Wednesday, June 06, 2007

Open access.

I've blogged about this once before, but Michael Geist is at it again. Here he gives two proposals for enhancing "open access" to research, one sensible.

The sensible suggestion is that "raw, scientific data currently under [government] control" should be available freely. That makes a certain amount of sense to me. When public agencies generate data -- say (his example), Natural Resources Canada creating topographic maps -- then there's no good argument that I can see for not making the data publicly available.

Here's the other suggestion: "Ottawa must pressure the three federal research granting institutions to build open access requirements into their research mandates." It's not totally clear what he means by this. It could mean that NSERC, SSHRC and CIHR should encourage (whatever "encourage" means in this context) grant recipients to publish in open-access journals. This would create an annoying hurdle for those, like me, who aren't big fans of across-the-board open access, but there's already enough annoying hurdles in the grant process that one more won't make a significant difference. So, on the whole, that would be a rather benign circumstance.

More dangerously, though, it could also mean that grant recipients are required to publish in open-access journals. There's two problems with that as a policy. First, cutting off access to traditional journals and the like, at this stage of the game, shoots the value of the publication to hell. Indeed, why bother applying for a grant when its eventual product will have to be published in some journal no-one's heard of and no-one reads? Keep in mind that the point of publication, at least in the early stages of an academic career, is tenure, and low-profile publications matter less than high-profile ones.

Second, it misconstrues the relationship between researchers and grant-giving agencies. Researchers don't work for the grant agencies. If I'm hired as a researcher by some private institution, then there's a prima facie argument to be made that the work I produce is work-for-hire for that employer. But even if I'm such a private researcher, I don't owe the grant agency my work, as they don't employ me. If I'm at a university, then there's even less of an argument for someone other than me owning my work. So, research publications are a form of created work which is not under the control of the grant agency. Which means it is, initially, up to the creator(s) how they are distributed, and even if they are distributed at all. (I note with interest that there is no discussion in Geist's short post of CCA grants for artistic works. Is the creative control of an artist over his work more important than that of a researcher over his?)

Furthermore, the point of giving grants to researchers is not to make them into employees of the state, it's to encourage them to continue to research, and give them some level of autonomy in doing so (e.g., profs get a slight break from teaching, private researchers may be able to take a leave of absence, etc. and all can concentrate on their preferred work). This would be completely undercut by requiring grant recipients to publish in ways dictated by the grant agency. Indeed, again, why bother applying for a grant if you'll just end up working for someone else?

Finally, if the grant agencies try to dictate where research must be published, is there anything to stop them from dictating which research must be performed? I don't see why the descent down the slippery slope wouldn't happen.

7 comments:

undergroundman said...

First, cutting off access to traditional journals and the like, at this stage of the game, shoots the value of the publication to hell. Indeed, why bother applying for a grant when its eventual product will have to be published in some journal no-one's heard of and no-one reads?

PLoS is one of the most prestigious journals around already. The only potential problem is that their standards are too high to accept the average paper. But all journals are moving quickly towards more open-access, and if the researchers publishing their work pressure them to make an article open-access, they will give in.

Second, it misconstrues the relationship between researchers and grant-giving agencies. Researchers don't work for the grant agencies. If I'm hired as a researcher by some private institution, then there's a prima facie argument to be made that the work I produce is work-for-hire for that employer.

At the root, researchers do work for the grant-giving agencies. When I get federal grant money to work on a project, I (or the group of which I am a part of) am de facto working for the federal government. The feds can cut off that funding and oversee the operation to see that the money is being spent wisely. Of course they can attach strings to their grants; its their (actually our) money, and they have an obligation to spend it wisely. Ultimately money has the control, and grant offerers can attach any strings they feel necessary to the recipients of their grants.

Why should research which is publically funded should not be publically available? I don't get it.

This would be completely undercut by requiring grant recipients to publish in ways dictated by the grant agency. Indeed, again, why bother applying for a grant if you'll just end up working for someone else?

Grant money can be viewed as contract work. Grants (ideally) go towards projects which have external, public benefits - the government contracts people to do the project. There difference between it and a private contract are largely semantic. Yes, it's generally looser to allow for creative license and ease of dispersal, but it's still a definite project with definite constraints. You can't take a contract for studying the effects of pollution on the environment and do simply philosophy without losing the money and/or being charged criminally for fraud.


Finally, if the grant agencies try to dictate where research must be published, is there anything to stop them from dictating which research must be performed? I don't see why the descent down the slippery slope wouldn't happen.


This is a very bad place to invoke the slippery slope argument. Research publically funded should be available to the public; that is the single aim of this movement. There's no reason for it to go any further than that. There's no reason that this particular movement would incite it to go further.

ADHR said...

PLoS is purely scientific (indeed, it doesn't even publish anything in the social sciences, let alone the humanities). Whether or not it is prestigious in the fields in which it publishes -- and I'm not sure it is -- there are still huge swathes of the research community with no prestigious open-access journals. Furthermore, to conclude from PLoS that "all journals are moving quickly towards more open-access" is a pretty fallacious move; it's just overgeneralization.

It also looks to me like you're just denying the nature of the grant process, in trying to read it as equivalent to an employment contract and thus work-for-hire. (I've got a government grant for this coming year, so, trust me -- I ain't working for the province.) I agree there are constraints on what you can do with grant money (although they're a bit looser than you seem to think -- depending on the grant, you're allowed to pursue alternate avenues that seem fruitful, if they flow from the original research proposal). But that doesn't transform grant-funded work into work-for-hire. When I do work-for-hire, I have zero control over what happens to my work. Someone else pays me to do it, and they own it. When I work on a grant, I set the research goals, I convince someone to support me while I do it, and hence I own the result. The grant agency may have an interest in what I produce, but they don't own it. (And the difference between "interest" and "ownership" isn't just semantic.)

Whether the feds have the "right" to attach strings to grant money really isn't the issue. The issue is that receiving a grant doesn't make one a government employee or contractor, so the government doesn't own the produced research.

The question of whether publicly-funded research should be publicly-available is actually tangential to what I'm arguing here, which is that the grant agencies have no business forcing research to be publicly-available. If they want that, then they should just hire the researchers -- provide office space, pay their salaries, give them health benefits and parking spots. The value of open access in itself is a different point.

I also disagree that the slippery slope argument is no good. While the open access "movement" might not want to control the direction of research, if they make the granting agencies their agents in compelling public access, then I don't see what stops those agencies from simply stepping in and directing research. Indeed, if, as you're claiming, accepting a grant makes one a government employee, then there isn't anything to stop this: one's employer is perfectly entitled to give one instructions on what work to do.

Michael Geist said...

Adam,

It might be worth learning a bit more about open access before criticizing it in such an uniformed manner. Nowhere in my column do I argue that an open access requirement would require grant receipients to publish in open access journals.

Rather, in what is known as the "green approach", grant recipients would be able to publish in the journal of their choosing - open access or not. The only additional requirement would be to provide a publicly-available version of the article in freely available form (usually an institutional repository, but it could be a blog like this one) within six months of publication. The overwhelming majority of publishers will agree to such a condition. That approach will continue to provide researchers will full freedom to publish where and when they choose but, as a condition of funding make their work available to the public who has funded it.


This is clearly not an employee issue. Just as granting institutions already require certain conditions for funding (ie. regular reporting), this inserts one further requirement that involves minimal additional effort yet has the potential to provide far broader access.

MG

ADHR said...

Michael,

Thanks for stopping by. I'd appreciate a slightly more careful reading on your part before hurling ad hominems, though. Nowhere in my post do I claim that you argue an open access requirement would require grant recipients to publish in open access journals. What I actually say is this (bolding added for easy reference):

... Here's the other suggestion: "Ottawa must pressure the three federal research granting institutions to build open access requirements into their research mandates." It's not totally clear what he means by this. It could mean that NSERC, SSHRC and CIHR should encourage (whatever "encourage" means in this context) grant recipients to publish in open-access journals. This would create an annoying hurdle for those, like me, who aren't big fans of across-the-board open access, but there's already enough annoying hurdles in the grant process that one more won't make a significant difference. So, on the whole, that would be a rather benign circumstance.

More dangerously, though, it could also mean that grant recipients are required to publish in open-access journals. ...


In other words, you didn't make your meaning clear, so I offered two possible interpretations, one of which I argued against.

That said, I have no idea what the difference is supposed to be between being compelled to publish in an open-access journal from the get-go and being compelled to publish in an open-access forum of some kind six months down the road. Is there some relevant difference between a journal and a forum (a repository, for example)? Or about the six months grace period? You certainly haven't explained, in anything I've seen, why these features are supposed to matter.

Furthermore, in your comment here, you repeat the argument that the public funds the research, and hence research should be publicly accessible. As I indicated in the earlier post (snarkily), that's pretty specious: the public funds lots of things to which the public only has limited access. Public buildings are a great example. You can't waltz into a jail cell whenever you feel like it, even though, strictly, you contributed money to its creation. Controlling public access to things the public funds is really nothing new and examples do abound. So, what is supposed to make publicly-funded research special?

ADHR said...

Upon further reflection.... The six-months business does discharge my first set of worries, regarding prominence of publication. (Although I'm not sure why six months, per se.)

It does nothing to touch my second set of concerns, though. Indeed, from your comment, it seems that you accept the claim that I'm arguing against: that receiving grant monies makes one an employee of the agency. That's the only reason I can see for saying that public funding creates any right to see the product of research (vs. a right to direct the activity of research).

If accepted, the principle used here would create a presumption in favour of public access to public buildings like jails and city halls (after all, they're the products of public money) as well as records of public-paid professionals, such as physicians. But that seems to get things backwards, certainly in the case of physician records: the presumption should be against access, not in favour of it. So, there has to be more than just "well, the public pays for it, so they (ceteris paribus) get access to it".

Furthermore, what about arts grants? If I write a novel after receiving a Canada Council grant, do I have to make it publicly-available after it's been on bookshelves for six months? For a year? What about a painting or a sculpture? If I'm a playwright, do I have to put on free public exhibitions of my plays? The principle you're using -- if the public pays, then the public has presumptive open access -- would say "yes".

Michael Geist said...

I won’t touch the notion of ad hominens other than to say you ought take a closer look at the tenor of both of your posts.

In answer to your first question, the most common approach for open access (indeed the one adopted by a growing number of funding agencies including CIHR) is not to put any requirements on where a work is initially published. The only requirement is to ensure that within six months of publication, that same work is made publicly available in an open manner. Think of it as a requirement to post a reprint of the publication. This gives the researcher the full freedom to publish where they like, the publication the advantage of exclusivity for a brief period of time, and the public full access to the work sometime later.

With regard to the connection between public funding and public access, obviously the link is not the public should have access to everything it funds. Rather, public funding is about serving the public interest. That interest can be served by building roads or jails. It is also served by funding research. However, a growing number of people believe that we should strive to maximize the public benefit associated with that publicly-funded research. That extends beyond merely funding its creation to ensuring greater access to the results.

MG

ADHR said...

Strange sort of "ought"; looks like tu quoque rather than an obligation....

"Public interest" doesn't answer the question, it just pushes it back a level. You need to specify what interest the public has that is satisfied by open access, and you also need to show that no other policy would satisfy it.

You need to do the first because "public interest" is an obviously vague term. The interest of the public's served by roads is not the same as the interest served by jails and neither one is the same as any interest that could plausibly be served by open access to publicly-funded research. The interest served by roads is something like ease of transportation of people and goods. The interest served by jails is something like security. Applying either to publicly-funded research generates nonsense: "open access to research serves the interest in ease of transportation" and "open access to research serves the interest in security". So, (1) what is the interest the public has that is (allegedly) being served by open access? It seems to me you're going to have to fall back on "the public paid for it, so the public has presumptive access to it", unless you've got some neato interest in your back pocket. (Keeping in mind that any interest you propose would also, potentially, catch artistic works produced through funding from Canada Council grants. I'm still in the dark as to whether you'd consider that a bad result.)

Moving on: you need to do the second because establishing that the public has a particular interest doesn't, in itself, get you all the way to open access. What it gets you is some policy or another must be enacted to serve this interest. But which policy? Let me give you an alternative: call it "improved access".

Right now, getting into research publications is a pain in the ass, even if you're in the research community. If, like me, you're a student, you have to keep your academic finances up to date, or the university can cut off your library access. If you're a member of the community at large, you have to pay a separate fee for access to the library. I'm not sure that there's an analogue for faculty, but I wouldn't be surprised.

Furthermore, once you've got library access, you quickly find that many research publications aren't actually available to you: you have to put in inter-library requests to get hold of them, which costs you money. On top of that, of the publications that are available, some you actually have to get in physical form -- meaning you run the risk of joining a long waiting list for some book or another, and you also run the risk that the resource has been misplaced somewhere amongst the stacks -- and, of the ones you can get electronically, every publisher has their own system by which you can view and download them and every discipline has multiple databases to sift through, each with different levels of coverage. (My current beef with a publisher is Oxford UP: they have a number of philosophy e-books available, but they're viewable as webpages, not .pdfs. Which means that, if you want to download them for reading later, you have to print them to Acrobat, 5 pages at a time. It's not quite as bad as going to the dentist, but it's getting up there.)

In short, then, the bar for access is quite high, not just in terms of cash for getting into the system, but also in terms of time and expertise required to sift through the resources and find what you want. So, to that extent, the public probably isn't getting their money's worth. It doesn't follow, though, that all the barriers should come down. As I argued in the previous post, a nominal hurdle (such as a small fee for access to a library, with an easy-to-use and comprehensive database of .pdf'd papers and books) would serve some worthy goals. So, (2) what's wrong with the nominal fee, improved access model over the no-fee, open access model?

Let me also note in passing that the improved access model focuses on institutional improvements -- publishers and libraries, primarily -- and doesn't force researchers to do anything. So, the improved access model would also solve my concerns regarding interference with researchers' control over the products of their research.