Re: The True Cost of the Essentials (Implementing Peer Review)

From: Mark Doyle <doyle_at_APS.ORG>
Date: Fri, 29 Mar 2002 16:47:22 -0500

Hi Stevan,

On Thursday, March 28, 2002, at 01:55 PM, Stevan Harnad wrote:

> On Tue, 26 Mar 2002, Mark Doyle wrote:
>
>> I don't see this $30/article [mark-up] price working for a highly
>> technical journal....
>>
>> ...a wholesale replacement of the current system [by] one based on
>> self- or institutional- or subject-based archiving without
>> tackling the underlying technical issues related to long term
>> archiving would be a major mistake....
>>
>> The most viable and cost-effective solution for solving this problem is
>> to develop authoring tools that allow authors to directly create a
>> truly
>> archival XML file. The later in the process you add markup, the more
>> costly it is...
>
> I agree that it is becoming more and more apparent that off-loading
> the XML mark-up to authors is the optimal solution, and will no doubt
> happen, as user-friendly, windows-based XML markup tools are designed
> and adopted.

But who is working on them? It is a serious question.

>
> But let us not get the causal sequence or timing mixed up because of
> this. There is no immediate "problem" for which that "solution" must
> first be found!

The problem is that many smaller alternative journals or pure
self-archiving
solutions don't even address or acknowledge that a true electronic
archive
is technically sophisticated and currently labor-intensive to produce.
PDF
files are not a true electronic archive. If literature is diverted from
publishers
who are building true archives, this is a loss to the community. If
comparisons
are made between what publishers charge and what low-cost archives do
without explicitly examining all things that publishers do, then the
comparisons
are unfair and dangerous. Of course not all publishers are cost effective
and not all publishers create the same level of electronic archives.
However,
a real analysis of the costs is needed. My point is that $30/articles is
extremely
low and it doesn't represent a fair comparison. It gives the perception
that
even well-intentioned, not-for-profit publishers are abusing the system.
This
I think is the root of why the ALPSP response and the response from many
society publishers hasn't been enthusiastic about the BOAI. We understand
that there are deeper things involved than just making the content
available
for the here and now.

> Priority #1, by far, is opening access to this (peer-reviewed)
> literature right now (yesterday!). There is absolutely no excuse for
> blocking its access or impact for a microsecond longer.

Of course we are in agreement that authors should self-archive their work
in a way that makes it available to as wide an audience as possible.
But we aren't in agreement when you say things like (from "Re: BBC News
SCI-TECH Boost for research paper access"):

> On Monday, March 25, 2002, at 02:42 PM, Stevan Harnad wrote:
>
> But which costs? The $500 for peer-review is uncontested. But it is the
> only remaining essential cost in the era of online institutional
> research archiving.

These archives don't yet exist at the same level as that of most
publishers
(whether it is the APS or Elsevier). Name one free long
term archive that has adequately addressed this issue and for which the
costs of
preparing the archive are not laundered by a publisher. (Hint: arXiv.org
isn't one
and neither is PubMedCentral). Anything that is a PDF/TeX/Word
repository isn't
one either. Name a university that takes on the responsibility for
translating
their author's output to a uniform, well designed marked up archive (this
essentially means XML these days).

> Meanwhile, however, journals continue as before, selling their paper
> versions and their markup, and their online page-images, etc. It is
> most definitely not a PRECONDITION for freeing access to this entire
> literature, right now, that authors should first be able to provide
> XML-marked-up drafts!

Yes, but it is a precondition for moving to pure self-archiving and
alternative
journals to replace current publisher archives as advocated in the BOAI
and similar
places. You can't throw the baby out with the bath water. If publishers
are the only ones focused on creating a true long term archive but their
way of paying for it is undermined before a new economic model is in
place,
then there will be a loss to the community. My main point is that
alternative
journals are easy to start, but that they need to do things beyond peer
review
and delivering PDFs.

> On the contrary: It will be the availability of this whole literature
> online and free that will DRIVE the downsizing of publication to the
> essentials, the development of authoring tools, and the upgrading of
> the author-version, as the need for that arises. The ONLY need right
> now is to free this literature; and an author-supplied peer-reviewed
> final draft is sufficient to do that.

I don't agree simply for the reason that authoring tools and rich XML
repositories aren't explicitly valued by current researchers because
they only
think in terms of delivering or receiving a PDF or HTML file. However,
derived benefits are valued now in terms of the searching and linking
that
researchers crave in online journals. Why should
the evolution of the system be a path that first goes to bare bones
throwing
away things that we know are important rather than building a new
system in a way that has all of the essentials in which the economic
model
recognizes the unappreciated costs in the current system?

> Please let us not needlessly mix, complicate, or hamstring agendas, at
> the risk of delaying this overdue benefit for research and researchers
> any longer.

It isn't needless. It should be part of the debate. There should be an
effort
to provide for authors a way to truly archive their material and it
should
be an explicit part of understanding the cost in moving to a new way of
doing things.

> Publisher practices, as well as author tools and author practices will
> evolve to adapt to the reality of open access.

What, after they realize they are missing some key feature such as robust
reference linking? Why not get it right to begin with?

> Open access need not
> wait for anything at all at this point. Those who have already
> self-archived have not waited, and there is no need for the rest of us
> to wait either.

Again, there is nothing wrong with self-archiving to increase the access
and distribution, and it should be done by all now. However, this
isn't the end all and there should be a broader discussion of what the
correct end point we are working towards is. It is not just free access
to article content.

> Perhaps under BOAI Strategy 2 (creating and converting to open-access
> journals) promoting the development of XML authoring tools would be a
> money well spent. But let us not make that a brake on Strategy 1
> (author/institution self-archiving, NOW), for it is not. Strategy 1 need
> not and should not wait for XML authoring tools.

We are in agreement that authors should avail themselves of the
opportunity
to self-archive now. But it is apparent that strategies 1 and 2 will
lead to
different, inequivalent, end points and my concern is that strategy 1 is
being promoted as just as good as strategy 2 for the long term end point.

>>> [S.H.: What about the cost of implementing peer review?]
>>
>> The archiving cost is just as, if not more, important than the peer
>> review cost and the fact that is it usually missing from your
>> discussions
>> is a major weakness. I don't think the $30/article number is
>> generalizable
>> to all fields of scholarly communication.
>
> I am afraid I have to disagree rather strongly here. Not only is it not
> the case that markup (and its costs) is more important than peer review
> (and its costs) -- what an idea!

Peer review is a layer built upon archiving. What good is a peer review
label
if it doesn't lead to something locatable, readable, and persistent?
Clearly arXiv.org
succeeds not because it provides peer review, but because it provides
excellent short and medium term access to the literature.

> -- but by the time markup becomes a
> salient factor at all (which will be when the literature has been freed
> by self-archiving and publishers are ready to downsize to the
> essentials), necessity will be the mother of invention, and the
> requisite XML authoring tools will be developed.

Markup is already a salient factor. It is why online journals provided
by publishers
have many more features than arXiv.org. I emphasis its importance because
recognizing this fact is important for figuring out how to transition
existing
subscription based journals to an open access model without undoing a
lot of what we provide. In 1995 arXiv.org had fair claim to saying that
it
provided the same (or better) functionality than what publishers were
providing.
It can no longer do that. However, with real markup underlying arXiv.org,
all of the features present in publishers' journals would be trivial to
implement
and this would further hasten the transformation.

> There is no problem of principle there, just one of practice, and the
> current absence of demand for author XML, within the current status quo
> (why should there be a demand?):

Why shouldn't the linking and searching within arXiv.org be as extensive
as that in publishers' electronic journals?

> But that is exactly the status quo
> that the author/institution self-archiving is meant to alter,
> demonstrating the huge utility of the free peer-reviewed drafts, even
> without proper mark-up.

Only to the point of access to the basic content. Not to linking,
searching,
or being able to reformat it 20 years from now for some new media.

> If freeing access diminishes subscription revenue, it means that this
> vanilla peer reviewed version has considerable market value; if it
> doesn't diminish subscription revenue, we don't need to worry about any
> of this, and author XML markup can take its time coming as long as it
> wishes.

No, this is flawed. One hears increasingly about how major publisher
platforms that cost libraries enormous amounts of money are extremely
heavily used by end users because of the rich feature sets offered. You
will not free up this money to pay for new models or get authors to move
to alternative journals unless you can duplicate or compete against
these kinds of services. Yes, people will be able to read the articles,
but
you are still giving large publishers a lock on the marketplace. By not
looking at how to increase the features offered in something like
arXiv.org
or how to help smaller publishers transform to new economic models,
you will extend the length of time needed undo the lock.

> So much for BOAI Strategy 1. Obviously BOAI Strategy 2 (open-access
> startups and conversions) will want to minimize costs, and one of the
> ways will be to offload XML markup on authors, and hence XML authoring
> tools would be very handy to have. So by all means let us develop them.
> But let us not mix up these two BOAI Strategies and their causal
> interaction, describing as "a major mistake" the "wholesale replacement
> of the current system [by] one based on self- or institutional- or
> subject-based archiving without tackling the underlying technical
> issues related to long term archiving."

> No one is proposing "wholesale replacement of the current system" by
> self-archiving! (And not primarily because of markup, but because of
> peer review!) An agenda like that would be incoherent, like proposing
> to replace all driving by hitch-hiking!

OK, perhaps I was being too telegraphic. Perhaps I misunderstand your
view of the transition. Currently a publisher (peer review entity) like
the APS does several different tasks: Peer review, digital archiving,
distribution
of article content, and provision of services based on content. My
understanding
is that what is being advocated is that we just focus on peer review and
let the
rest take care of itself, that peer review would be a layer on top of
some
institutional/self/subject archives. My assertions are 1) digital
archiving
is an extremely important aspect of what publishers currently do, often
left out of the discussion, 2) peer review is tied to some kind of an
archive,
whether it is external or internal to the peer review entity, 3) that
researchers
rely now on services built on top of current proper electronic archives.
This leads me to conclude that peer review cannot simply transition to
being a layer upon an external archive until the external archive is of
the same quality as our internal archive. Given that both the peer review
and archive components are somewhat costly, a new economic model
must take both into account so that existing journals can become free.
Any alternative journals have to face the same issues or not offer the
same
benefits to researchers.

> Self-archiving is a means to an end, and that end is open access, now!
> As the "system" adjusts to open access, many things may happen. The
> demand for the old products might continue, and continue to pay the
> bills, in which case no further adjustment will be necessary, apart from
> the normal drive to minimize costs. But if the old demand flags because
> of open access, then cost-cutting will become more urgent, and the
> transition, which I think is likely, to author-institution-end payment
> for peer review as well author-end mark-up, will take place in
> earnest.

My objection is based on the "serialization" you are imposing. The way
you phrase it here advocates the fastest possible path to self-archiving
and then seeing what comes of it. You predict that author markup
will be discovered to be important after all and then there will be
another
transition (perhaps now without publishers) in which the repositories
are enhanced to bring them up to the level of what is part of electronic
journals today. What I would like to see is that this occur in parallel
with
institutions and others who take on the tasks of building the
infrastructure
also look deeper and tackle the all issues at the same time. The sooner
work is started on addressing the complete problem, then the sooner
the final transition can take place. Also, by being explicit about what
is important,
one can create a sound basis for a new model in which current publishers
can go through a safe transition to open access. But if peer review is
the
only focus for the first step of some staged transition, then it will be
much harder, if not impossible, for a publisher like the APS to continue
to do what we think is important for
the long term viability of the literature we review.

> But right now, it is just another excuse for lingering longer in the
> status quo.

Not at all. The focus needs to be broadened so that the status quo
can change without imperiling the benefits of the status quo.

> Now I know this is not what Mark means to encourage, so I am unwrapping
> these causal and temporal factors here simply to make it more
> transparent here that he is inadvertenely mixing apples with oranges.

Thanks, but that mischaracterizes what I am I saying. The apples and
oranges
are linked together (they always have been - we printed (and thus
archived)
what was peer reviewed in the past and now we electronically publish
and archive what we peer review). Yes, the economics have changed because
raw dissemination as demonstrated by arXiv.org is so low, and thus it
should
be possible to separate them further, but they will still remain linked.
Both
are important to the long term health of the scholarly literature.
Without care,
you are going to end up with a basket of apples and a craving for orange
juice.

Cheers,
Mark

Mark Doyle
Manager, Product Development
The American Physical Society
doyle_at_aps.org
Received on Fri Mar 29 2002 - 23:15:56 GMT