciences also constantly change, with the time-lag from research frontier to teaching use being correspondingly far shorter than it is in STEM subjects. So for better or worse, what gets taught in our disciplines influences how social scientists see their subjects quite broadly and deeply. As ‘intermediate text’ and professionally read books increasingly stopped being set for courses, even for specialist courses, and only first year textbooks remained, so the academic status and relevance of books as whole declined across the social sciences. Because we set only or mainly journal articles in our teaching, and increasingly also discussed only them in classes and seminars, so we increasingly read only or mainly articles and books dropped out of sight.
In parallel, the ‘scientization’ of many social sciences was proceeding apace on many different fronts (allegedly driven by ‘physics envy‘ in fields like economics and political science). Whatever the reasons, ‘scientizers’ mostly focused single-mindedly on journal articles and suggested that books are not ‘properly’ peer-reviewed. For them, books became an artefact increasingly associated with disciplinary ‘obstructionists’ and reactionaries
In the UK a long series of bureaucratically structured ‘research assessment’ exercises (the RAE) led up to the current Research Excellence Framework (REF) in 2014. All were driven by a Treasury concern to demonstrate that the specific amount of money pumped into research support in universities was delivering results. Universities ensured that this was denominated in the form most pleasing to academic-bureaucrats (let’s call them ‘bureaudemics’) – that is, articles in prestigious or international uber-academic journals, whether or not these papers were ever subsequently read, used or cited by anyone. Journals responded by publishing more (often unreadable and unread) stuff.
In bureaudemic terms, books were too hard to assess – each of them was so sui generis and so unique. The myriad RAE panels, involving hundreds of senior academics – who should have known better – maintained the pretence of ‘reading’ (i.e. eye-balling) 200,000 research ‘outputs’. But everyone involved knew perfectly well that these mass-manufactured judgements were just going on journal reputations for articles (easy to do with the actually meaningless averages of journal impact factors).
When it came to books the bureaudemic process relied on unspoken and unexamined biases about types of books and the reputations of publishers. So an obscure monograph with an Oxbridge university press (a dwindling numerous product) would always rate a top mark, but a book from a (shudder at the thought) ‘commercial’ publisher must be less ‘research-based’. Over time in theUKthe numbers of books submitted to the RAE progressively shrank as these bureaudemic pressures cumulated.
If books were hard for panels of bureaudemics to assess, book chapters were impossible – they were steadily excised from the sacred ‘4 items’ to be submitted to the RAE. The number one casualty were chapters in edited books, which moved from being a well-used and well-respected format in the soft social sciences and humanities especially in the late 1980s, to becoming an increasingly suspect offering. To the rushed readers on RAE panels, and increasingly in the eyes of promotion committees within universities, hard-to-assess chapters in books smacked of an effort to evade peer review, by publishing with mates instead of going through the rough and tumble of journal submissions. Of course, the bureaudemics in many fields also knew that knowledge of chapters in books in academic professions was anyway slender.
Because, of course, chapters in paper books have become less and less visible in the digital world. The proprietary bibliometrics systems defined in the 1980s, like Web of Knowledge, covered only journal articles, ignoring all forms of books and book chapters (because they were unimportant in the STEM world). The enormous benefit conferred on the academic community by Google Scholar did begin to redress this, covering both books and book chapters along with articles.
But even here, chapter were so numerous and so untagged that they were hard to track. Chapters had very low initial audiences compared to journal articles circulated in print form to many professional academics and publicized through email alerts. For chapters in edited books though, unless you could physically handle the paper book in a bookshop or a conference, you had no way of knowing they were there. Even publishers’ catalogues now rarely give complete titles and authors for chapters.
Google Books, another extraordinary gift to academia from Messrs Page and Brin, functioned well at the whole book level, but it was not indexed for chapters. So the pathways to uncited publications lead increasingly broad and true to chapters in books. And as citation metrics began to become more important in academia, so the lesson for any remotely ambitious academic became plain – don’t write chapters in books.
Who publishes what in the social sciences now?
All of this brings us to where we are now in the social sciences. For a random sample of UK social scientists across twelve disciplines, we tracked all their publication outputs, of which just over three fifths were journal articles and a fifth were books, book chapters or edited books. We next looked to see how citations to these outputs in Google Scholar and Web of Science were distributed across disciplines. Figure 1 below shows the breakdown of results for all forms of book outputs versus journal articles.
Figure 1: The relative importance of book outputs and journal articles in citations within each social science discipline
Source: LSE Impacts of Social Sciences database.
In economics and psychology there were eight times as many citations of journal articles as there were of books, but in sociology and media studies the ratio was more like two to one – with the remaining social sciences spread out in between. In philosophy and history which are predominantly in the humanities, book outputs were also as or more important than in sociology in terms of citations.
However, this Figure puts together all citations to three different kinds of book outputs –authored books, edited books and chapters in edited books. In my second Figure below I instead look at citations to these different kinds of outputs separately, and find that there are some sharp variations. The importance of whole books is shown in orange and the disciplines are arranged in their ordering on this dimension, with media studies and sociology at the top and with geography and anthropology at the bottom. Edited books are cited in a fairly standard way across most disciplines, but are absent or almost absent in psychology and economics. Chapters in books fluctuate sharply in importance. They remain most cited in philosophy by a long way, followed by sociology, media studies, history, law and geography. Book chapters are least cited in economics, followed by psychology and business studies.
Figure 2: The relative importance of books, edited books and chapters in books in citations in each social science discipline
Source: LSE Impacts of Social Sciences database.
The turning point
So this is where we stand today in the social sciences, at a very bad pass for books and even more for edited books and book chapters. But luckily the story does not end there, because many of the trends of the last two decades have now begun to change. We have in short reached a turning point, a moment in history when a renaissance of books’ influence can be foreseen across our disciplines. The key to this change is that books stop being only or even primarily paper products, and make the transition to ebooks and other digital forms. In addition, there are also many other newly favourable influences, such as the rise of the impacts agenda.
Read the second part of this blog post.
Patrick Dunleavy is Professor of Political Science and Public Policy at the London School of Economics and Political Science, where he has worked since 1979. He has authored and edited numerous books on political science theory, British politics and urban politics, as well as more than 50 articles in professional journals.
Looking forward to part 2 of your post and what it contains, but one thought which occurs to me is that the article is far more suited to the digital world than the book – as exemplified by David Weinberger’s description of the Internet as “Small pieces loosely joined”.
How have you dealt with the fact that Web of Science only. Tracks citations in journal articles? And is the Google Scholar citation index fully incorporating books? This is the big validity problem for bibliometric measures in disciplines that publish in books. Citations of books in articles are captured but citations of anything in books are not.
As for teaching, I think many people stopped assigning books long before Moodle, largely because students wouldn’t read that much. Chapters or excerpts, yes, but articles were often the staple of reading lists.
I too look forward to the second part.
“Google Books, another extraordinary gift to academia from Messrs Page and Prin” [sic., should be Brin ]
Thanks for pointing out the correction, now updated.
Cracking article – I am a fogey so won’t be giving up paper books anytime soon but the points you make are really good ones – many publishers are now looking to incorporate books into their online platforms so that content within is as easily searchable as within journals – which should open a lot of people’s eyes to the significant amount of research being undertaken in book format – even in economics, one should bear in mind that neither The Theory of Games and Economic Behavior nor The General Theory were journal articles. The real game changing contributions are arguably always going to be in book form.