Friday, September 28, 2007

Open access series: The status quo.

Today, I'm finally continuing my open-access series. Here is the introductory post, for those who want to refresh their memories as to what the hell I'm talking about. In today's post, I'm going to be talking about arguments for and against the status quo. Or, at least, that was my original intention. As it turns out, though, it's incredibly hard to find positive arguments in favour of the status quo.

Let me explain what I mean. There's basically two ways to make an argument for some claim or proposition, p. First, I can advance a positive argument. That is, I can tell you some reasons to think that p is true. Second, I can advance a negative argument. That is, I can tell you some reasons to think that not-p is false. In introductory formal logic textbooks, the distinction is really irrelevant, as the "reasons" in question are usually premises in deductively valid or invalid arguments. That is, the reasons either guarantee the truth of the conclusion, or don't. From this, it follows that the two ways are equivalent.

In more complex arguments, though, the second way of arguing usually turns out to be the weaker one. This is because it only indirectly gives reasons to think that p is true, i.e., if not-p is false, then you should think p is true. The first way directly gives you reasons to think that p is true. And because the second way is indirect, there's two possible sources of weakness in the reasoning, as compared to only one for the first way. The first way will produce a weak argument if the reasons give little or no support to the truth of p. The second way shares this weakness, in the indirect form: if the reasons give little or no support to the falsity of not-p. The second way also, though, has a weakness in the connection between p and not-p. Again, in strict formal logics, not-p is the contradictory of p; and, outside of these systems, when the second way goes right, this is true as well. But it need not be. What fills the place of not-p can amount to "what I (the person making the argument) believe to be the alternatives to p". And, obviously, these are not always co-extensive.

So, whenever possible, it's better to give a positive argument for your view as well as (or instead of) a negative argument against opposing views. Unless you're dealing with a strict formal system, you can never be sure that you've got all the possible alternatives covered by not-p, which opens up the unique flaw of the negative argument.

What's this got to do with open access? Well, simply that the arguments I've managed to find in favour of the status quo are -- surprise! -- negative. That is, they tell you why it's bad to move away from the status quo, not why the status quo is any good. Now, given that the status quo is the status quo, this is not an incomprehensible move. After all, institutional inertia being what it is, as long as change is not clearly beneficial, the status quo will prevail. But institutional inertia is a really weak reason to maintain some practice; it only works if every possible reason for change is a failure. And it's extremely unlikely that this is the case.

Still, let me look at some of the major negative arguments I've managed to pull together and see how well they do (or don't) work. They fall into three broad classes. The first is the increased or initial costs of changing the current model. Most arguments I can find fall into this category. A few, however, fall into a different category, that of the threat to something fundamental to the academy. And a few fall into a third category, that of the irrelevance of changing the system. That is, it simply isn't an important or necessary change. For the sake of having easy labels, I'll refer to both categories as the "cost" argument, the "academic" argument, and the "relevance" argument, respectively.

The cost argument goes like this.
  1. If the status quo is altered, it will consume certain kinds of resources and/or deprive certain groups or persons of gain.
  2. This is bad.
  3. Therefore, the status quo should not be altered.
To make the argument work, we first have to fill in details of the resources lost or the gain deprived. So, we could say that publishers will be deprived of profits. Or we could say that an open access system would require money (and other resources) to be instituted and maintained. Or we could say that professionals (e.g., health care practitioners) would feel obligated to understand even more research than they currently do, which would consume more of their time and energy. Or we could adopt a sort of combo platter approach and take a little of each of these, and others.

As far as that goes, I have no problem with the argument. It states something that I'm willing to grant as true or, at least, not obviously false. So, (1)'s okay. And, given (1) and (2), (3) really does follow. The problem is the evaluative premise (2). We're supposed to be shown be the argument that something important is being lost by moving from the status quo to either TOA or IA. But the argument just assumes this. If there were a positive argument for the status quo, then (2) might work. There might actually be something important to the status quo. But, there is no such argument; hence, there is no support for (2).

Therefore, the cost argument fails.

Moving on to the academic argument. Publishers and other supporters of the status quo sometimes argue they have an important gatekeeping role to play in the research community. And a move away from the status quo would eliminate that. Putting it in a schematic form:
  1. The status quo serves an important function.
  2. Changing the status quo would eliminate this.
  3. Therefore, the status quo should not be altered.
Like the cost argument, the logic here is fine. Unlike the cost argument, I think both premises are false.

(2) is false because the important function does not (or at least does not obviously -- it needs argument) require the status quo in order to be served. Peer review is a usual candidate for the "important function", as is a more generic, keeping the hoi polloi away, sort of function (call this the "elitist" function). Peer review has no necessary connection to the status quo, though. You don't need to have a system of limited access, subscriber-only journals in order to encourage academics to review each other's work. (Really, what you need is paid review, but let's put that aside for now....) There's no evident connection between the barriers against accessing material and the process of peer review.

There is a connection between barriers and the elitist function. And I'm not opposed to the elitist function. I'm too much of an elitist myself to take seriously the idea that everyone, no matter their education, ability, or interests, should be allowed to access any information they like. Information can be incredibly dangerous in the wrong hands. (How do we tell which are the wrong hands? Topic for another time, perhaps.) The problem is the barriers are a very ham-fisted way of fulfilling the elitist function. If you really want to stop people who don't have certain qualifications from accessing particular material, doing so on the basis of ability to pay or access to a library is indirect at best, irrelevant at worst. Even a "need to know" basis works better, for at least then access to information is indexed to a relevant quality (i.e., need for the information).

(1) is also false, because the status quo actually fails to effectively serve either peer review or the elitist function. As said above, the elitist function could be better served by directly addressing the question of who should have access to information, not by cost or level of library service as a (poor) proxy. And, as also said (although parenthetically), peer review is not obviously best served by the status quo. Paying peer reviewers might be a better way to encourage peer review. But certainly access barriers to research don't seem to do anything to make peer review easier or more common. Indeed, it's hard to see how these could be connected.

So, the academic argument is no good, either.

Finally, then, we come to the relevance argument. This argument goes something like this:
  1. There is no need to change from the status quo.
  2. Therefore, the status quo should not be altered.
Now, this is, of course, enthymic: we have to add the claim that we shouldn't do things we don't need to. This makes the argument:
  1. There is no need to change from the status quo.
  2. If there is no need to change, then there should not be change.
  3. Therefore, the status quo should not be altered.
I'm noticing that these arguments are actually getting worse; this is actually quite pathetic. The most obvious problem with this argument is that it begs the question in (1). One way of phrasing the dispute between supporters and opponents of the status quo is that the latter assert, while the former deny, that there is a need to change the status quo. Since this is the very bone of contention, it is fundamentally fallacious to claim there is no such need in the premise of an argument in support of the status quo.

To make matters worse, (2) is not necessarily true. There is no necessary connection between need and obligation. Let me start by considering whether obligation is exhausted by need. (In an unrelated, but relevant, note, this unreflective invocation of need is what tends to drive me crazy about environmental and anti-poverty groups.) Most of what I do in my day is not done out of need. Most of what I have I don't need. But I like what I do (usually), and what I have, and they add some kind of value to my life. The idea that obligation gives out when need ends is thus incredibly radical. It implies that I'm never obligated to do more than I need to (as said in (2)), and I'm never entitled to what I don't need (as environmental and anti-poverty groups sometimes say). And I take this as a reductio of the position. My obligations go beyond what I need to do: they extend, at least partly, into what it is good that I do. Similarly, what I am entitled to goes beyond need, at least partly into what's good for me (or, perhaps, what I take to be good for me?).

To proceed in the other direction, need does not entail obligation. I'm not obligated to do for you everything that you need, nor are you so obligated towards me. Some people have these sorts of obligations to each other, because of close relationship, explicit agreement, or other such mechanisms. But not everyone does. So, the fact that there is some need doesn't necessarily create a specific obligation.

For (2) to be true, it's the first that's really problematic. (2) asserts the contrapositive of the claim that obligation entails need: it asserts that a lack of need entails a lack of obligation. (Given a sentence, "if A, then B", its contrapositive is "if not-B, then not-A".) To shore this up would require showing that any time there is an obligation, there must be need. As I've argued, though, many genuine obligations I have are in no way related to need. They're instead related to things like explicit agreement or personal relationship; and, more generally, to value.

(Edit Sept. 30 4:08pm EST:: undergroundman suggests in the comments that while the inverse of (2) may be true -- necessity implies obligation -- it doesn't follow from that that (2) is true -- lack of necessity implies lack of obligation. So, rather than failing to prove the contrapositive (obligation entails need), it may be an erroneous use of inversion (which can generate falsity from truth) instead of contraposition, which is a valid rule of inference.)

Thus, in conclusion, there are no good arguments in favour of the status quo. None of the cost argument, the academic argument, nor the relevance argument really works. So, the only thing speaking in favour of the status quo -- the only thing that can -- would be the failure of a case supporting all the other alternatives, including the two I will consider over the next few weeks, TOA and IA. So, I need to consider the case in favour of each; and, next week, I will start with TOA.

2 comments:

undergroundman said...

As it turns out, though, it's incredibly hard to find positive arguments in favour of the status quo.

How about Elsevier's arguments that the under a different model the costs would be passed directly on to scholars, rather than being shouldered by institutions paying for subscriptions? This is a negative argument, but arguments like this are also positive, really, depending on how you want to spin it: "The status quo charges institutions rather than individual researchers and thus does not impose a publishing disincentive on individuals."

I think the argument is bunk, of course, because really it's just cutting out the middlemen -- the institutions could surely afford to pay for the submissions of their scholars directly, and would likely save a lot of money that way. Small universities like mine would have to pay hardly anything, while large universities with lots of funds (and lots of subscription journals) can afford to pay for their faculty's individual research to be published.

The first way directly gives you reasons to think that p is true. And because the second way is indirect, there's two possible sources of weakness in the reasoning, as compared to only one for the first way.

This second paragraph is hard for me to follow, but I basically disagree. P in this case is that the status quo provides net benefits over alternative systems which is equivalent to saying that the alternative systems imposes net costs. These two equivalent arguments in most cases and I have trouble seeing how arguments in favor necessarily lead to a better coverage of the alternatives, as one can easily ignore the benefits of the alternatives while arguing in favor of the status quo. I suppose I can see how the "not-P" arguments can seem to slip into strawmen. Is that kind of what you're implying?

The problem is the evaluative premise (2). We're supposed to be shown be the argument that something important is being lost by moving from the status quo to either TOA or IA. But the argument just assumes this.

Which arguments are you referring to, out of curiosity? It's interesting to contrast this to an economic argument, which might evaluate the net benefits of the status quo vrs the net benefits of an open-access model. Of course, many of the net benefits in the open-access model are external and tricky to value, although certainly not impossible. I ordered Elsevier's Ecological Economics journal archive for 2004 through ILLiad-- it must have weighed 10 pounds! Then I had to photocopy it, so it's tricky to work with it digitally. The papers in it are on ScienceDirect -- for $10-$30 a piece. Open access can reduce a huge amount of bureaucratic wasted time and energy and increase organization dramatically. Plus, it likely just makes sense from a fiscal perspective to cut out the pricey middlemen, who are accruing all the benefits from the prestige of the researchers who get very little.

You don't need to have a system of limited access, subscriber-only journals in order to encourage academics to review each other's work. (Really, what you need is paid review, but let's put that aside for now....) There's no evident connection between the barriers against accessing material and the process of peer review.

Indeed, and is the closed-door peer review process even that effective? It may be better to put up preprints for several interested experts to assess before something is officially published. I believe this is starting to be done.

Information can be incredibly dangerous in the wrong hands.

The most dangerous information is already on the internet (arXiv.org).

If you really want to stop people who don't have certain qualifications from accessing particular material, doing so on the basis of ability to pay or access to a library is indirect at best, irrelevant at worst.

Honestly, can you give me a reason why you would want to do that, especially to the extent that it becomes policy? And don't bring out the nuclear physics example -- let's put that one aside.

The relevance argument is really just a dismissal of any trouble with the status quo, which seems to me to be actually a flat statement that either the alternatives suck or the status quo is overwhelmingly great. Or maybe it really is that there is no need, as you say. I suppose you're right in that it is a common and crude fallacy that what is not necessary should not be done: it's the inverse of the more logical idea that what is necessary must be done.

To really put weight behind your statement that the cost arguments don't work, it might be better to articulate what those arguments are and why they don't work. That economic study from the European Commission was interesting in that it found significant benefits to open-access. It's also interesting to look the process. Right now the money flows from the government to the research groups, who then publish it (usually without being paid) to a journal run by Elsevier, who gets a 26% profit margin while selling the journals for egregious prices back to the schools which produce the research. Cut the journals and suddenly you have a lot more money to fund research.

ADHR said...

Isn't that cost-transfer argument just a variant on the cost argument I considered? i.e., there's some new cost that appears under a different model.

P in this case is that the status quo provides net benefits over alternative systems which is equivalent to saying that the alternative systems imposes net costs.

As you've given them, those aren't equivalent, unless you add the assumption that lack of net benefits = net costs (or, alternatively, lack of net costs = net benefits). If you want to assume instead, as I tend to think one should, that lack of benefits can also be lack of costs -- an area of neutrality -- then that doesn't quite fly.

As far as negative vs. positive arguments, the problem with the negative arguments is something like the strawman problem, only a bit worse because it also covers the possibility of a false dilemma or a false choice (or even a complex question).

Ignoring the benefits of an alternative is a real worry with positive arguments, but it's also a worry with negative arguments -- you may fail to consider some benefit, and thus fail to appropriately criticize it.

When it comes to peer review, I'm not sure that an open peer review process is necessarily better, though I'm sympathetic that it can't be worse. Closed-door peer review does do something to reduce the possibility of irrelevant bias, though.

Regarding the relevance argument, I should've put the inverse in there. The inverse works -- necessity implies obligation -- even though the original -- lack of necessity implies lack of obligation -- does not. I think there's probably a confusion there between inversion (which can transform a true sentence into a false one, or vice versa) and contraposition (which is a valid logical inference).

I'm going to get to considering whether TOA or IA will reduce costs in the next few weeks. ;) And I'll also get into the elitist argument -- that's one of the reasons I'm going to try to draw out in order to favour IA over TOA.