I was recently invited to give a talk as part of a lecture series titled The Open Access Future, sponsored by the Smithsonian Libraries, the Smithsonian Institution Archives, and the Smithsonian’s Office of the Chief Information Officer. I decided to focus on the issue that has been troubling me most lately: why is it so hard to have conversations about OA that don’t devolve into shouting matches and accusations of bad faith? What has led to this state of affairs, how bad is the problem now, and what can we do to create a more open, inclusive, and reasonable environment for discussion of the complex issues surrounding OA and the economics of scholarly communication generally? I came up with a provocative title (“Is a Rational Discussion of Open Access Possible?”) and delivered the lecture on March 10, 2014. Here’s the archived video of the lecture and of the discussion that followed. (The full text and accompanying slides are also available; comments are welcome both there and here. However, I will not be monitoring or responding to any related Twitter traffic, having learned from experience how difficult and frustrating it is to carry on meaningful discussion of complicated topics in 140-character bursts.)
47 Thoughts on "Rick Anderson at the Smithsonian: "Is a Rational Discussion of Open Access Possible?""
Just thought I’d chip in here to say that I watched the video, and found it well worth my time. To others who are inclined to skip it, I’d encourage you to give it a go. The actual talk is about 30 minutes, and the rest is (interesting) Q&As.
What the hell is wrong with you Mike? It’s like you think we can have a sound discussion about OA at the kitchen or something?
Considering that Rick at one point in the Q&A refers specifically to Mike as one of the OA advocates he’s implicitly criticizing, this is a generous response by Mike.
That’s actually not quite accurate. In the Q&A, I referred to Mike as someone who would certainly disagree with me that “no one would be saying the scholcomm system is broken if all science journals cost $15” etc. No criticism was implied there, just the expectation of disagreement.
On the other hand, criticism of Mike was implied in the link I provided to his reaction to the Bohannon sting. (You’ll find that link in the lecture text, though obviously not in the lecture video.) I provided that link as one in a string of references to reactions that I felt were inappropriately negative.
Either way, Mike’s response here is still generous. 🙂
No worries at all — I was actually pleased to be called out by name at that point in the Q&As, because it was a welcome confirmation that my actual position on this is properly understood (at least by Rick). I’ll avoid discussing the Bohannon article here, since that’s a bit off the actual topic, beyond saying that I stand by what I wrote in the article that Rick linked, and that I doubt Rick himself would find much to disagree with in its substance (as opposed to its tone).
here is something perhaps of interest to you re open-access publishing:
Tötösy de Zepetnek, Steven, and Joshua Jia. “Electronic Journals, Prestige, and the Economics of Academic Journal Publishing.” CLCWeb: Comparative Literature and Culture 16.1 (2014): http://dx.doi.org/10.7771/1481-4374.2426
with best regards,
totosy de zepetnek, steven phd professor
purdue university & purdue university press
I’ll start with the disclosure that I’m one of two Editors-in-Chief at PLOS Genetics, an open access journal, so apply whatever grain of salt you like to my comments. I like Rick’s call for rational discussion. That will certainly benefit all stakeholders. The point that most interested me in your talk was the critique that OA necessarily creates a perverse incentive that degrades quality because the end product isn’t sold (my paraphrasing). That’s an insightful observation but I wonder if it assumes that sale of product is the only (or at least the primary) currency that incentivizes quality through market pressure. Academic publishing may be more complex because there are at least two important currencies. There’s the traditional monetary currency, but there is also (for lack of a better term) the “currency of prestige”. Market pressure to maintain quality is exerted on OA journals because academic researchers want to publish their work in high quality venues. This preference stems from real-world pressures because tenure, promotion and compensation decisions typically use quality of scholarly output (as measured by publication outlet) as a metric. For most small to midscale OA journals (ignoring “mega-journals” which have their own unique economy) lowering their quality means sacrificing their customer base of academic researchers.
One final comment – at the end of your talk an audience member asked about high impact OA publications that use rigorous peer-review. You mentioned PLOS ONE. I just wanted to point out that PLOS is a family of journals (that all use rigorous peer review) and have the following (2012) impact factors (with the typical recognition that it is an imperfect metric):
PLOS Medicine: 15.253
PLO Biology: 12.690
PLOS Genetics: 8.517
PLOS Pathogens: 8.136
PLOS Computational Biology: 4.867
PLOS Neglected Tropical Diseases: 4.569
PLOS ONE: 3.730
The point that most interested me in your talk was the critique that OA necessarily creates a perverse incentive that degrades quality because the end product isn’t sold (my paraphrasing). That’s an insightful observation but I wonder if it assumes that sale of product is the only (or at least the primary) currency that incentivizes quality through market pressure.
It’s certainly not the only incentivizer, and that’s why I was careful generally to speak in terms of quality incentives “weakening” rather than disappearing with OA models. The PLOS family of journals (thanks for providing the IF data on those, by the way) provides one example of how an OA product’s quality can continue to be strong despite weakened market incentives; unfortunately, the output of many other OA journals (especially those from “predatory” publishers) provides evidence for the reality of the incentive problem. This is certainly not a binary issue — it’s not like journals are either expensive or lousy. But market incentives do matter.
It must also be noted that the dominance of market considerations can have adverse (perverse?) effects on quality. That has certainly been a trend in scholarly book publishing where decisions about what to publish have increasingly been influenced by market, rather than purely scholarly, criteria.
You’re right that “what people are willing to pay for” doesn’t always correlate perfectly (or even well) with “what is of the highest quality.” But it’s important not to jump from that observation to the conclusion that, when it comes to scholarship, quality and demand have an inverse relationship. In this context, relevance is a very important intervening variable. The greatest book in the world on shoelace aglets is going to sell relatively few copies, simply because there are relatively few book buyers in the world with an interest in that topic. Does that mean market forces have an adverse impact on the quality of books on shoelace aglets? No — it means that some markets are small.
There is a strong correlation between quality and sale price in most commercial exchanges, but there are also notable exceptions. Many people are willing to pay large premiums for a Rolex watch even though there are alternatives with equivalent mechanical and aesthetic quality. These consumers act on perceived added value that comes from the prestige of owning a Rolex. A similar (but not entirely parallel) force exists in academic publishing. Cost is not the only driver, the social currency of academic prestige (with all the real-world implications) distorts the quality-price correlation. That distortion can enforce quality in the absence of sale price pressure. There are predatory elements in almost all markets, but they don’t necessarily undermine the market rationale of the fair players. Extending the earlier metaphor, the guys in trench coats lurking in alleyways saying “hey, wanna buy a Rolex” don’t undercut the market rationale of genuine Rolex or their well-made competitors. Similarly, predatory OA practices shouldn’t necessarily be used to impugn high-quality OA publishers. The bottom line is that the publishing ecology has enough different species that broad brushstroke assessments often fall short on closer inspection.
Similarly, predatory OA practices shouldn’t necessarily be used to impugn high-quality OA publishers.
This is a point on which I very much agree with the OA advocacy community. My problem with their response to Bohannon, however, is that in too many cases I see them using this valid and important point as a smokescreen to distract people from another valid and important point about the perverse incentives created by author-pays publishing. Yes, it’s very important to bear in mind that most OA publishers are not predators. It’s also very important to bear in mind that predatory OA publishing is an expression of a structural flaw in the Gold OA model. Sweeping either one of those truths under the rug hinders rational discussion of OA.
Rick, this is a fine argument for reasoned discussion but I was disappointed that you did not directly address your title question. Depending on how that question is interpreted I think the answer is either yes or no. Is some rational discussion possible in some venues? Yes, it is occurring already. Is a disruptive social and political movement like OA possible without the various features that you complain of? Probably not, at least I cannot recall ever seeing one and I have been studying these issue driven movements for many years.
These nasty features occur for good reasons, alas. Policy making is intrinsically an advocacy business, which means each side pushes a one sided case. When the issue is deemed to be moral, moral arguments must be made. I note that at one point you complain of something like war propaganda, but later you characterize the OA case as a war. In fact I would describe your essay as an advocacy position, namely advocating for the middle ground. There are other positions in this policy debate, each of which hopes to win. To that end you might consider the downside of rational discussion, which is that certain positions lose.
One of the most interesting thing about these debates is that debate about the debate is still part of the debate. This is a major source of confusion.
Rick, this is a fine argument for reasoned discussion but I was disappointed that you did not directly address your title question.
I was trying to do that in the final section of my talk, in which I proposed six principles for “critical and constructive discussion of OA.” (By suggesting those principles I was implying that the answer to my title’s question is “Yes, if…”)
In fact I would describe your essay as an advocacy position, namely advocating for the middle ground.
I hope I didn’t give most readers/listeners that impression, since it’s emphatically not my position. My position is not that we should all converge on the middle ground, but that we need to be able to discuss these issues critically, openly, candidly, and respectfully, keeping our minds open to both the upsides and the downsides of any proposal or program. The result of a conversation built on that framework may, in any given case, be a policy or program that represents a middle-ground compromise, or one that represents a lopsided “win” for one side or the other. I don’t think there’s anything inherently wrong with any of those outcomes. But I think it matters very much whether relevant information is excluded because one side or the other doesn’t want it to be addressed, or stakeholders are excluded from the conversation because they’re regarded as morally unworthy, or arguments are misrepresented during the discussion in order to make them easier to refute. In other words, I don’t necessarily see the inevitability of one side losing as a downside of rational debate. Compromises and win/win aren’t always either possible or desirable — sometimes everyone will be better off if one side loses.
One of the most interesting thing about these debates is that debate about the debate is still part of the debate. This is a major source of confusion.
In the case of the OA debate, I don’t think confusion is the central problem (though it certainly can be one problem). I think the central problem is that we have different parties who want mutually exclusive things.
perhaps after you all read “Electronic Journals, Prestige, and the Economics of Academic Journal Publishing.” CLCWeb: Comparative Literature and Culture 16.1 (2014): http://dx.doi.org/10.7771/1481-4374.2426 we can have a discussion?
I think the central problem is that we have different parties who want mutually exclusive things.
This unpalatable truth is at the heart of the conflict. It’s nice to imagine that if we can only find the right way to look at things, we can find a solution that satisfies everyone. But we won’t; it can’t be done. The disagreements go much deeper than that.
Doesn’t mean we can’t be civilised in arguing our way through it, of course.
Rick, you use the term excluded but no person or argument is being excluded. Objection is not exclusion. Policy making is an advocacy process and what you are objecting to is normal, so probably deeply reasonable. Advocacy has it’s own norms and standards. Moral arguments, questioning motives, and objecting to lines of argument are all part of it.
As for positions, there are several that claim that further discussion is unwarranted so your call for discussion is an opposing position. For example those who want immediate strong action will see it as a delaying tactic.
If you see the debate per se as a problem then of course that is the biggest problem. I was referring to a specific kind of confusion as a problem within the debate. As an issue analyst I study the debate.
Just to note that “immediate strong action” (which I favour) does not exclude continuing discussion (which I also favour).
In public policy this is not true. Once a law is passed the discussion of what it should say is over.
All laws are subject to revision. Discussion is not over at that point: it simply restarts with a different set of premises.
This is true in principle, and perhaps in academic circles where discussion never ceases, but the time frame for revision is typically 10 to 20 years, if ever. At the Congressional level once a law is passed the discussion ceases for a long time. For example they just passed an OA law for several agencies. Those agencies no longer have the flexibility that the OSTP memo allows. For them the only thing to discuss is how to implement the law.
Wonderful speech! So glad you posted it here. I am looking forward to seeing the comments on the blog. I think you very fairly echoed what Alice Meadows was getting at a few weeks back that some in the advocacy world are damaging their own efforts by refusing to engage in any discussion. The “my way or the highway” approach rarely works and they are grossly exaggerating anything that can be perceived as a small success.
Although it’s understandable why Rick would focus his talk exclusively on OA as it applies to scholarly journal publishing, that has been part of the problem with the OA debate, viz., the reluctance to move the debate beyond journals to books, which I tried to do in drafting the AAUP’s Statement on Open Access in 2007. Rick did have an opportunity to widen the discussion to books when one of the questioners asked about the Smithsonian’s own publishing program under its new director, but Rick apparently wasn’t familiar with the long history of the Smithsonian’s publishing through its own press: http://www.scholarlypress.si.edu. At any rate, one of my frustrations in discussing OA rationally with other OA advocates is that some of them insist that only publishing that uses the CC-BY license qualifies as “true” OA. Using such a stipulative definition has the unfortunate consequence of ruling out most of what is done as OA by book publishers since they generally use the CC-BY-NC-ND license instead, for good reasons. Rick doesn’t note this aspect of the debate in his talk, but it has been a particularly annoying one to me. I should also credit Rick with being a good debater in dialogues we have had over various issues in scholarly communication, such as PDA, where our e-mail debate on Liblicense got formally published as a “dialogue” in an issue of Against the Grain.
Thanks for the kind words, Sandy. As for books and CC licensing: I would have liked to address both of those topics (especially the latter, which I think is particularly important and fraught), but I was working within serious time constraints and had to focus on the area that has been most central to the conversation. (I realize that part of your point is that you feel books should be more central to that conversation than they have been, but for better or for worse, they haven’t been. I think a big reason for that is that a book represents such a tiny expenditure as compared to a journal subscription. The fiscal crisis is by no means the only thing at issue here, but it’s a major driver of the discussion.)
But, as you well know, monograph purchasing has suffered from the economic crisis in journal publishing, so the two are closely interlinked, and OA for book publishing has been one attempt to respond to this situation. It is also this interlinkage that has exacerbated the problem of market criteria playing a more important role in monograph publishing than pure quality; as library sales have declined, presses have looked to a wider market for sales, and that concern for a wider market has muted the influence of purely scholarly criteria in deciding what gets published. Part of the virtue of OA for book publishing is that it can move decisionmaking back in the other direction.
There is an interesting logical paradox with the ad hominem argument. It is indeed a fallacy in deductive logic but to consider the speaker is often valid in inductive logic, especially when the speaker is making self interested claims. For example, used car sales people. Advocates too, for advocates often exaggerate, and this is the form the argument often takes in policy debates, where it is not a fallacy. It is not that the claim must be false, just that there is a reasonable possibility that it is false, given the speaker. It is a buyer beware argument.
I’m sorry to be obtuse, but can you unpack that a bit?
I can try but it might help if I knew what part you were questioning. It goes like this. The ad hominem fallacy is when one argues that what someone says must be false because of who they are. This is a deductive argument, hence the “must be.” But in inductive logic, or the logic of evidence, who the speaker is, is often an important consideration, one that may argue for skepticism about what they say.
The classic case is when the speaker has a financial interest in their statements being accepted, such as someone trying to sell you something. But that the speaker is an advocate presenting their cause is also grounds for increased skepticism, because they too have a personal interest in having their statements accepted. In fact it is the nature of advocacy to present one sided arguments. The trial lawyer is a clear case of this.
This advocacy case often arises in public policy debates because the policy making process is an advocacy process. Characterizing calls for skepticism in such cases as ad hominem fallacies is therefore incorrect. This confusion occurs frequently.
Here is a good general book about fallacies: http://www.psupress.org/books/titles/0-271-01416-4.html.
Congratulations on a very cogent and even-handed deconstruction of the open access debate.
I really enjoyed your explanation of the ‘things that cost money’. I sometimes feel that one of the things missing from the debate is an acknowledgment that publishers really do contribute value to the process. This is partly due to a failure on the part of publishers to adequately explain that value. You mentioned that making documents available is now not as costly as it was. I’m not an expert in the economics of digital versus paper dissemination but when talking to many publishers, I often get the impression that the amount of effort that goes into not only creating and maintaining their platforms, but ensuring global, fast access to that content under often very high traffic loads is quite substantial. What challenges are involved, for instance, in making sure that your content can be read by academics in China? I wonder if one set of costs hash’t just been replaced with another.
There’s also the question of discoverability, which has become more of a challenge, I think in the age of web-based publishing. Putting an article up on wordpress may make it available but it’s certainly not discoverable. Publishers have to coordinate with a raft of abstract databases, discovery layer providers, link resolvers and databases. The publishing industry has traditionally done a pretty good job of facilitating at least some of that, with efforts like crossref. With Green OA, I feel like discoverability becomes more problematic. Currently, if a researcher’s only legal access pathway to an article is the institutional repository of one of the co-authors, the likelihood of them finding that instance is pretty small. I hope that either CHORUS, SHARE, or both help alleviate that problem, but the fact that these two collaborations exist really points to the fact that discoverability is a issue in green OA publishing.
The last point I want to make is on the other the false dichotomy; the danger that author-pays models might cause a race to the bottom in terms of editorial standards. Of course, you’re absolutely correct that predatory OA journals are a problem, but I don’t really agree that OA publishers necessarily have a perverse incentive here. One of the most important factors that goes into where a researcher chooses to publish is the reputation of the journal. A journal that is stringent in it’s acceptance policies would presumably gain a better reputation and be able to charge more for it’s publication services. On the other hand, there’s also a risk of a perverse incentive for toll publishers to accept more papers in order to create spin-off journals that they can charge subscriptions or to consider author charges as a revenue source and have double-dip business models. I think that irrespective of business model, there will always be predatory companies operating on the margins of any industry, but for the most part, most publishers are trying to honestly fill market need, even if they sometimes get it wrong.
“Putting an article up on wordpress may make it available but it’s certainly not discoverable.”
I’m surprised you’d say that. I would have said the exact opposite. Third-party search engines can and do give excellent results. Google for, say “sauropod kebab”, “rabbit neck posture” or “xenoposeidon week” and the top hits are all for our palaeontology-themed WordPress blog. There are many, many other such searches. I would say that discoverability is much closer to a solved problem now than it was ten or even five years ago.
You make a good point about Google. when I wrote that, I was thinking from the perspective of a researcher looking for a specific article, or looking for content inside a discovery service. Maybe I was unusual but as a researcher, I would mostly use an abstract database or a reference in an article that I’ve already read to identify a citation. Then I’d have to go find that in the maze of possible locations.
Perhaps if a researcher copies and pastes the title of an article into google, they might look far enough down the list of links to get a free WordPress instance of it, but that instance is competing with a host of links to instances of the full-text and the abstract, that are more cleverly SEO’d. The other place you might look for the appropriate, legal copy is a discovery layer or link resolver, which don’t have blog links in them. In fact, they don’t even have repository links in them yet.
You could simply publish on WordPress without going through a publisher and getting the work certified and I think that’s what you’re thinking about. My concern about that isn’t really discovery so much as certification. I feel that, unless you are already famous as a researcher and have clout, content published in non-peer-reviewed format wouldn’t be taken seriously by most researchers. It may be seen in a general google search for a topic but would be in danger of being passed over. I know that I run the risk of sounding dismissive here, so please forgive me, but I can’t imagine post-docs, junior or tenure track faculty thinking of self-publishing in a blog as a viable alternative.
Well, I didn’t really mean that publishing in a blog is a substitute for more formal routes — not yet, anyway! We’ve agonised over this issue on our own blog quite a bit, but never reached a satisfactory conclusion. The best we can do is to say that things are in flux, and that the degree of respect that a scholarly blog-post is held in varies a lot in different places, but is never equal to what a “proper” published paper achieves.
Still, my experience has been — to my surprise — that Google and other such third-party crawler-based engines are the best discovery tools for published papers as well as blog-posts. They dig out publicly posted PDFs, repository copies and suchlike, in a way that (you rightly note) more academia-anointed discovery systems do not.
Google Scholar is very good indeed when it comes to journal articles. Their more-like-this feature is great. I use it in an algorithm that finds all and only those papers related to a given topic and ranks the papers by closeness to the core topic (called the inner circle method). Their repository finding feature is also getting good, as you note.
Well, actually it is, Mike, in some fields like economics and law. Some senior economists have gotten fed up with the time it takes to get their papers formally published, so have opted instead just to self-publish, relying on their own reputations as “brands” to attract readers. And since law reviews do not use peer review anyway, some well-known law professors have taken to blogging and their blogs are recognized as disseminating some of the best legal writing available today.
Interesting! You can certainly imagine something similar happening at the better-known universities.
I use wordpress to publish my own work. I cite my blog in my more formally published work and vice versa. The immediacy of self-publication allows me to do things I could not do in any other way; it’s certainly a very good way of achieving scientific communication — I do get feedback on my work from colleagues. I think it should be a much more heavily adopted publication route.
In terms of discoverability, my best read blog articles have page reads similar to some of my papers; whether this says more about my blog or my papers I leave others to decide.
What really amazes me is how easy it is; compare the self-publication route — whether blog, arxiv or whatever. It’s quick, simple and cheap. My experience of formal publication is quite the opposite. I don’t know why.
Thanks for these thoughts, Phill. A couple of responses:
You mentioned that making documents available is now not as costly as it was. I’m not an expert in the economics of digital versus paper dissemination but when talking to many publishers, I often get the impression that the amount of effort that goes into not only creating and maintaining their platforms, but ensuring global, fast access to that content under often very high traffic loads is quite substantial.
For publishers who are disseminating content in the manner of publishers, you’re right — the cost of creating and maintaining strong and durable platforms can be quite high. But for an author who simply wants to make a paper freely available to the world, the costs will usually be negligible: set up a free account on WordPress or some other freemium blog platform, copy and paste the paper’s content in, and there you go. (That’s just what I’ve done with my lecture text.) For many authors, this isn’t enough, of course, and for some papers this approach will work better than others — and that’s one major reason that publishers exist.
The last point I want to make is on the other the false dichotomy; the danger that author-pays models might cause a race to the bottom in terms of editorial standards. One of the most important factors that goes into where a researcher chooses to publish is the reputation of the journal. A journal that is stringent in it’s acceptance policies would presumably gain a better reputation and be able to charge more for it’s publication services.
This is an important point. In addition to perverse incentives (to accept as many papers as possible and thereby get as much APC revenue as possible), Gold OA does also carry with it a virtuous incentive (to attract paying authors by providing rigorous review and therefore prestige). But while that virtuous incentive exists, I think it’s strongly overpowered by the perverse one. Here’s why: offering false prestige is relatively cheap and easy — all you need is to create a new journal with a prestigious-sounding name and an editorial board populated (or at least claiming to be populated) by reputable scholars. Promise rigorous peer review. Claim to have a high impact factor, even if you haven’t existed long enough to have an impact factor at all. Advertise your journal to a large population of authors that will lose their jobs unless they publish a certain number of articles in peer-reviewed journals, take their money, accept everything they submit, and then, as the reputation of your journal declines (or when you get caught by Jeffrey Beall), shut down the journal and open a new one. On the other hand, actually establishing a rigorously-reviewed journal and building a reputation for it based on rigor and quality is very expensive and takes a lot of time. So while there are arguably real incentives to do the latter, I don’t think it’s surprising that we’re seeing so very much of the former. The problem isn’t that the marketplace provides no incentives to be a good publisher; the problem, I think, is that the rewards for being a scam OA publisher are immediate and rich, whereas the rewards for being a good publisher (under any model) are relatively modest and slow-coming.
Thanks for the response, Rick.
I see your distinction better now between the costs that publishers absorb to support their platforms and the minimal cost that throwing up a web-page on WordPress represents.
My only other thought is that putting something up on a blogging platform doesn’t guarantee global delivery and availability of the article. I mentioned China in my comment. I wrote it as if it were throwaway line but it’s a good example. China is of increasing importance scientifically and WordPress is blocked by the Chinese government. I don’t know if I’m over-stating the value of it but monitoring and ensuring global delivery is part of what I think of as the delivery service that publishers provide.
My only other thought is that putting something up on a blogging platform doesn’t guarantee global delivery and availability of the article.
That’s true, and a good point — but it usually does guarantee much broader and more immediate access than what a publisher will provide. (At the cost, of course, of the added value that a publisher also usually provides.)
I would like to make a point about inclusion, rather than exclusion. The Scholarly Kitchen has featured extensive and lengthy discussions of open access among publishers, librarians, OA advocates and many other voices. In fact to date there have been 540 articles that include the term open access in the text (often in the title, which shows that OA is the focus of the article) and many of which feature numerous comments. While some points are made frequently, most of this vast discussion is non-redundant. We have explored a lot of sub-issues in great detail. Clearly OA is a very complex issue, which cannot be simply summarized, and that I think is the real problem.