Marcus' World

The purpose and future of academic research libraries

On Friday afternoon I attended a seminar presentation by Tom Leonard, University Librarian at the University of California Berkeley. Leonard has a doctorate in history and has been on the faculty of Cal's journalism school since 1976. He's also been the University Librarian since 2001.

Ever the historian, Leonard began with an illuminating historical overview of academic libraries in the US. He divided academic libraries pre-Web into two broad phases: the "slacker phase" and the "overachiever phase."

Slacker phase (late 1800s): In this period students and faculty viewed the campus library as irrelevant, removed from the realities and heart of campus life. Hours were terrible, since the library closed at sundown.  So at the exact time students could most easily go to the library, after the end of the class day, it was closed. And collections decisions had no bearing upon what students learned in the curriculum. What was this library thing anyway?

Overachiever phase (turn of the century up until dawn of the digital age): The big turn happened in the late 1800s and early 1900s--first on the East Coast, then spreading to the West Coast. As doctoral programs blossomed, so too did collections budgets and the desire to tie collection decisions with the curriculum. Academic libraries transitioned from peripheral entities into the "heart of the campus." The stated desire was to collect "everything," which in its way was Google-esque (Google's mission, after all, is to "organize the world's information.") 

Leonard pointed out that "everything," in practice, meant much deeper collections in German and French than in Chinese or Korean. There was no attempt to collect with true comprehensiveness, despite rhetoric to the contrary. As he spoke, I remembered an analogy from the Dewey Decimal System. Melvil Dewey's treatment of Christianity in the Religion bin is fine-grained, while all the rest of the world's religions are dumped at the end of the line.

Although the great academic library collections were not truly comprehensive, they did become massive. As the 20th century continued, the key metric of an academic library's prowess was its "number of volumes held." Leonard referred to this competition between libraries as an "imperial" "arms race." I recall that "x million volumes" figure bandied about on my tour of the Northwestern campus as a prospective student, and this metric is very much with us today as we go ever deeper into the digital age.

The trouble is that this concept--"millions of volumes held"--is more about keeping up with the Joneses than anything else. How much difference does it really make that Cal has twelve copies of a book while Harvard has eleven? The more rational thing to do, way back when, would have been for the elite academic libraries to coordinate their collection decisions to ensure true comprehensiveness and to stretch dollars the furthest. Alas, human nature tends to be more narrow and competitive than this.  And so, over a century we built up impressive collections that are now very expensive to maintain and to a large extent seldom used.

Today: After outlining this history, Leonard spoke about the challenges of making substantial changes within a large bureaucracy that rewards settled habits. Even if organizations know they need to change, it is very hard to do so when there is no shared language available to define what those changes should be. This is the current state of academic libraries.

Leonard described the work of a new Commission which will soon make changes about how the UC Berkeley Libraries should operate in the coming years, given budget cuts and the continued evolution of digital scholarship. Even humanities scholars, long-perceived as monograph fetishists, are now building Centers for Digital Humanities.  In theory, this confluence of new modes of scholarship and budget strains will lead to bold thinking about how the Libraries should operate. But I suspect the Commission will offer a modest set of recommendations, locating operational effiiciencies that preserve the core function of collection development, and the core metaphor of the library as a keeper of books.

I could be wrong. We will have to await the Commission's forthcoming report.  My prediction about the Commission's likely modestness is because most of its members are UC Berkeley faculty members.  For at least a decade librarians have been grappling with the challenges of preserving "born digital" materials, and creatively re-purposing library space.  But the deeply ingrained notion that the library is a warehouse for printed work has not yet budged very far in faculty consciousness, and I suspect it will carry the day. Arguments about the "imperial" over-reach of 20th century collection practices probably do not mean much to the Chair of the History department who is counting on an ample monograph collection next year.

The Future: So, where are we headed?  Assuming that established notions of what an academic library should be remain in force, I'd like to see a bifurcated organizational structure evolve.  The apparatus for physical collections and preservations would remain in place; alongside it, a skunkworks devoted to facilitating and preserving new modes of scholarship would be established. What's a skunkworks? A "project developed by a small and loosely structured group of people who research and develop a project primarily for the sake of radical innovation."

Bethany Nowviskie has already developed one such skunkworks with some success at UVa. Other ideas are available in the "Libraries" section of "Hacking the Academy."  The goal of these projects is to inspire and facilitate impactful projects in libraries that position us to thrive in the 21st century.

Perhaps these skunkworks would need to be funded by soft money first to prove their worth, and then later be added to the main library budget (at which point--when money's on the table--the real battles about the library's purpose would commence.) Or perhaps faculty members called to serve on the next library Commission in 15 years will demand serious treatment for digital scholarship, finally severing the tie to those imperial 20th century collection development practices. If that happens, these skunkworks will have been proving grounds.

Old Wine in New Bottles: Lorcan Dempsey in EDUCAUSE Review

Thanks to the member of my staff who pointed me to Lorcan Dempsey's excellent piece in the December 2012 EDUCAUSE Review, "Thirteen Ways of Looking at Libraries, Discovery and the Catalog: Scale, Workflow, Attention."  In this piece Dempsey, who now works for OCLC and used to work for JISC, synthesizes many ideas that he's previously developed on his blog.  He also weaves in work from elsewhere, all in an effort to better understand the current status and potential evolution of the library catalog.

It's a meaty piece, best read closely to allow time for Dempsey's many ideas to germinate.  The gestalt of the piece is a call to place old wine in new bottles.  One fundamental contribution of librarianship is to provide the means of organizing and retrieving information objects--books, serials, maps, DVDs. etc. This has required classification schemes, union catalogs, MARC records...the entire cataloging apparatus. 

All of that intellectual work is still valid and useful and important--this is the "old wine."  But we haven't yet figured out to graft these principles onto the "new bottles" of born-digital objects.  Library cataloging rules date from a time when you had to physically publish or manufacture a work.

Cataloging protocols evolved over centuries, so it will take a while to make a fundamental transition toward describing digital objects. (Just linking to web sites from within the catalog doesn't count.)  This transition will only occur if librarians acknowledge the fundamental shift that Dempsey describes--that the library catalog, as a record of a locally selected and managed collection, will represent an ever-decreasing share of the user's information base.

Google, Wikipedia, Amazon, iTunes, Pandora--these are all information sources that library patrons have access to without any need to use the library.  If we keep perfecting the catalog as we currently understand it, we'll eventually have a perfect catalog that nobody uses.  It's better to embrace the world entire, both in and outside our walls, and give what's floating online the organization and curation it so desperately needs.

Small Victories

One of my most tangible accomplishments at Samuel Merritt has been the creation of our open access publishing fund. This allows our faculty researchers to pay the authors fees to make their articles open access, thereby increasing their reach. I took a liberal stance--as long as the article itself is open access, even if the entire journal is not, authors can utilize the fund.  In 2012 monies for the fund came from cancelling little used online journal packages, now the open access fund is a line item in the library's budget.

As someone who has long followed conversations about open access publishing, I have mixed feelings about the author pays model. It's ended up providing a new revenue stream to publishers without challenging any conception of the outdated notion that for scientific communication the "paper is king." Open access journals produce PDFs, principally...a PDF stands-in for the old fashioned printed page, even though the solutions to many problems cannot be fully depicted in ink, paper, and two dimensions.  The Web allows multi-factorial depictions of complex problems. But we still insist that the formal paper-a necessarily reductive method--is an effective dissemination tool for scientifc work.

Ahem.

On the other hand, when I started my position at Samuel Merritt discussion of open access was not on the radar. I attended several faculty meetings in which people did not know what I was talking about. I had to find the resources to support the fund, and keep on promoting it.  Last spring, at budget time, I ruefully worried that nobody would ever listen to me. 

Then, on Friday I pulled together a campus news article about the open access fund.  Turns out we've supported publication of four papers, in almost every school of the University.  Not so bad.

The trouble with being a "visionary" (pardon the pomposity please) is that you always see ten moves down into the chess game. Maybe even twenty.  But the game is played one move at a time. Small victories matter, and become triumphs in the end.

Gartner "Hype Cycle" for Emerging Technologies

Thanks to Joe Murphy for today's excellent post about Gartner's "Hype Cycle" for emerging technologies.  Gartner has provided an analysis of technologies that will soon be ready for productive use, those that are at the "peak of inflated expectations," and those now wallowing in the "trough of disillusion." 

All is not lost in that trough, which currently contains technologies like virtual worlds and home health monitoring tools.  Eventually some of these technologies will triumph, edging first into the "slope of enlightenment" before reaching the "plateau of productivity."

In sum: don't believe the hype at first, but don't dismiss new ideas or technologies out of hand either.  Skepticism and caution are good, as long as this is in the service of improving new ideas rather than quashing them from the start.  Organizations are risk averse and rule bound places, which is why inertia and mediocrity usually prevail.  But it does not have to be this way.  The change agent needs to have patience, political savvy, and endurance.  Take the long road, not a shortcut.

Weinberger and Wieseltier

Last week I finished reading David Weinberger's book Too Big to Know, which I wrote about last year after hearing him give a talk about it at Cal.  Weinberger brings his philosopher's eye and library experience together in a very appealing and unique way.  I spoke with him soon after his visit to Cal, and Weinberger reminded me that philosophers have been grappling with questions about what counts as authoritative and truthful for decades.  Now that our means of sharing knowledge no longer must go through traditional filters--a tiny example: I am writing this blog post without any editorial intervention besides my own--librarians must contend with these questions too. 

Should we only send patrons to sources we've long known and trusted?  What if that now feels stale, or at least incomplete?  In that case, how do we determine how widely to cast our nets and what should we tell our patrons? These are among the questions confronting librarians today.

Weinberger offers a way to reframe these questions, by reminding us how previous limitations of form have structured our notions of how to express knowledge.  When your only tool is a hammer, everything is a nail. Likewise...when your only tool is a printing press, your only output is a book. (A less pithy formulation, admittedly, but you get the point.)  The Web has now solved the distribution problem, even if we are not yet anywhere close to assuring quality control.

Indeed, most of what's online is shallow or base.  Weinberger acknowledges this several times.  It's enough to make one pine for the good old days of scrupulous editors and reputable publishing houses.  Or at least, it's enough to make Leon Wieseltier feel this way.  Lately he's been on a kick, in his back page of The New Republic, about the various plagues--moral, epistemological, volitional--of the digital age. This month he extolled the virtues of editorial gatekeepers, who guard the culture from dreck and thereby save us from intellectual ruin.

Oh...Leon! Yes, now there's a wild and wooly online world to navigate.  In this online space, though, the dedicated amateur can improve the work of even the most impressive expert.  Crowdsourced data gatherers can solve problems that a fleet of researchers could not have addressed pre-Web.  New types of inquiries, that quite literally were impossible before, are now not only possible but inevitable. That's the gain Weinberger describes in his book, while acknowledging the loss of our previous epistemological certainty. 

Wieseltier can describe a half empty glass if he wants to, and his very pugnacity about it is compelling.  But not convincing.  The glass is half full.

Copyright Reform and the Digital Public Library of America

I've been following the efforts to build the Digital Public Library of America, an idea Harvard University Librarian Robert Darnton has proposed on various occasions in the New York Review of Books (such as here.)  To date the DPLA has established working groups (called, oddly, "workstreams") and a governance structure. It will soon be in search of an executive director and permanent headquarters.

I attended a planning meeting the DPLA held in San Francisco in April of this year, where the fractious interests of the various parts of libraryland were on full display.  Some members of the audience railed against the publishers and their restrictive e-book licensing schemes, others were more mellow.  The Archivist of the United States spoke to us by phone.  At the end of the day it was unclear what the DPLA would actually be--a website? a portal to other web sites? server space? a set of software tools? etc.  This is not to fault the DPLA's creators, but rather a reflection of the almost impossible task of creating a digital public library for a nation as sprawling as America. (For a better take on that day's events, read Karen Schneider.)

The fly in the ointment of all digital schemes, including DPLA, is our antiquated copyright law.  Google's proposed settlement with publishing groups regarding its book scanning project was rejected, and now it looks like legal proceedings between the two groups awaits.  The core issue is whether Google's scanning efforts constitiuted a permissible fair use of library material (as Google claims) or "massive copyright infringement"" (as publishers and authors allege).  These same issues bedevil the DPLA, as Nicholar Carr documents in a recent issue of the Technology Review.  For now, all the DPLA will scan and make available in full text are guaranteed public domain documents (ie, those works published before 1923).

That's right--with no change to our copyright laws, the very ambitious DPLA can only provide access to works that are almost, or in some cases more than, 100 years old.  Nothing against those works, but history isn't everything.  In the most recent Chronicle Review, Pamela Samuelson suggests modest remedies for this state of affairs (modest because she feels comprehensive copyright reform is impossible to achieve in Washington anytime soon).  Samuelson's purpose is to ensure that the DPLA will actually have substance.

Her ideas are both intelligent and obvious: grant libraries a blanket privilege to scan in-copyright works for the purpose of text mining (not for the purpose of posting the complete scans); allow orphan works--those for whom rights holders can't be located after a diligent search--to be posted open access; respect reader privacy; improve access for people who can't read print.  None of this really challenges our paradigm of copyright law or our notion that a textual object is forever and always the best way to convey information.  It is sad that such modest thinking represents the outer limit of our discourse, but there we have it.

So we'll see.  Hopefully Samuelson's prescription will find purchase in Congress, and therefore the DPLA will find legs.  If not the DPLA will end up linking to records of in-copyright items that people must purchase or visit a library to obtain.  Hardly the stuff of dreams, but that's not the DPLA's fault.  As Karen noted, it may turn out that the DPLA is a "wayfinder" rather than "the path itself."  That's still something.

Thoughts on the Research Works Act

The Research Works Act, just introduced in the 2012 Congress, would gut the National Institutes of Health's Public Access Policy. Furthermore, it would prevent other federal agencies from following the lead of NIH.  This is the latest incarnation of similar bills that appeared, but did not pass, in 2008 and 2009. If this passes taxpayers would always have to pay twice to access the research they funded--first via taxes, then via subscription or pay-per-view access to an article.

Advocates for open access are justifiably aghast.  As the always wise Kevin Smith points out, this is a brazen attempt to enshrine legacy publishing models into law.

So brazen, in fact, that I am not worried that the Research Works Act will pass.  I support discussion among parties that each have legitimate if irreconcilable views.  Librarians wish to maximize access, publishers must protect their business interests.  This has always been the problem. But this particular bill is stacked absurdly on one side of the ledger.

That said, there would be a silver lining if it actually becomes law.  The consequences would be so outrageous that authors may finally break free of the chains of a publishing model that dates from the 17th century.  Knowledge is what matters, not articles, and we now have new and exciting ways to spread knowledge.  Publishers should be careful what they wish for.

David Weinberger, "Too Big to Know": Talk at UC Berkeley School of Information

Last week I went to Cal for an afternoon talk by David Weinberger, author of several highly regarded books about the Internet age including The Cluetrain Manifesto and Everything is Miscellaneous. Next month his book Too Big to Know will appear, and in his talk Weinberger presented some of his ideas from that work.

You can listen to the audio file here; the full event was 80 minutes.  You can also read Quentin Hardy's write-up on the NYT Bits blog, and Weinberger's response to Hardy.  Herewith is my own take.

Weinberger's central point is that, thanks to the Web, knowledge is now lateral and associational rather than hierarchical and structured.  In an academic setting, hierarchy is asserted via peer review and the desire to appear in prestigious journals.  In a library, structure is maintained via classification schemes that attempt to render all aspects of human experience in relation to one another.  Thus, the trees and branchings of tools like the Dewey Decimal System or NLM's Medical Subject Headings (MeSH). 

The thing is, a user needs to know a lot to make effective use of a classification scheme.  In MeSH it is good to know the difference between a major subject heading and subheadings.  In Dewey, scholars of Eastern religions must know that there is very little space for Buddhism because Melvil Dewey devoted almost all of the bandwidth for religion to Christianity. This is not to fault poor Melvil--we all harbor assumptions by which we make our way.  The point is that these seemingly objective bibliographic instruments are actually highly subjective markers of their place and time.  And they make many assumptions about how people should organize and perceive information.

Weinberger gave a good example of this.  In a library catalog, to find Moby Dick you need to know either the author (Herman Melville) or the title.  If all you know is the opening sentence--"Call me Ishmael."--you are sunk, even though this is one of the most famous sentences in American literature.  Online you are liberated from these constraints, and more able to approach the world in your own way.  Type "call me Ishmael" amd many representations of Moby Dick appear.  The difference between metadata (how cataloging fields are encoded) and data (what the user actually wants to see) is fading. Today it's all data.

As to peer review, Weinberger noted this fall's posting of the important news that neutrinos are faster than light.  This news arrived not via a peer reviewed journal article, but posted to ArXiv for all the world to see.  Sure, physicists have long embraced ArXiv more than chemists or biologists.  Sure, the deeper changes required to the peer review system are political and cultural rather than technological.  Weinberger's point was different--that the very ability to conduct online post-publication peer review is a game changer.  We no longer have to rely on the vetting mechanisms developed in the pre-Web era.  If we choose not to evolve, because of inertia, that's a different matter.

The Web is intentionally chaotic and "loosely joined," to use a Weinbergian phrase.  And so the Web's very nature will challenge attempts at order.  This could be good or bad, depending on your point of view; the reality is that it's a mixed blessing. But attempts to put the genie back in the bottle--to get people to use classification schemes developed by others when they can finally use their own--seem destined to fail.

I can hear the objections now.  What about authority? What about truth?  How can librarians possibly do their jobs in a world in which everyone writes their own rules?  One answer is that librarians will need to evolve.  Weinberger--who is also the co-director of the Harvard Library Innovation Lab at Harvard Law School--suggests that librarians could move away from guiding people to particular resources and instead move people toward knowledge networks composed of other people within their communities.  I'm hard pressed to see a world in which the glorious stacks at Harvard's libraries remain standing unaltered in 50 years time. So librarians might become people brokers rather than organizers of information objects.  Maybe not, it is just one hypothesis.  Part of the solution to our future is being unafraid to try new ideas, whether or not they all pan out.

Another answer is that we've been down this road before.  Right now I'm reading Simon Garfield's fascinating book Just My Type, about the history of fonts.  Soon after Gutenberg invented the printing press in the 1400s type setters and font developers appeared all over Europe--these were the entrepreneurs who brought Gutenberg's invention to life.  And before you knew it they were printing these dastardly things called...wait for it...books!  Books!! Who needs them when we already have scrolls?  Garfield quotes a scribe in 1470s Venice, worried he would soon be out of business because the city is "stuffed with books."  Or how about Hieronimo Squarciafico (what a name!), who argued that the "abundance of books makes men less studious."  Can you imagine?  Today we are despearately trying to get kids to read printed books, precisely because we want them to be studious.  But had we lived in Europe in the 1400s books would have been feared as scary interlopers. 

Eventually, of course, books became commonplace and modern research libraries--with their cataloging apparatus and copious shelf space--were born.  If we are open-minded this time around, suffused with a creative spirit about how best to maneuver in the digital age, libraries will prosper again.  This is the whole point of the Harvard Library Lab where Weinberger works...to bring a sense of adventure to the development of the next generation of libraries.

But we will only succeed at this if we're willing to challenge the status quo. As Garrett Hardin argued in "The Tragedy of the Commons," the status quo will forever have an unmerited edge.  Hardin was speaking of status quos in the natural sciences, but his point holds more broadly.  Here is Hardin, emphasis mine:

"It is one of the peculiarities of the warfare between reform and the status quo that it is thoughtlessly governed by a double standard. Whenever a reform measure is proposed it is often defeated when its opponents triumphantly discover a flaw in it. As Kingsley Davis has pointed out (21), worshippers of the status quo sometimes imply that no reform is possible without unanimous agreement, an implication contrary to historical fact. As nearly as I can make out, automatic rejection of proposed reforms is based on one of two unconscious assumptions: (i) that the status quo is perfect; or (ii) that the choice we face is between reform and no action; if the proposed reform is imperfect, we presumably should take no action at all, while we wait for a perfect proposal.

"But we can never do nothing. That which we have done for thousands of years is also action. It also produces evils. Once we are aware that the status quo is action, we can then compare its discoverable advantages and disadvantages with the predicted advantages and disadvantages of the proposed reform, discounting as best we can for our lack of experience. On the basis of such a comparison, we can make a rational decision which will not involve the unworkable assumption that only perfect systems are tolerable."

And yet...I still love print and still treasure the feel of a book in my hand.  Part of me, therefore, loves the status quo.  But I also love the ability to download books wirelessly to my iPad within seconds; the night I heard Weinberger speak I did just that with the book Reinventing Discovery by Michael Neilsen, a book Weinberger recommended.  No waiting for it through the mail, no putting it on hold at the library, no dropping it in my backpack. Just there, right away, right now. 

Of couse, I could afford the download.  Librarians of now and the future have a HUGE opportunity in making sure that the e-reading market does not become fully commoditized and has a library/community role.  But to do this we'll have to let go of some of the practices and focus we developed over centuries of print.  We'll have to reshape our sense of purpose and in so doing forge a new (temporary) status quo.

Lastly, the Web is not wonderful in all respects.  Weinberger is among a group of thinkers--he cited Clay Shirky, who is another--who run the risk of being branded a digital cheerleader, with little awareness of the darker corners or potentially nefarious uses of the Web.  At the talk Weinberger squarely addressed this risk, acknowledging that much of how people use the Web is to watch pornography or to spawn racist, hateful theories.  People are base, and the Web makes that more plain than ever. 

This is distressing, but until humanity collectively evolves we'll be stuck with it.  In the meantime, we should embrace the liberating aspects of the Web, namely its power to gather people together and to let more voices be heard.  This is chaotic and confusing, but also wonderful.  I am not concerned that the world is too big to know. Instead I'm grateful to be living in the period of history in which I have so many new tools to learn about it.  Thanks to David Weinberger for a provocative and inspring talk.

Annotum: New Word Press Theme for Open Access Publishing

Yesterday a colleague informed me about Annotum, a new Word Press theme that facilitates open access publishing.  A great deal of technical expertise by people at Google, PLoS and NLM (among others) went into its development.  Learning about Annotum reminded me of my experience with Open Journal Systems several years ago.  Now you don't need a whole platform to make open access happen, you just need a WordPress extension.

This is all to the good.  But we can build as many tools as we want to facilitate open access and not actually get there.  Building software is hard, but it's much easier than changing the culture of publish-or-perish tenure.  That's about changing people's minds, not spiffy technology.  This is where we need to focus, no matter how slow the journey may seem.

"Decoupling the scholarly journal" by Priem and Hemminger

Over the past few years I've often returned to the topic of evolving the scholarly communication processes. In so doing I've referenced Michael Clarke's brilliant post about why scientific communication hasn't yet been disrupted.  Here's the thumbnail version (you should read Clarke's piece for the full argument):  the advent of electronic communication has not replaced all the functions of a peer-reviewed journal.  Yes, it is much easier to publish than it once was, and to note the existence of an article in an electronic database.  This solves the dissemination and registration functions of the scholarly communications chain. 

But we still have validation, filtration, and designation to contend with.  Validation is peer review; filtration slots the article within a recognizable, branded journal that scholars know to read; designation is the space on the CV where a scholar lists all their publications in high impact factor journals.  The impact factor may be highly dubious, but it is still highly used.  These latter three functions are less amenable to electronic solutions, which is why the scholarly communications system remains conservative.

These five functions are bundled together within every single journal, so any proposed change to one part of the chain must simultaneously account for all of the other functions to have an impact.  We have a systems problem on our hands.

Along come Jason Priem and Bradley M. Hemminger, with an innovative proposal to "decouple" the traditional functions of a scholarly journal as a means of solving this systems problem.   There is no reason that every single journal must continue to facilitate all of these functions, it is just the way it has always been done.  Priem and Hemminger note that our current template for scholarly communication dates to 1665, when scholars who had previously mailed letters to each other agreed to mail the letters to a central location so they could be bound together into the print journal.  That was very innovative then, less so now. 

Priem and Hemminger: "We should not expect a fully decoupled metajournal to emerge in the next year or even decade. However, neither should we expect the current system, based as it is on the paradigm of the 17th century, to continue with only small, evolutionary changes. There will be a revolution in scholarly communication, as the fundamental potential of the Web compared with traditional models puts increasing torsion on our system."

How do we embrace and harness this torsion?  By "decoupling" the functions now maintained journal-by-journal, using the Web to manage publication processes instead.  It would be very easy to post an article online, get initial feedback, and revise it before final "certification" by a qualified body. In this scenario there are three services--hosting the article; peer review; final certification.  But here they are managed in a "Web-like environment of loosely joined pieces," rather than in a vertically integrated publication process.

Separating the pieces of the scholarly communication system apart makes it possible to fix/improve/modify one element without disrupting everything.  The market can work its magic, as scholars pick and choose among registration services rather than simply submitting to the same old journals.  The unit of discourse becomes whatever the scholar wishes to share, rather than only the article.

There is a lot of interesting work for librarians in this vision, to manage a scholarly communications chain that facilitates the ebb and flow of discourse rather than only presenting polished final products.  One can quibble with the details Prier and Hemminger present. But their logic of separating out the functions of the system, to provide space for innovation, is unassailable.  This is much easier said than done, but first someone has to say it.