can there be great textbooks without great authors? 10.27.2005, 8:08 AM
posted by kim white
Jimmy Wales believes that the Wikibooks project will do for the textbook what Wikipedia did for the encyclopedia; replacing costly printed books with free online content developed by a community of contributors. But will it? Or, more accurately, should it? The open source volunteer format works for encyclopedia entries, which don't require deep knowledge of a particular subject. But the sustained examination and comprehensive vision required to understand and contextualize a particular subject area is out of reach for most wiki contributors. The communal voice of the open source textbook is also problematic, especially for humanities texts, as it lacks the power of an inspired authoritative narrator. This is not to say that I think open source textbooks are doomed to failure. In fact, I agree with Jimmy Wales that open source textbooks represent an exciting, liberating and inevitable change. But there are some real concerns that we need to address in order to help this format reach its full potential. Including: how to create a coherent narrative out of a chorus of anonymous voices, how to prevent plagiarism, and how to ensure superior scholarship.
To illustrate these points, I'm going to pick on a Wikibook called: Art History. This book won the distinction of "collaboration of the month" for October, which suggests that, within the purview of wikibooks, it represents a superior effort. Because space is limited, I'm only going to examine two passages from Chapter One, comparing the wikibook to similar sections in a traditional art history textbook. Below is the opening paragraph, framing the section on Paleolithic Art and cave paintings, which begins the larger story of art history.
Art has been part of human culture for millenia. Our ancient ancestors left behind paintings and sculptures of delicate beauty and expressive strength. The earliest finds date from the Middle Paleolithic period (between 200,000 and 40,000 years ago), although the origins of Art might be older still, lost to the impermanence of materials.
Compare that to the introduction given by Gardner's Art Through the Ages (seventh edition):
What Genesis is to the biblical account of the fall and redemption of man, early cave art is to the history of his intelligence, imagination, and creative power. In the caves of southern France and of northern Spain, discovered only about a century ago and still being explored, we may witness the birth of that characteristically human capability that has made man master of his environment--the making of images and symbols. By this original and tremendous feat of abstraction upper Paleolithic men were able to fix the world of their experience, rendering the continuous processes of life in discrete and unmoving shapes that had identity and meaning as the living animals that were their prey.
In that remote time during the last advance and retreat of the great glaciers man made the critical breakthrough and became wholly human. Our intellectual and imaginative processes function through the recognition and construction of images and symbols; we see and understand the world pretty much as we were taught to by the representations of it familiar to our time and place. The immense achievement of Stone Age man, the invention of representation, cannot be exaggerated.
As you can see the wiki book introduction seems rather anemic and uninspired when compared to Gardner's. The Gardner's introduction also sets up a narrative arc placing art of this era in the context of an overarching story of human civilization.
I chose Gardner's Art Through the Ages because it is the classic "Intro to Art History" textbook (75 years old, in its eleventh edition). I bought my copy in high school and still have it. That book, along with my brilliant art history teacher Gretchen Whitman, gave me a lifelong passion for visual art and a deep understanding of its significance in the larger story of western civilization. My tattered but beloved Gardner's volume still serves me well, some 20 odd years later. Perhaps it is the beauty of the writing, or the solidity of the authorial voice, or the engaging manner in which the "story" of art is told.
Let's compare another passage; this one describes pictorial techniques employed by stone age painters. First the wikibook:
Another feature of the Lascaux paintings deserves attention. The bulls there show a convention of representing horns that has been called twisted perspective, because the viewer sees the heads in profile but the horns from the front. Thus, the painter's approach is not strictly or consistently optical. Rather, the approach is descriptive of the fact that cattle have two horns. Two horns are part of the concept "bull." In strict optical-perspective profile, only one horn would be visible, but to paint the animal in that way would, as it were, amount to an incomplete definition of it.
And now Gardner's:
The pictures of cattle at Lascaux and elsewhere show a convention of representation of horns that has been called twisted perspective, since we see the heads in profile but the horns from a different angle. Thus, the approach of the artist is not strictly or consistently optical--that is, organized from a fixed-viewpoint perspective. Rather, the approach is descriptive of the fact that cattle have two horns. Two horns would be part of the concepts "cow" or "bull." In a strict optical-perspective profile only one horn would be visible, but to paint the animal in such a way would, as it were, amount to an incomplete definition of it.
This brings up another very serious problem with open-source textbooks--plagiarism. If the first page of the wikibook-of-the month blatantly rips-off one of the most popular art history books in print and nobody notices, how will Wikibooks be able to police the other 11,000 plus textbooks it intends to sponsor? What will the consequences be if poorly written, plagairized, open-source textbooks become the runaway hit that Wikibooks predicts?
dave munger on October 27, 2005 9:18 AM:
This is one of the many reasons why I think textbooks should be open source, but authors should be paid. How? Well, state educational boards could hire publishers to create books -- nearly the same way they do now, but instead of publishers retaining copyright, publishers would be doing the work under contract, promising to release the work under the GPL, a Creative Commons license, or some other public license allowing reprints free of copyright fees. In the first years, this would cost as much or more than a regular textbook, but down the line, much could be saved.
If the state needs 5,000 more books the next year, they need only pay the printing costs -- say, $5 per book, instead of $50. If they wanted to change just one or two chapters, they wouldn't have to pay for revising the entire book. And if they wanted to produce an online or CD version of the book, this could be done essentially for free.
Such texts could then be used to populate those $100 computers we've heard so much about, but they'd be high quality texts, because they would have been subjected to the usual peer review process, not just slammed together in a wiki.
charlie on October 27, 2005 1:55 PM:
I'll disagree about the payment process. Not that it might not be a bonus to the project, but the problems suggested above are a consequence of the process.
liblamb on October 27, 2005 5:35 PM:
Your understanding of collaboration of the month isn't totally correct. You say, "This book won the distinction of "collaboration of the month" for October, which suggests that, within the purview of wikibooks, it represents a superior effort." Yet, the Collaboration of the Month Project page says only that "The hope is that the Wikibook will advance quickly over that period." It doesn't lift the book up as a good example. In fact I think the opposite is true. Books that really need work are nominated for collaboration of the month. Was there maybe confusion with the featured articles from Wikipedia?
That being said, the plagiarism (which I can't validate but believe you) was up for 8 days and that is a problem. It's still up but I put it on the discussion page for evidence to be provided before complete deletion.
I have no doubt that the text books produced by Wikibooks will be different than single author texts. The question is how will they be different and how will that affect instruction/learning?
kim white on October 27, 2005 8:30 PM:
It doesn't lift the book up as a good example. In fact I think the opposite is true. Books that really need work are nominated for collaboration of the month.
I stand corrected, thanks for clarifying that point. I wonder, though, if the selection process for these featured books might be a good place to require scrutiny of the contents so far. The collaboration of the month page also says that Collaborations of the Month are selected by a simple majority vote from books nominated by registered Wikibooks editors. That means that a lot of people (including at least one registered editor) looked at and voted for "Art History," but didn't notice the plagiarism.
I agree with Dave that some kind of compensation model might motive writers and scholars who are currently working with traditional publishers to make the move to digital. I also agree with Charlie that there are some fundmental flaws in the process that need to be addressed.
Nick Carbone on October 27, 2005 9:21 PM:
It's not only great authors that elevate prose, it's also great editors working with those authors, producers and designers bringing the book together, compositors, marketing managers and sales representatives helping the book finding an audience so that it doesn't die on the vine, and lots of other people working with and for the author.
The advantage to commercial book production is that it has professionalized some very important skills. In part because the economics of print require books --especially textbooks, which are among the most difficult books to write and produce-- to be right in order to go on. And it takes time to get a book right and patience.
This is not to say that an open source/community model might not evolve to get similar results over time, but really, it will take having someone in charge, someone making editorial decisions, someone researching and double-checking, and so much else.
You might find some books where volunteers can pull this off well, but they're likely to be the exception. If the economics of it all change, then you might have a future with no professional produced books and world where, perhaps, books-via-anonymous-community find a way to be as well written as books by dedicated authors (who have a royalty in mind) working with professional editors.
I do think such a thing might be ossible. But I don't think the Wiki model is going to get it there.
dave munger on October 27, 2005 9:36 PM:
Hi Nick, Hi Charlie -- great to see you here, what a blast from the past.
Anyway, I don't think we should give up on the wikibooks concept, but one big problem with wikis is the very different role of authorship. I can see a model working very similarly to the current scholarly publishing model, where authors get credit for contributing to open source books. I can also see a model like I described above, where authors are paid, and so it works in a similar way to commercial textbook publishing.
However, being involved in a wiki is a very different thing, and I have a hard time figuring out how scholarship in wikis can get recognized and rewarded -- especially since it can get deleted in the blink of an eye -- and since such obviously shoddy work can be "published." The wiki model is exciting, but in the end, making textbooks is just a lot of hard work. Is anyone willing to do the work it takes -- and to be so selfless as to not be particularly concerned about whether that work is recognized? I suppose it remains to be seen.
Juan Lopez-Valcarcel on October 28, 2005 12:20 AM:
Interesting comment in terms of setting up a collaborative payment process.
However, I think the key issue is how to balance subjective views as to what should be the "right interpretation" of literature, art or history.
The wikibooks model arguably shows more potential for maths, language and other less controversial or less opinionated subjects.
In any case, an excellent initiative indeed that hopefully will one day lower the costs of textbooks.
gary frost on October 28, 2005 8:56 PM:
I like the Wiki text book model provided that it is used by students in a given class to produce their own textbook. This approach would have a number of instructional advantages, the first of which is that it would induce the students to read their textbook.
ben vershbow on October 29, 2005 11:10 AM:
I agree that getting students in on the production of the book is perhaps the most distinct advantage of the Wikibooks idea. The task of creating a good knowledge resource leads to real engagement with the material. Take as an example this project that we featured on Next\Text. It's not so much a textbook as a photo essay anthology. But it's in a wiki, and I think suggests the power of this idea.
charlie on October 29, 2005 7:15 PM:
Dave and Nick,
Right, Dave. And I think I'm sort of in aggreement with Nick here, too, although I believe that open source is capable of creating a professional quality product. The production model used in wikis, I think, does not lend itself well to being recognized for one's work. In open source communities, people get recognized for strong contributions and quite often become leaders in the community, recognition which does translate into career advancement and opportunities. Open source communities don't have the "anyone can edit the code" model of wikis. Rather, individuals contribute suggestions for changes which are then evaluated before being applied to the source. That process of both being evaluated and evaluating is also part of the social process for gaining stature in the community. With wikis, that process is much less visible, partially because peer review occurs because someone edits the already made contribution, not because of an involved discussion about making changes. Maybe I'm wrong here about the peer review aspect of wikis, but making the process transparent is important, I think.
Meanwhile, the other thing that wikibooks (and indeed, wikipedia) could do is move toward the unstable (experimental) / stable (production) versioning of Linux and other open source projects. I know that wikipedia locks pages, but does it also have an unstable version of those pages under development? In open source, once a project has reached a certain point, a stable version is always maintained for public consumption while developers work with the unstable version. At some point, the unstable version reaches feature freeze, at which time it is polished up to become the new production version, and at which time, contributors then begin working with a new unstable version. The problem with wikis is that the unstable and stable version seem to be the same.
So if I were going to create a good collaborative process for wikibooks, I would start with contributors proposing 250 word abstracts for new book pages/sections. Once those abstracts were created, they would be the first stable version and community members interested in working on that section would start an unstable version elsewhere. Perhaps the goal for the unstable version would be 1000 words. Once that one was stable (and properly peer reviewed), the next might be a 4,000 word entry. Or perhaps a revision of the 1000 word version that reflected a different rhetorical strategy and writing style. What's important about this is that each unstable-stable cycle be composed of incremental changes in scope, but not wholesale "let's write the perfect 7,000 word entry right from the beginning."
kim white on October 31, 2005 1:19 PM:
With wikis, that process is much less visible, partially because peer review occurs because someone edits the already made contribution, not because of an involved discussion about making changes. Maybe I'm wrong here about the peer review aspect of wikis, but making the process transparent is important
I agree with Charlie on this and I would go a step further and say that the fact that one writer/editor can change the entry does not really correllate with peer review. Contributors are not necessarily peers. A high school student could replace the writings of a college professor and make the entry worse not better. To make improve the process I would suggest vetting contributors. Each writer should provide a complete and verifiable profile that is made available to the reader. For example, when I click on the link for ExtraKryspi, who I believe was entered the plagiarized text, I get a message reading: You've followed a link to a page that doesn't exist yet. What I need to know is, who is this ExtraKryspi and why should I entrust my understanding of Art History to him or her? Warranting is a huge problem for open source book projects and solutions need to be in place before we can engage with them in a serious way.
Also wanted to note that the plagiarism I cited has been removed from the book. That kind of responsiveness is one positive aspect of the wiki book.
liblamb on October 31, 2005 4:59 PM:
Kim, the page that doesn't exist yet might be more useful than you think. :-) (At least slightly.) Go back to that page and look for a link to the left that says "User contribution." Click on that and you will notice that ExtraKryspi has only made one edit in Wikibooks. The one edit that was plagiarized.
The Wikimedia communities do place value on author reputation but it's different than the traditional author reputation. It isn't linked to the "real" person as much as it is to the online username reputation. That being said, a mixture of online and real reputation may be the most respected. It has less to do with degree or occupation and more to do with consistent accurate contributions to the community (in some way). I often look at a user's contributions when I begin collaborating with him/her on a module or article. How I handled removal of ExtrKryspi's text was different since I knew he/she wasn't a regular contributor.
I will go so far as to say someone known as a scholar in academia wouldn't be a very respected Wikipedian (or Wikibookian) until they proved their understanding of the online Wiki community. At least when they have created a user name and made a few low key changes. Is this good or bad? Probably some of both but doesn't it often work similarly when people meet in the real world?
Nick Carbone on October 31, 2005 5:03 PM:
We're describing a morphing thing, sort of like trying to say what the view into a kaliedoscope is as its turned.
The Wikibooks process --a wide-open volunteer model where anyone can edit a page or make a change and contribute-- is good for some kinds of books and not others.
The wiki technology can be modified. Even if it's not, a Wiki project doesn't have to follow a Wiki Encyclopedia model.
So a critique of the process is not a critique necessarily of the technology.
Textbooks are --apologies for being redundant on this when it comes-- less about being books and more about being pedagogical tools. Most textbooks capture the work of scholar-teachers who have a good idea about how to teach something. They can an approach, a pedagogical view. They are tools in that teachers can, if they choose, use the books to help teach their students. Books help teachers do things.
Textbooks are collaborative. They are in fact reviewed and revised. I've seen books take four years to be developed into just a first edition. Editors send out drafts of proposals, drafts of chapters, sample TOC's on behalf of the author over and over. Authors and editors go through reviews and work together to make decisions about which paths to follow. Often there is conflicting, even diametrically opposed advice. Often a beloved idea is scrapped. Sometimes an newly contributed idea becomes beloved.
Wiki-based books *shouldn't* try to be like print based books. They're a different technology. I think a textbook in Wiki that tries to achieve the tone, scope, and voice of a traditional print developed textbook is a bad idea.
Wiki's are interactive in different ways than print. Students can engage the content and the teacher can assign things much differently. Reading and recall are not always the best learning strategies. So why have Wiki books try to be like print books?
When you compare them to print and how print books are produced, they will necessarily be different and often --not necessarily, but often-- inferior to print native books.
What could teachers do in and with a Wiki textbook? What could students do? What can a wiki textbook do to best bring that out? How does that change how students will engage and learn?
Those are the questions that matter, and the processes and actual shapes of books will arise from finding good answers. So too will any markets and opportunities for someone to get paid --whether in a good/better reputation or money or, if one is lucky, both.
kim white on November 1, 2005 12:47 PM:
Nick, I really appreciate your comments on this. For me, they help distill and focus the goals of this discussion. I hope you don't mind, but I've quoted you on next\text">http://www.futureofthebook.org/next/text/">next\text. I'm hoping a few of our readers will migrate over here to participate in the discussion.
Robert Horning on February 21, 2006 2:42 PM:
I just came across this Blog entry, and I'd like to point out that the Collaboration of the Month proposals are just that... a request to collaborate and make the content better. These are books that a majority of the Wikibooks community want to improve and make better, not make an example as some of the best stuff available.
We, on Wikibooks, have a serious problem at the moment in trying to display the best of what has been developed there. Most of the stuff on the front page is of low quality and is mainly a way to advertise for each independent book project in an attempt to recruit content developers.
There have been critiques of Wikibooks that have been done on a much better informed basis that include some of the much better Wikibooks. See http://www.lightandmatter.com/article/infrastructure.html for a not too sympathetic critic about Wikibooks, whose author surprisingly has actually started to contribute to Wikibooks after he wrote this review.
BTW, the 11,000 modules (now over 14,000) is not the number of textbooks, but the number of different pages of content in a collection of about 250 different textbooks at various stages of development. Still, I would have to agree that some better peer review does need to happen for the content that already exists on Wikibooks. Some of the content on Wikibooks is much better reviewed, especially the content about technology issues. Art History is one area that is in general an area of weakness for most Wikibooks contributors.
bowerbird on February 21, 2006 7:23 PM:
i guess i don't see why a wiki-textbook couldn't
start out as a single-author entity which is then
modified and improved and updated by others.
a good many assistant professors write textbooks
which are not picked up for publication, and thus
could be quite ripe for donation to the commons.
and ben cromwell over at lightandmatter.com
-- mentioned in the last comment -- is a good
example of someone who has written textbooks
and set them free. the .pdf versions of his books
are free; he makes money by selling print copies.
but many people won't want to go to that bother
of becoming a salesman, they'll be happy just to
turn their book loose and let others worry about it.
so i think there is great potential in a single-author
start that is followed up by community maintenance.
on this, also see: http://www.textbookrevolution.org
Takuya Murata on February 23, 2006 10:24 PM:
I think it depends. Maybe the art history is a hard task for a army of volunteers. But creating say books like language-learning books can be a very different story. In writing such a book you can really use a lot of inputs around the world, and I doubt any single person can accomplish such a job, in theory. And of course there are many books that don't have to be inspiring at all but have to have solid materials like math.
A price is certainly an issue, but it is quite possible wiki process proves to be effective in certain kinds of books if not art.