Depending on the audience, the case for open access (OA) varies. Opponents of intellectual property, for example, may favor OA simply on principle. To a researcher you might argue for a broader dissemination of his or her own work. A funding agency may accept the dissemination premise as well and tie it to an exercise in branding, where each published OA article becomes an ambassador for the sponsor of the research. A librarian may be persuaded on the basis of cost (the money that does not have to be spent on OA material can be used elsewhere), and an established publisher may see the Gold variety simply as an additive revenue stream.

I don’t intend to navigate through the merits of any of these arguments here; from my point of view, OA is a fact, not a philosophy. What has caught my attention is how often I am asked to justify a recommendation that a professional society look into the creation of an OA service. In my experience (which I do not pretend to be comprehensive), OA is a tough sell. The officers of professional societies are often skeptical about the claims for OA and need to be persuaded that it is good for science and good for their own particular society. If you are going to try to make a case to a group of accomplished individuals whose stock in trade is to study all assertions with great care, you have to toss the abstractions out the window and bring some facts to bear on the situation.

This exercise is made more difficult by the fact that no professional group speaks with a single voice. I have never run across a group that has no OA advocates; what differs from society to society is density–what proportion of the society members are in favor of OA–and the relative position of the advocates within the society’s power structure. Thus a large part of the challenge in introducing OA to a professional society is the need to “socialize” the recommendation–deliver it in a way that provides fodder for discussion but does not presume to impinge on the prerogatives of the practicing professionals. No two societies make decisions the same way; there is no way of predicting in advance the outcome of a board discussion or who will or will not speak up on a conference call. The case for any recommendation, whether about OA or anything else, is best made with a written document that creates context for discussion. PowerPoint won’t do it, as there is a need to work through all the options at some length. This document must be shared with the participants well in advance of a meeting. No one who creates such a document should expect to come through the deliberations of such a board unscathed.

Contrary to popular belief (if the measure is the blogosphere and the runic declarations of various organizations active in or observers of scholarly communications), many scientists don’t know much about OA. In particular, they don’t know about the practical implications of the growing OA publishing ecosystem:  how many OA articles are published and where, and what does this mean or could come to mean for the society’s own publications? If these publications are continuing to see an increase in the number of submissions every year, as many of them are, OA is likely to seem peripheral. A rising tide is of little consequence if you are sailing on a sturdy ship.

One particularly useful exercise is to suggest that the society review and catalog all the OA articles published in the past year that are relevant to the society’s discipline. This is a tall order, of course, as the number of OA articles is very large and defining relevance is not without ambiguity, but a useful way to begin is simply to analyze the publications of PLOS ONE, whose sheer size provides a healthy sample (depending on discipline, of course). Many societies are surprised to discover how many articles have appeared in PLOS ONE that could have been published in the societies’ own journals, assuming they passed muster under peer review. It is not unusual for a society to “wake up” to OA when it realizes the sheer range of articles that their own editors never got to see.  Such an analysis could lead a society to change the editorial positioning of its own journals, to start a new journal, to create an OA service–or any and all of the above.

A more common exercise (I don’t know any journal that has not done this) is to study the journal’s own publications to see which articles are subject to OA mandates. Increasingly traditional journals are having their “Swiss cheese” moment, where some articles are mandated to be OA (perhaps with an embargo) and some are safely tucked behind a pay wall. But it is not enough to stop there by identifying the scope of the mandates. It’s useful to go back in time to some of the journal’s most-cited papers. When we consider the funding sources for these articles retrospectively, how many of these articles would now fall under an OA mandate? Not having an OA option (at a minimum, a hybrid OA policy) could mean that some of the finest scholarship will bypass the society’s publications and find a home elsewhere. This puts the journal’s reputation on the line, which grabs the attention of the society’s editors and officers very quickly.

The question I personally find most interesting is where a rejected paper eventually ends up. While we know that some authors become desperate and seek publication with publishers of dubious merit (whose exposure by Jeffrey Beall we all should be thankful for), most authors shop around until they find a suitable venue. That venue may not be as prestigious as the first one, but it also may be more appropriate in terms of its topic, something that only the journal’s editors can determine. Are these papers good papers? Good-enough papers? Good papers, but with the wrong editorial profile? Bad papers? Except for the truly bad papers, should there not be some publishing outlet for this work? And considering the time and expense a society puts into rejecting a paper, does it not seem practical to find a way to monetize the sunk editorial costs? Once a society begins to consider that publication is more than a question of good versus bad, accepted articles versus rejected articles, the case for an OA service appears to be more reasonable.

Discussions of OA policy rarely take place in the abstract. It’s common for the debate to arise when the society is facing marketplace challenges, the most common of which is declining institutional subscriptions, typically brought on when libraries opt for “big deal” packages from the largest publishers, leaving small society publishers to pick up the crumbs. This is more than a financial matter, though the money can be significant, especially for a society that uses the surplus from its publishing program to underwrite other activities. The real risk is that declining subscriptions may mean that authors’ work cannot reliably be brought to the attention of other people working in the field who happen to be affiliated with an institution that no longer subscribes to the publication. It’s worth remembering that most research is conducted at a small number of institutions, which subscribe to the leading publications: researchers affiliated with these institutions receive little or no benefit from OA publishing. But if an institution cancels a subscription, then OA will begin to seem like a useful alternative. So a vicious cycle begins: cancelled subscriptions mean that some researchers no longer have access to the publications they need, which discourages them from submitting their work to those publications, which lowers the perceived editorial quality of the journal, which makes it easier to cancel the journal if the library budget is tight, which means that even fewer researchers have access to the content, which further discourages them from submitting papers to that journal, and so on. A journal in such a situation may be faced with the choice of partnering with a very large publisher or attempting to flip their publication to Gold OA as a means of preserving the journal’s independence.

What doesn’t work in these discussions is the invocation of the histrionics of OA advocacy. Of course the commercial publishers are greedy capitalists; of course there is an unnamed researcher in the developing world who wants to get access to research materials; of course it is simply unjust that taxpayers should fund research and not get to see the results at no additional cost; of course there are laypeople everywhere who wish to keep abreast of breaking scientific developments–of course. But the leaders of a professional society may not feel that the fight against the world economic order is within their writ or that finding a readership beyond the walls of the small number of institutions that create and consume the majority of research is of paramount importance to them. There is no point in whistling a tune if no one wants to dance. What is important is attracting the best researchers, the prestige of the society and its publications, and the economic viability of the society itself. Speak to these issues and OA gets on the agenda, and not as a Trojan horse but as a practical extension of the society’s activities.