[This is a rough first draft of a presentation for the STM conference in Budapest in May 2006. The material here has not been reviewed or edited. It is being posted to the PBOS site to give it an address on the Web. Comments on this document are invited. Please send them to me at firstname.lastname@example.org or use the annotation tools on the PBOS site to place comments directly on the document (the latter option is preferred). For anyone who is viewing this document in a context other than the PBOS site, note that PBOS stands for Processed Book Operating System and the site is located at http://prosaix.com/pbos. This material is copyright (c) 2006 by Joseph J. Esposito.]
Several years ago I wrote an essay entitled “The Processed Book” in which I attempted to define various aspects of the book in the digital era. One of these aspects was the Platform Book, a dimension of an electronic book in a networked environment in which other books, notes, and commentary were built upon the original book, the platform. The word “platform” was intended to be a pun, however bad, on a platform in the computer world. Such a platform (Microsoft Windows is the best example) is something that other kinds of software are built upon, and that other software, the applications, have the ability to “call” or invoke various properties of the platform. A Platform Book is thus a book that can be “called” or invoked. As new work is created, whether that work is an annotation or even another book, the Platform Book sits behind it or beneath it.
A Platform Book is distinct from a Portal Book, which is a translucent text, both providing content of its own and passing a user through it to other, third-party content. A Platform Book is distinct from a Recursive Book or Self-referencing Book, which is a book that comments on itself, a book wholly of metadata. A Platform Book is distinct from a Robotic Book, in which the content of a book is built into a computer algorithm, making the book into a component of a machine. An example of a Robotic Book can be found in search engines, which silently incorporate such things as dictionaries and thesauruses. A Platform Book is also distinct from the Nodal Book, in which the book is part of a connected network of information. A Nodal Book potentially points in many different directions (it can be both Portal and Platform, for example), whereas the work of a Platform Book is to enable others to work with it and on top of it. And a Platform Book is distinct from a Primal Book, which is the antithesis of a Processed Book; a Primal Book is traditional in character, even if it is rendered in electronic form. A book is primal when it is self-contained and appears as the fixed expression of an author. A Platform Book could start out as a Primal Book, but as a user base develops around the Primal text, invoking it in various ways (the most obvious way being simple quotation--O brave new world that has such books in it!), the Primal Book comes to be at least partly a Processed Book, taking on aspects of a Platform Book.
One other point about categories before moving on. The Processed Book and its aspects (including the Platform Book and the Nodal Book) are different from the concept of the networked book that is currently being advanced by Bob Stein’s group at the Institute for the Future of the Book. I much admire the thinking behind the networked book, though it derives from a different set of interests and emphases. The networked book, a la Stein, is very much oriented toward production, technology, and users. The Processed Book, on the other hand, is oriented more toward the creation of content than its use and can be relatively limited in its technical aspect if its creative impulse has been satisfied. Another way of putting this is that the Processed Book is an upstream concept, whereas the networked book is downstream. The concepts are thus usefully complementary. In any event, one hopes that reasonable people can disagree, though in the Internet Age, that does not always appear to be possible.
What I would like to do here is spend some time developing the idea of the Platform Book. Rather than leave it as an abstract category, however, I want to think about how to make the Platform Book a reality, and by a “reality” I mean a commodity, something that exists in the real world where it can be bought and sold. If the Platform Book can be bought and sold, capital will be invested in it. If capital comes in, innovations will flourish. A Platform Book, or any other kind of theoretical notion of publishing, that can be tested in the real world has a better chance of surviving, even as a concept. The meme of the Platform Book needs a host in the marketplace.
There is another reason as well to be thinking about the Platform Book, and this has to do with my kinship to book publishers, who are now struggling to find growth. Except for isolated pockets, the book business today is a mature business. The growth of one company mostly comes at the expense of another. Growth seems forced. One way to grow is for one company to gobble up another--but though that increases the revenue of the predator, it does not enlarge the book business overall. Or a publisher might grow by developing various means to increase prices, sometimes deviously. An obvious example of this is in the journals arena, where a handful of publishers have found ways to boost sales through aggressive subscription pricing. A less obvious way, but infinitely more clever, can be found in the consumer book market, where inexpensive mass-market paperbacks have largely been replaced by higher-priced trade paperbacks, some bearing prices that one would have expected to see only on hardcovers only a few years ago. But unit sales? At least in the U.S. at any rate, they seem to have plateaued at around 2.5 billion a year, sold at an average price (net to the publisher) of around $10. If investors are faced with the choice of putting money into a new aspect of information technology, alternative energy, genetic engineering, or book publishing, to whom do we expect them to make out the check?
Book publishers are struggling mightily and creatively to punch their way out of the small-growth or no-growth situation. I will get into some of their strategies below, but allow me to note here that most attempts at growth are likely to be modest in nature, as they deal in mostly incrementalist tactics--such things as new and improved marketing plans or the “repurposing” of content, which is essentially an attempt (and a sensible one) to leverage sunk investments in content development. The sad truth, though, is that the traditional book has fallen from its proud position in cultural history and the economy; like Humpty Dumpty, it can never be put back together again. A path to robust growth requires a more radical solution, one that seeks growth not through tweaking and refinement but through innovation. One path to growth is the Platform Book.
There are a number of broad trends in the environment today that have to be taken into account in planning any publishing venture, whether in print or in electronic form. A Platform Book, if it is going to be successful, will have to address at least some of these. This does not mean that a publisher cannot be successful by swimming against the current, but it is simply easier to grow when the market is growing with you. Many of these trends are further along in the journals area, but they are emerging for books as well.
Undoubtedly other people would come up with other trends to highlight here; this list is by no means definitive. Nor does the list have some clear coherence or hierarchy. I have tried to order this list in a number of different ways. For example, should the universal interface be clustered with disintermediation or does it occupy a category of its own? None of the taxonomies I came up with satisfied me, however. I am inclined to think that these trends, and others I did not name, are all organically connected and draw their energy from underlying social and economic phenomena, most notably the increasing tendency to see ourselves and the world as parts of networks: social networks, networks of commerce, and, with the Internet, networks of networks. The good news for publishers is that we don’t have to understand the underlying issues. Rather our task is to test the current, get out into the strongest channel, and paddle just a little bit faster than the next fellow, who may be distracted by looking backward at a legacy business of printed books.
If we were to gather all these trends together and start with a blank sheet of paper, what kind of publishing business would we come up with? In one such formulation we would have a product--actually a service--that was created by a community that was sold as a subscription online directly to end-users. It would likely have an Open Access component and users would most often find it on the Internet through the mediations of a universal search engine, which in 2006 most likely means Google. This raises two interesting questions: Is there any product or service in existence that has this profile, and are the steps that publishers are currently taking leading in this direction?
There are so many things going on in the publishing industry today that it is impossible to keep up on them all. About two months before I began to draft this paper, I commented to someone in the content-management software business, who has a number of STM publishers as clients, that I was surprised that despite all the hoopla over Google’s various attempts to index books and academic material, no one had yet struck a deal with CrossRef, which would have been the first place I would have gone. Well, great minds work alike: six weeks later an announcement appeared to the effect that Microsoft was launching Windows Live Academic Search with the aid of CrossRef. This is but one example of many, where illustrations get overtaken by events. For this reason I will try to summarize some developments in publishing at a relatively high level so that the overarching point concerning trends is not undermined by an ill-timed example.
This list can go on and on; no doubt if we added enough elements to it, we could then engage in an argument as to which of these actions are strategic, which tactical, which are likely to lead to top-line growth, and which are useful for cost-savings, building the bottom line, though not necessarily the top. What is perhaps more interesting, however, is to map these actions against the broader trends outlined earlier. How closely aligned are publishers’ actions with the broader trends within their industry?
Our paradigmatic publishing endeavor would be created by a community, have an Open Access component, be sold in some way on a subscription basis directly to end-users, and it would be in electronic form. At least that is one way to construct the paradigm; there are no doubt others. Publishers that are working in hardcopy, even if the products are delivered online, are bucking this trend. Publishers without an Open Access component (and very few have a revenue model encompassing Open Access) or are not working with community-created content are not swimming with the broader currents in the industry today. Some publishing activity is aligning itself with these trends, however, such as the direct sale of books from a publisher’s own Web site and the increasing use of search-engine optimization, the first step toward the acknowledgment of an emerging universal interface. By and large, however, in fairness it has to be said that publishers are embracing the broader trends in the information industry only timidly.
There are very good reasons to be timid about the broader trends in the industry. To cite two: many of the trends seem to be devoid of economic sense; and legacy publishers do indeed have a legacy to protect. And here is the conundrum: although the broad trends do not promise clear economic benefits and may indeed put the legacy businesses at risk, failure to pursue these trends almost certainly puts a publisher back in the slow-growth or no-growth situation. A publisher that wants growth must invite risk and innovation.
It is hard to be very encouraging to publishers who choose to adopt a strategy that does not embrace a paradigm based on emerging trends, not only because the growth prospects are at best modest, but also because every other publisher is doing the same thing. This is an important point: it’s not enough for publishers to run with the pack; they have to run ahead of it. Someone whose strategy is based on a bit of content management, some Web bookselling, and the occasional acquisition is going to find it very hard to differentiate his or her publishing operation from another. What differentiation there is will be largely of the legacy variety: editorial discrimination. It’s hard to argue with this (who doesn’t want to publish the best books?), but it is also hard to outpace the competition in this way.
This leads to a situation where investors begin to say, Do you really need all that cash? What growth areas are you investing in? The pressure builds for the management to turn over more and more cash to owners; balance sheets get leveraged; and ultimately the financing is not there to pounce on new initiatives when they appear. This situation has been well-analyzed in Clayton Christensen’s The Innovator’s Dilemma, and I won’t pursue the point further beyond saying that many publishing companies are inviting start-ups into their market by failing to think and act like entrepreneurs.
The most radical of all the emerging trends listed above is community-created content. Such content goes right to the heart of what historically makes publishing publishing: it removes the author from the equation and, in doing so, posits the disturbing idea that one can have authority without an author. Now, the elimination of authority has a certain appeal to anti-establishment types, but community-based content has its own regulatory regime, one based not on the voice of the expert but on the rules and procedures of a community united by a common cause. The truly astonishing thing about the Wikipedia (to take the most famous example) is not whether a particular article is any good, but the process that gives rise to an article that may indeed be good. Wikis are about the process that gives rise to content, not the content itself; and the content that is formed through this process is not allowed to set for long, as a new idea, a new perspective comes along that adjusts the previously “set” or fixed version. This is, of course, anathema to anyone who was raised in traditional publishing (as I was), but the closer one observes this evolving phenomenon, the more merit it appears to have.
But, someone will insist on insisting, is it any good? Well, yes and no. The Wikipedia is good in its breadth, covering far more topics than any comparable reference; and it is good in its timeliness, updating events as they happen, whether they take place in Iraq, Yankee Stadium, or a science laboratory. It is bad in its balance, in its occasional admission of incompetent or malicious contributors, and in its voice, which is jarringly inconsistent, sometimes within the same sentence. Many of the pieces lack real seriousness, something that is only evident to someone who has knowledge of the topic under consideration. But is it any good? Well, it flows in a good direction.
What is more important, though, is not whether it is any good but how it could be better, a question that can and should be put to far more works than just the Wikipedia. If a publisher can solve this problem, then this would be a true innovation.
The Wikipedia falls short, when it falls short, for two reasons: the uneven talents of its contributors and its apparent boundlessness. As for boundlessness, where, precisely, does the Wikipedia end? Unlike its unavoidable comparison with hardcopy reference works and Encyclopaedia Britannica in particular, the editorial guidelines for the Wikipedia potentially allow for a continually expanding definition of what belongs in a reference work and what does not. No print encyclopedia would ever think to include the results of last night’s Yankee game, but in an online product, why not? And with a community determined to keep things up to date, even late-breaking sports scores are potentially the stuff of reference publishing. Some of the editorial policies of hardcopy reference publishers sound odd nowadays, as they were created in an era when information had to be fixed in place. So, for example, there is Merriam-Webster’s Biographical Dictionary, an outstanding work. Thumbing through its pages, however, a prospective reader may wonder at all the names that are not included. Napoleon gets an entry, but not Bill Clinton. Yes to Jesse Owens, but no Barry Bonds? No Vladimir Putin, Margaret Thatcher, or Saddam Hussein? It takes a while for it to sink in that in order to get into this particular book, you have to die first. That’s one way to make sure that there will never be a need to revise an entry.
The uneven talents of the Wikipedia’s contributors (and the right word is uneven, not poor: some of the contributors are outstanding, but the problem is that a reader does not necessarily know this) are a function of the wide-open policies of this particular instantiation of the wiki form. Anyone can read the Wikipedia, and anyone can write for it. So who, precisely, comprises the “community” that is at the heart of the Wikipedia? This is the crucial question for community-created content: who is a member of the community? And this is precisely where publishers can play a role, by developing means to define a community and thus facilitate community-generated content not by restricting access but by restricting membership.
Let’s spend a minute on the term “community.” A community can be big or small, universal or tribal. Often the term becomes laden with political ideology. For example, in my hometown of Santa Cruz, CA, the word “community” is mostly used to refer to one segment of the population: people with little money who don’t own a house or commute to nearby Silicon Valley. No one in this town talks about the community of real-estate developers or the community of technology entrepreneurs or the community of affluent retirees who enjoy the temperate climate and the views of Monterey Bay. In Santa Cruz, in other words, the term “community” is mostly associated with the constituents of the area’s politically progressive activists, a very large number since this is, after all, a college town.
A community is larger than an individual, larger than a couple or a family. It is most often defined in terms of common interests, as the community of scholars or the community of auto-racing enthusiasts. All of us belong to multiple communities, though the connective tissue for some communities is more attenuated than for others. We all belong to the community of humankind, though few of us are conscious of our connection to that community on a regular basis in our day-to-day lives. I happen to belong to a loosely defined professional community, the community of people interested in the strategic implications of digital media. This community membership is not at odds with some of the goals or activities of other communities I belong to, such as that of parents of teenagers in our local public schools or of people with whom I have some kind of political affiliation.
The problem at the heart of the Wikipedia is that the community is too broadly defined. Membership is open to anyone with a Web browser. I could usefully contribute to articles about digital media, and I could beneficially read articles on some aspects of the sciences (maybe), but I cannot contribute articles on particle physics or any number of other fields where specialized knowledge is essential. This is different from having opinions on different fields, where an opinion is a point of view that is not supported by expertise. I have opinions on American military involvement in Iraq, but it would be foolhardy for me to pretend that I could make serious recommendations about the form and actions of troop deployments. The Wikipedia does not discriminate. The fundamental error of the Wikipedia is that it confuses the political rights of an individual with the qualifications to participate in intellectual discourse. And as to the argument that the credentialing bodies that declare one individual an expert and another a fool, are corrupt and out of touch, well, yes, there are mistakes. Not every college student earns his or her degree, not every Yale graduate has the talent to go on to become president.
Publishers can define communities of interest in part because this is what they have always done. They know that the market for a book on genetic engineering is distinct from a market on search-engine algorithms, and they know how to evaluate and attract experts in their respective fields. The opportunity for publishers is to take the audience and authors for particular topics and convert that group into a self-conscious community that works together to create and maintain community or civic resources. A researcher in genetic engineering is a very good person to writer about genetic engineering, but the same individual may be a poor choice to offer an article on professional hockey or a textual analysis of Chaucer. If we gather the experts on genetic engineering together, we now have a community of genetic engineers. Communities created in this way get around the two problems inherent in the Wikipedia. The contributors are competent by virtue of their belonging to a narrowly defined professional community and the boundlessness of the Wikipedia is replaced by the more targeted concerns of a professional.
Membership communities, as distinct from open communities, also sidestep the problem of Open Access, which is not a small problem in some contexts. OA poses serious economic challenges to publishers whose fundamental business model depends on the benefits of content being conferred primarily on its readers--who, thus, are asked to pay for access. In a membership community content is created and consumed by the same individuals. If the particular domain of this community is a specialized topic such as orthopedic surgery or radiology, the likelihood of outsiders, people who are not members of the community, wanting to get access is small.
Publishers exploring community-created content will see that this form is both seditious in that it undermines the traditional notions of authors and authority and also not all that extreme, provided that the community is thoughtfully defined on less than universal terms and membership requirements are enforced. For those publishers who have to pay large sums to authors, community-based content may also provide some economic relief in that authors, rather then being paid, will pay to become members. For a publisher the switch from the traditional author-driven model to the community model involves a change in emphasis, from selecting the best author to facilitating the communications of a virtual gathering of experts in a particular field.
Theoretically even a novel could be a Platform Book because even a novel can be thought of as a kind of reference book, but as a practical matter, it is material that is created for reference to begin with that presents the best opportunity. The short entries of a reference work counter the difficulties in sustaining a narrative by a community of writers, and a reference work is defined as something that is referred to; there is a long tradition of writers citing (calling) reference works in their own work (“According to Webster’s dictionary . . .”). For all its limitations, the Wikipedia demonstrates this point very well. There are a stunning number of references to the Wikipedia now, so many that some texts seem to be a collection of hyperlinks to the Wikipedia. New texts create narratives or whatever size, but the information objects embedded in those narratives increasingly draw on the Wikipedia. The task for publishers is to define a field where a good reference work is needed and to experiment with community-content tools to build the reference.
Let’s take an editorial leap and come up with a prototype for a Platform Book, which we will playfully call here the Dais Ex Machina (DEM). To be sure, this is not the only kind of Platform Book that could be concocted, but it has the virtue of providing some concrete detail for people to look at and comment upon.
Our Platform Book is going to be an online reference book in some undetermined area of science, technology, or medicine. As a reference book it will include a great number of relatively short articles. The content itself will be created by a scientific community; the software used will be a form of wiki software. The reference work will be a kind of subject-matter encyclopedia; just what subject is appropriate we will discuss in a moment. But because our Platform Book is in the scientific area, each entry is liable to require constant updating (which, of course, is what the wiki form is best at). This raises the tantalizing possibility that each article could be a synthesis, as it were, of all the current thinking on a particular scientific topic. The author of a particular article would review all the literature in that particular area and summarize it. If a particular author left something out or wrote about something that was subsequently superseded, other members of the community would make adjustments. If an author had an unwelcome slant, other writers would attempt to redress the balance. It is entirely possible that arguments could become fierce, but insofar as this is a membership community, there is likely to be some moderation of extreme or reckless behavior, just as it is unlikely for a fist fight to break out in a faculty club.
Choosing the specific domain for DEM is not going to be easy. It can’t be as broad as the Wikipedia for any number of reasons, the greatest being that it is its sheer lack of discrimination that creates many of the Wikipedia’s problems. But how narrow can the domain be? One starting point is with research journals; around every research journal there is a community of scholars. There are 24,000 peer-reviewed journals, including humanities journals, but it is doubtful that there could be the same number of community-based reference works. For most topics there is usually only a single reference, and many of the 24,000 journals compete with each other. So the total number of potential community-based references is some subset of 24,000. It is likely that the number will continue to shrink as we examine the situation further, as some journals cover truly narrow topics, too narrow to support a reference work. The final determination is going to vary with the publisher, who will examine the company’s areas of expertise and access to authors.
To get a reference work started, a publisher will have to create a flexible outline. The operative word here is “flexible.” Since this is going to be a wiki, flexibility is everything. On the other hand, without some starting-point, authors would not know where to begin. The publisher may choose to commission a certain number of articles to prime the pump, as it were; such articles would establish precedents for length, tone, citations, etc. The outline must include the ability to revise the outline--to add new topics, to adjust the taxonomy, to determine when an article should be subdivided. Authors would be invited to add to the outline, to write articles in accordance with the outline, and to edit articles already written.
At this point we should assert a key point about managing such a community: every author must sign his or her name every time he writes or edits for DEM and each signature must link back to the author’s biography. There are multiple reasons for this. The first and obvious one is that such identification will suppress particularly poor behavior. Another reason is that such identification becomes a primary motivator for an author to be involved with the work to begin with. Authors will want to be known as the person who wrote the definitive entry on a particular topic; and to keep their names identified with that topic, they will monitor the emerging literature and continually update the article. An author who does not stay on top of a topic will find new authors will enter the fray, editing and revising the piece and thus associating their names with the topic. This form of professional rivalry is good for the academic discipline and it can be good for the publisher. It can work for the simple reason that reputation, not money, is the primary currency of most researchers.
The benefits of the DEM to readers (who are also community members) are partly the traditional benefits bestowed by good reference works. A reference book provides core information, of course, but beyond that a good reference is also a map of a particular field. The form of the reference that I am proposing here goes beyond that as well by taking advantage of the real-time updating features of an online wiki: the reference is not only a backgrounder but also a repository of late-breaking research, which has been integrated into the reference’s taxonomy. Insofar as each article is a synthesis of the literature, another benefit is that it permits a researcher to review a topic quickly, sometimes reducing or eliminating the need to read all the back-up literature. Of course, links to the primary literature would be part of the service as well, assuming that the proper business arrangements can be put in place. (This is a matter of editorial opinion, but in my view, a good reference work is a self-contained universe, linking only to other sections of the same work. The art of article-writing is in presenting a clear and comprehensive summary, which should require no external support.)
From an editorial point of view, the DEM is thus far a community-created wiki in a particular scientific domain. The editorial form is that of a reference work, with each article being a synthesis of the current thinking on a particular subject. But this does not make it a Platform Book; it simply makes it a wiki. It becomes a Platform Book when other works “call” it. This requires placing the prototype into its living context, where it will be marketed.
A Platform Book has several potential revenue streams:
It is when we reach business-to-business subscriptions that the membership-created wiki crosses the line and becomes a Platform Book. (Theoretically an individual may seek such a license as well, but as a practical matter the cost of managing individual licenses is likely to be too high.) Included in this subscription will be a license granting certain privileges for reuse, not unlike what one typically finds in most open source software licenses. This is important: I am proposing here something that goes far beyond simple linking and includes the incorporation of extensive portions of text into authorized works. It may be that the licensee will also have the right to resell content from DEM, provided that content remains within the context of the licensee’s original work. While publishers will naturally fear creeping piracy, the fact that the original work is being updated continuously undermines the value of the pirated work. The cure for piracy is not digital right management but dynamic content.
DEM or any Platform Book, in other words, has to do two things in order to be effective (above and beyond providing valuable content to begin with): it has to be structured technically in such a way as to permit third-party licensees to use and reuse its content easily and efficiently, and it has to provide prenegotiated terms for the business arrangements. Observers of the open source software world will recognize in this formulation some of the aspects of open source licenses, though it should be noted that there are many kinds of open source licenses, not all of which would be suitable for the Platform Book as proposed here. For example, the best-known open source license, known as GPL for “General Public License,” which was invented by Richard Stallman, would not apply here, as it sides too much with the reusers of intellectual property, whereas the Platform Book attempts to straddle the interests of the creators and the reusers. Coming up with the right form of the license and communicating its salient points easily and effectively is as much a part of the success of a Platform Book as good content and robust technology.
How does DEM or any Platform Book align itself with the currents in the business environment? It meets some of the challenges, fails to meet others, and meets yet some others halfway.
Alignments with broad environmental trends, of course, are no guarantee of economic success, but at least publishers who look in this direction will have the wind at their backs. The point can be made in different ways, but the plain truth of the matter is that publishers seeking growth will find the Humpty Dumpty book of yesteryear, no matter how it is patched, pushed, and computerized, of no avail. To put the matter starkly, I defy any established book publisher to post year-over-year growth in the double digits without resorting to price increases or the acquisition of other companies. The only path for the book industry to sales growth is product innovation.
Proprietary material; “The Platform Book,” copyright (c) 2006 by Joseph J. Esposito
Copyright (c) 2006 by Joseph J. Esposito