The World Into Which This Journal Was Born

It is hard from this distance to re-envisage the world into which the British Journal of Political Science was born in 1971. There may or may not have been appreciably fewer political scientists in the world back then.Footnote 1 Certainly, anyway, political scientists are currently much more career-active than they were in the early 1970s. You notice that in the numbers attending conferences. The size of the American Political Science Association Annual Meetings has trebled over the period – and with it the number of conference papers people are trying to place in journals. You notice that in general expectations concerning productivity. APSA’s 1971 president (Robert Lane, no slouch when it came to publishing) had published only two-thirds as much by the time of his presidency as his 2008 counterpart (Peter Katzenstein).Footnote 2 The change in Britain is even more striking. Brian Barry recalls:

Such stubborn refusal to publish became utterly inconceivable in British universities in the wake of the Research Assessment Exercise (RAE) that started in 1986.

By the same token, back then there were far fewer journals in which to publish academic political science. Most of the long-established journals were designed to appeal to the interested general public. For American examples, think of Political Science Quarterly (founded 1886), Foreign Affairs (1922), Public Opinion Quarterly (1937) or even World Politics (1948) in their early years. For British examples, think of International Affairs (1922), Public Administration (1923), Parliamentary Affairs (1947) or even Government and Opposition (1965). Many articles in those journals were by non-academics. When academics did write for them, they were obliged to do so in suitably non-specialist ways that would appeal to a genuinely general audience.

Of course, there was the American Political Science Review. For most of its first sixty-five years, however, it had broadly shared those same features.Footnote 4 By 1971, though, it had been publishing a steady stream of modern political science (alongside a continuing but declining stream of older-style work) for a decade or more. Much the same was true of the Journal of Politics (dating to 1939). Then there was the Midwest (later re-badged American) Journal of Political Science, which had been leading the behavioural charge for fourteen years by the time BJPolS appeared on the scene.Footnote 5

Perhaps the most striking feature of the journal landscape circa 1971 was the paucity of well-established subdisciplinary journals. All the journals that we now regard as premier journals of the subfields of political science were, in 1971, either fledglings or yet to be born. That was true of the major journals of political behaviour (Political Behaviour and Electoral Studies).Footnote 6 That was true of the major journals of comparative politics (Comparative Politics, Comparative Political Studies and Politics and Society).Footnote 7 That was true of the major journals of political methodology (Political Methodology and Political Analysis)Footnote 8. That was true of the major journals of political theory (Philosophy and Public Affairs, Political Theory and the Journal of Political Philosophy).Footnote 9 That was true of the major journals of political economy (Public Choice, the Journal of Theoretical Politics and the Quarterly Journal of Political Science).Footnote 10 That was true of major journals of international relations, both British (Millennium and the Review of International Studies) and, in a different way, American (International Studies Quarterly and International Organization).Footnote 11

Note, finally, the relative rarity of new generalist political science journals post-1971. After the European Journal of Political Research was launched in 1973, there was really nothing until the European Political Science Review in 2009.Footnote 12 The explosion of subdisciplinary journals around the time of the launch of the BJPolS seems to have sucked the air out of the publishing atmosphere for new generalist journals.

Crafting A Niche

What are the editorial implications of those ecological circumstances? The principal one, I think, is that it was easier for generalist journals to be truly general in earlier periods. Before around 1970, there was simply nowhere else for specialist articles to go.

That is not to say that all generalist journals necessarily opted to take advantage of that opportunity. The American Political Science Review did pretty well – certainly through the 1960s and 1970s.Footnote 13 The Journal of Politics did reasonably well, with a few more eccentricities. Other notionally generalist journals remained largely mono-cultural. The re-badged American Journal of Political Science, for example, remained largely devoted to the testing of empirical theories, albeit applying an increasingly broad range of methods to an increasingly broad range of theories.

In so far as it first had to establish a bridgehead in the British academic world, it was inevitable that the initial niche for the BJPolS would be as an empirically-oriented counterpart to Political Studies’ emphasis on political philosophy and old institutionalism. That is what made BJPolS distinctive within Britain, after all, and over four-fifths of the contributors to the first volume of BJPolS were based in Britain.

Although BJPolS did occupy that niche, it was also insistently eclectic from the start. Categorizations are inevitably contestable at the margin. By my count, however, only about half the first volume was given over to political sociology, broadly construed. Fully another quarter was devoted to political theory, with the rest splitting evenly between comparative politics and public policy and administration.Footnote 14 Remember that I said at the outset that the personality of the journal is inevitably partly that of the editor; and the proportion of that first volume devoted to political theory (a proportion never repeated since) doubtless has to do with the fact that Brian Barry served as editor of the BJPolS in its initial year. Still, throughout its early years, BJPolS remained insistently diverse, in differing ways under different editorships.

As the new subdisciplinary journals became better established, they increasingly became the primary loci of conversations within the subdisciplines. Anyone who wanted to communicate with and make an impact on the discipline as a whole, however, still had to publish in the main generalist journals of the disciplines, where their work would be more widely seen. To some extent, most people want to do that. To some extent, everyone is obliged to do that to satisfy tenure committees in the United States and RAE panels in Britain, both of which are anxious to see evidence of a wider impact.

With the new subdisciplinary journals siphoning off a substantial proportion of work in their fields, less was available to the generalist journals. The effect of this varied across the subfields. International relations tended to keep itself to itself, particularly in Britain.Footnote 15 Political theorists and law-and-politics scholars did likewise pretty much everywhere.Footnote 16 In particular, there were other whole discussions – of the American presidency, of legislative organization and behaviour, of local government – that almost never broke out of specialist journals.

It is unsurprising that specialists should want to talk to one another. Hence it is unsurprising that, once outlets dedicated to that have emerged and become established, the bulk of subdisciplinary output should appear there. What is more interesting, in a way, is what happened at generalist journals in the wake of that.

Political scientists know about path dependency.Footnote 17 The Polya urn models nothing better, in my experience, than the pattern of submissions to a journal. People looking for a place to send their next paper will naturally think first of sending it to the journal that has recently published the last good article on broadly the same subject. (‘Clearly, that journal is interested in publishing papers on that topic: after all, it just has done so,’ one says to oneself.) So editors invariably find themselves offered more and more papers on broadly the same topics. And assuming (as is typically but not invariably the case) ‘more papers’ also translates into ‘more good papers’, editors who simply choose the best of what they are offered will invariably end up publishing more and more of the same sort of article that they previously published. It really is like drawing a ball out of the Polya urn and putting that ball and another one of the same colour back into the urn before the next draw.

The most notorious instance of this is widely thought to be the ‘capture’ of the American Political Science Review by rational choice modellers in the 1980s. The capture was less complete than Perestroika rhetoric suggested.Footnote 18 Still, by the editor’s own count, the proportion of APSR devoted to ‘formalization’ rose from 10.9 in the decade beginning 1977 to fully 21.3 per cent in the following decade.Footnote 19 Something similar happened to BJPolS, which over the years has become very heavily devoted to the testing of empirical theories of a relatively homogeneous sort.

Editorial Styles

Not only do journals have their own personalities. They also have their own distinct editorial styles. That manifests itself, in the first instance, procedurally. On this, there is a striking contrast across disciplines.

• — Economists, being economists, often charge submission fees and reward on-time referees by passing on some (usually about half) of that fee to referees.

• — Sociologists, being sociologists, often charge (substantially lower) submission fees, but treat those purely as contributions to the common good of running the journal.

• — Natural scientists, who could not conduct their experiments without huge grants, charge huge fees but just write those off against their grants.

• — Philosophers, being philosophers, often favour the Philosopher-King style of editor, who makes decisions in the great majority of cases without asking the opinion of anyone outside his or her own department.Footnote 20

• — Political scientists, being political scientists, often treat journal submission as a game where winning and losing should be purely a matter of counting the votes of referees.

In that latter respect, however, there is a major divide between ‘association’ journals and ones not tied to any membership organization. Where there are members to answer to, editors tend to be more timid. They are inclined to send more papers out to referees. APSR, for example, used to send virtually all out for refereeing, just so long as the formatting and other formalities were correct;Footnote 21 and tellingly, when APSR editors began rejecting more submissions without refereeing, they did so only with the express consent of the Council of the American Political Science Association.Footnote 22 Association journals also tend to defer heavily to the weight of referee opinion in making editorial decisions. Journals like BJPolS that are not beholden to any membership association enjoy more latitude in both respects. There, the editor is at liberty to (and typically does) pre-screen submissions, rejecting a much larger proportion of submissions that have no realistic hope of publication in that journal without wasting the time of both referees and authors with inevitably futile refereeing. And there, the editor is able to (and typically does) weigh the arguments contained in the referee’s reports rather than just counting their votes when coming to a decision.Footnote 23

Each journal – indeed, each editor – has a distinctive editorial style. Still, some elements of style simply come with the job of editing a generalist journal. The most delightful catalogue comes from the pen of Nelson Polsby, in his ‘Help Wanted’ editorial announcing the search for someone to succeed him as editor of APSR.Footnote 24 Polsby recalls: ‘The last editor of the [APSR] once stated within earshot that the private standard against which he measured potential candidates for the … editor’s job was John L. Lewis’s famous description of John Nance Garner: “a poker-playing, whiskey-drinking, evil old man.” ’ Polsby opines, ‘the tacit criteria he foreshadowed …, if not the overt ones, are not bad guidelines.’Footnote 25

Anyone in a position of power of any sort is morally bound to behave in a responsible fashion, of course. But the burden falls particularly heavily on editors of the main generalist journals, because (as Polsby again observes) their decisions ‘affect the course of so many careers and the shape of intellectual discourse in our discipline’ overall. At the same time, an essential part of an editor’s job description is ‘a certain joyfulness of spirit, a sustaining capacity to have a little fun in the most unpromising of circumstances’.Footnote 26

Many of the characteristics required of would-be editors of generalist journals are obvious, once the desiderata have been stated, as follows:

• — ‘It is important in order to tap the most competent advice available that … editors have reasonably good access to informal networks of communication within the discipline.’Footnote 27

• — They should ‘make an effort, and … succeed, in informing themselves about currents of thought in political science’.Footnote 28

• — ‘Field of specialty within the discipline should count only a little. There is certainly no one mainstream in political science, so the notion of selecting a “mainstream” political scientist, while no doubt more appropriate for the central journal of the profession than the deliberate selection of an intellectual maverick, is in fact a will-o-the-wisp.’Footnote 29

• — ‘It is not quite self-evident that the managing editor should be a scholar of some accomplishment… [since] being a skilled and tasteful appreciator of other people’s scholarship does not entail also being a scholar. The reason … editors must meet a test of scholarship is to confer legitimacy upon their letters of rejection. A letter of rejection from editors who have not personally subjected their work to the same rigorous scrutiny that [APSR] routinely accords its submissions, is a document of questionable persuasiveness.’Footnote 30

• — ‘Successful … editors must also have a well-developed curiosity about what is going on in other parts of the discipline [besides their own], a sense of where the emerging interests are, a sensitivity to cliques and schools of thought, and something more hospitable than tolerance toward work in modes and traditions different from their own.’Footnote 31

That last desideratum might seem particularly demanding in a world of increasing career pressures on scholars to specialize early and narrowly. It neverthless certainly seems true that the most successful editors of generalist journals have been well plugged into multiple subdisciplinary conversations, in addition to those in their own home subdiscipline.Footnote 32 Whether the supply of such polymaths will continue in the future time alone will tell, of course. Remember, however, the pressures towards narrow specialization at ‘separate tables’ is nothing remotely new.Footnote 33 So far, the major generalist journals of the discipline have managed, often enough, to find editors with sufficient intellectual curiosity to resist those pressures. Long may that continue.

One of the major choices that editors of any generalist journal must confront concerns the form their journal’s generalism should take. Is the aim to offer a ‘sample of the best work’ from the several subdisciplines, without regard to whether the samples are of interest to (or even accessible to) scholars in other subdisciplines? Or is the aim, more ambitiously, to launch a ‘genuine conversation across the discipline as a whole’?

Most generalist journals – certainly most that are beholden to some membership association – in effect opt for the first course. They do so out of a sense of ‘fairness to authors’, born of the thought that it is ‘only fair’ that specialist papers be refereed by specialists in their field.

Of course, no one would want to publish something that specialists in the field regard as substandard. Some generalist journals, however, strive not only to sample subdisciplinary excellence but also to help non-specialist readers actually engage with it. Editors of generalist journals who embrace the latter aim will have among their stock of form letters one that says: ‘Thank you for your submission, which seems to me excellent of its kind; but it is simply too narrowly specialist, given the nature of this journal’s readership.’

There are many obvious desiderata and pitfalls for a generalist journal of the latter sort. Demanding that authors situate the problem, explain the methodology, define technical terms – all of those are clearly among them. The largest hurdle to overcome is often more interesting than that, however. Hausman and McPherson describe the ‘clash of disciplinary cultures’ they discovered in editing Economics and Philosophy as follows:

The same probably generalizes to conversations across most disciplines and subdisciplines. That is the real challenge a generalist journal is always facing.

The Future Of A Generalist Journal

So what is the future of the generalist academic journal in the digital age?

On one way of reading the question, the answer is clear and undeniably important, but frankly a little uninteresting. All journals are available online these days, and most of them do increasing proportions of their business (submission, refereeing, production, dissemination) online as well. Doing business electronically suits all concerned: it is cheaper for publishers, it is quicker and (apart from some truly evil journal-management software) more convenient for both authors and editors. Librarians much prefer the electronic version to save shelf space. Users prefer it too, or anyway they would if (like JSTOR) journal backlists were all electronically searchable. So there is no real question whether the digital age will change the way journals do business (it already has) or whether the electronic version will replace hard copy of journals (it will before long, where it has not already).

There is a little (but methinks only a little) more of a question whether commercial publishers will be displaced in the process. Certainly, it is true that one major thing we used to need publishers to do for us – to distribute the journal – has ceased to be an issue with the advent of online delivery.

There is no reason, in principle, that editors could not do everything publishers currently do (including typesetting) themselves, subcontracting out various tasks just as a commercial publisher would. The question is just where the money would come from to pay for all the professional services – for editorial interns, copyeditors, proofreaders, typesetters, website managers and so on – if not from subscriptions.Footnote 35

The only real alternative much mooted is to collect submission or publication fees from authors. That may be fine in a discipline where everyone with anything worth publishing will have a research grant to which they can charge it. In political science, that is not so. Or, worse, that is true of some but not all members of the profession. Such a policy would have a differential impact, probably most disadvantaging people at the beginning of their careers. Given a choice between soaking university libraries or soaking scholars just starting out, the former must surely be the professionally more responsible course.

The third and most interesting question is whether the digital age offers some alternative to current editorial operations, other than in the purely mechanical ways canvassed at the outset. The thought that is sometimes floated is that dispersed networks of scholars, posting their own work on the web and recommending to one another good postings from others that they have come across, might take the place of current refereeing processes. In place of hierarchical systems with editors in control, there would on this scenario be a non-hierarchical network of scholars who respect one another’s opinion ‘doing it for themselves’.

Anonymity of the current refereeing process would be lost in this way. So one threshold question here is just how important anonymity is, and why. There is the obvious issue of fairness, of course. The real importance of anonymity from a profession’s point of view, however, is that it ensures that submissions are judged on their merits rather than on the notoriety of their authors. That advantage would be lost if journals were replaced purely by networks of scholars creating links to papers they liked on one another’s home pages.Footnote 36

Online cross-links are already occurring, of course, more in some subdisciplines than in others. That, however, is precisely my point. That sort of thing can happen more easily at the level of subdisciplines. There, networks are smaller and tighter. Maybe it is no longer literally the case that everyone knows everyone within the same subdiscipline, as once it might have been. Nevertheless, most people at the top know one another, or anyway know of one another. You have a pretty strong sense of whose taste you trust, whose recommendations you can rely upon. Thus, there seems to be a fair chance that, within (at least some) subdisciplines, that networks of scholars ‘doing it for themselves’ might take the place of formally published journals.

There is much less prospect of that happening at the level of the discipline as a whole, with networks and networks-of-networks supplanting the main generalist journals of the discipline. It is not merely that there are too many people. It is more of a matter of there being too many topics, too many discrete groups of scholars working in substantially autonomous corners of the discipline. Someone whose job it is – the editor of a generalist journal – can make enquiries, establish contact, and build bridgeheads to these semi-autonomous provinces. Yet it simply has to be made someone’s job to orchestrate a network of disconnected networks. It will not happen automatically. These networks are not naturally linked all that closely to one another. The degrees of separation are just too great. Thus, while distributed networks might supplant specialist journals, they probably will never supplant generalist ones.

Generalist journals are likely to survive in their current form for one last reason, which has to do with their credentializing function. As already noted, tenure committees and research assessment panels invariably want evidence that individuals’ work is making an impact on the discipline as a whole. For evidence of that, they look to how frequently those individuals publish in the main generalist journals of the discipline. And the reason that that counts as persuasive evidence is that there are a strictly limited number of slots in each of those journals, determined by the ‘page budget’ imposed by their publishers.

A ‘page budget’ is a print-based concept, to be sure. As even commercially published journals move increasingly to electronic delivery, the pressure to keep the size of the journal down to any fixed number of pages is reduced. Reduced, but not eliminated: copyediting, typesetting and proofing costs, as well as the costs of putting each article on-line, all remain a function of the number of pages (and hence articles). Furthermore, so long as there is a print edition at all, there are all those other printing and distribution costs that vary with the size of the physical product. There is no such ‘hard budget constraint’ on the number of articles that might be commended by a network-based group of scholars recommending links to one another.Footnote 37

There would be no one, in a non-hierarchical network-based system, who would be in a position to ‘reject’ a submission. Instead, in the network-based model, everyone would simply ignore it (although everyone would have had to waste a certain amount of time establishing, to their own satisfaction, it was ignorable). Worse, there would be no reason, in the network-based model, to reject a submission. There is no page budget, there are no limits (within huge limits) as to how large the list of recommended readings could be. So there is no incentive to say ‘no’.

Thus, my prediction would be an inverse Tragedy of the Commons, as the common pool of recommended readings is flooded with far too many pieces that someone or another has recommended as being ‘worth reading’, and no one has authority or incentive to say ‘no, not good enough’.Footnote 38 Current journal editors all have in their file a stock letter reading, ‘That is a nice piece but it just does not make a sufficient advance over the existing literature to merit publication given the other articles competing for a place in the journal’s scarce pages’. In the network-based model, there would be no such scarcity – and indeed no one with authority to send such a letter.

That would have two consequences. One, already noted, is that everyone would have to waste a huge amount of time sifting through the recommended papers to determine for themselves which are really worth reading. In the journal-based model, that sifting is done for readers by editors and referees (with complaints about referees being increasingly overloaded, as is). In the network-based model, it would be a case of ‘every reader his/her own referee’ – or, perhaps worse, a case of ‘follow the leader’ where the leaders know no more than you.Footnote 39

It is the second consequence that will, I think, prevent this from ever coming to pass. Tenure committees and research assessment teams, looking for markers of contribution to the wider discipline, would not be able to infer anything in particular from the fact that someone in the network recommended someone’s article be included in a repository that has no natural limits. Without some filtering process, the credentializing function is lost: and that is something everyone in the profession has an interest (of the basest careerist sort) in preserving and protecting.

Here again, the problem could be solved by an internet repository that was space-limited and tightly refereed. But again, that just amounts to replicating journals by recreating the editorial procedures that I take it define a journal. Self-organizing networks of scholars could in principle do that for themselves, instead of waiting for commercial publishers or professional associations to invite them to do so. Yet what they would be creating is, I submit, a highly organized, structured, hierarchical system of scholarly assessment that is extensionally equivalent to a ‘scholarly journal’ of the current form.