Are all textbooks created equal?

August 1, 2010 at 11:55 am 20 comments

An interesting piece in this morning’s NYTimes that I’ve been thinking about all morning (since Alan kindly sent me the link):

Mr. McNealy, the fiery co-founder and former chief executive of Sun Microsystems, shuns basic math textbooks as bloated monstrosities: their price keeps rising while the core information inside of them stays the same.

“Ten plus 10 has been 20 for a long time,” Mr. McNealy says.

Early this year, Oracle, the database software maker, acquired Sun for $7.4 billion, leaving Mr. McNealy without a job. He has since decided to aim his energy and some money at Curriki, an online hub for free textbooks and other course material that he spearheaded six years ago.

via Ping – In School Systems, Slow Progress for Open-Source Textbooks –

I’m really glad that school districts are finding ways to incorporate open-source textbooks.  That alternate path should exist, and it’s disappointing that it’s taking school districts so long to work that through their system.

I do have concerns and questions about open-source textbooks.  Some of them are obviously biased by my being a for-profit textbook author.  Others are questions that I have as a computing education researcher.

  • Quality process? Commercial publishers have a long series of checks over the quality of their textbooks, from the prospectus, to external reviewers, to copyeditors.  While there can be arguments about the effectiveness of the process (e.g., are professors-as-reviewers really our best prediction of the quality of the product, as measured in student learning?), there is a process.  I am sure that open-source textbooks can construct a similar process, but without the same teeth.  Commercial publishers hold a contract and royalty checks as the carrot at the end of the process.  If an author doesn’t like the result of an open-source textbook quality check, does he just release the book anyway on his own?  I worry about insuring quality for millions of schoolchildren when the track record of most open-source software projects is that they tend to care about the usability of a handful of expert users, rather than making sure that everyone can use it.  That won’t work with schools.
  • Does quality matter? “Ten plus ten has been 20 for a long time,” says McNealy.  Is that all there is?  Does the quality of the textbook matter at all for the student learning, or are all textbooks essentially the same — as long as the facts are presented and the exercise opportunities are there, the learning difference is insignificant?  As an author, I hope that the effort I put into exposition and interesting examples matters, but as a researcher, I know that often such fluff distracts from learning rather than enhances it.  I know that there are textbooks that stood out for me as a student and do now as a teacher, like Structure and Interpretation of Computer Programs and How to Design Programs.  Will such notable, standout textbooks arise from an open-source process?  (Why haven’t they so-far?)
  • Can innovation arise from an open-source process? Some of those textbooks that so stand-out for me are because they are innovative, a markedly new approach, regardless of quality.  I’m no expert, but my sense as a computing user and a computer science professor is that open-source software tends to copy others’ innovations rather than be innovative itself.  I got the chance to spend several hours with Andrew Tanenbaum once, the grandfather of Linux.  He told me in some detail about how Linux was derivative, lacking innovation, and lacking good design choices. I note that GIMP is often referred to as “open source Photoshop,” and I rarely hear about features of GIMP that designers wish Adobe would adopt into Photoshop. Will open source textbooks fare better?  Will the next innovations in textbooks (because we certainly need them!) arise from an open-source effort?  I am skeptical that the economics will work.  Innovations arise from people who know their stuff really well (see the ongoing discussion about creativity on an earlier blog post).  People who know their stuff well get paid for that.  Might they also volunteer time in their area of expertise, when they might get paid for that time?  Maybe.
  • Is the innovation in the approach or the textbook? While there is a lot of evidence that Media Computation improves on school’s “traditional” approach (for what that’s worth, since we don’t have a strong measure of what “traditional” means across schools), I can’t say that that’s because Barb and I wrote such great textbooks.  It could be the approach (e.g., the examples, the focus on a motivating context, the libraries and tools), which some schools are adopting (without the textbooks) by grabbing the free materials from  Do we need the innovation in the approach or in the textbook?  Will it work if someone pays for the innovative approach development, while the book might be free?  That’s the approach that Deepak Kumar and colleagues took with the IPRE textbook.  Microsoft paid for the development of the IPRE robotics approach, with NSF sustaining the effort, but the textbook is free.  Does this result in the quality (in terms of student learning and motivation) that we want?  Does it result in enough innovation to improve computing education?
  • Sustainable? Perhaps my biggest concern about the open-source textbook model is sustainability.  Who makes sure that material gets freshened up regularly and new editions come out?  Maybe McNealy is right, and once you show that 10+10=20, you’re done, and the examples don’t need to be updated, the language stays the same, and no mistakes are ever discovered.  Commercial enterprises offer an incentive to sustain effort, to keep making things better, and economics shows that people respond to incentives. Open-source textbooks offer an opportunity to serve and to have impact, which is certainly important.  Is it enough of an incentive to keep the effort going?  It would be great if it was enough, but I’m skeptical.

Entry filed under: Uncategorized. Tags: , , , , , .

AAAS evaluates NSF BPC Alliances Respond to the NRC today!

20 Comments Add your own

  • 1. Free textbooks « Gas station without pumps  |  August 1, 2010 at 1:48 pm

    […] his blog, Mark Guzdial argues that free textbooks lack the quality control and incentives for maintenance […]

  • 2. gasstationwithoutpumps  |  August 1, 2010 at 1:49 pm

    I’ve put my comments on free textbooks at

    • 3. Mark Guzdial  |  August 1, 2010 at 4:20 pm

      Point well-taken: I am confounding open-source and free. Part of that is intentional. It’s a huge amount of effort to build a book, and few individuals will put that much effort in to create a book for free. Nonetheless, I should be more careful to distinguish the two.

  • 4. James Howison  |  August 1, 2010 at 2:57 pm

    I think these questions are very well put. For me “the open source way” works when an artifact can be built in small relatively independent layers each building on top of what is there already, in a process of superposition.

    But the crucial thing is that such layers have to have their own independent payoffs hence their own independent motivations.

    You have me thinking about where the motivations come from in this domain. I’d think that fixing errors in the books will come easily enough, since users (teachers and students) are sufficiently motivated to point them out (think Lisa Simpson ;).

    Adjusting the whole pedagogical approach of a book, though, that is unlikely to happen, since that’s akin to the Version 2.0 trap (where effort gets siphoned off into Sisyphean re-factoring efforts with delayed payoffs).

    As far as innovation goes, I buy your examples. In fact a shared reference point is a very useful for a distributed project. I’d assert, however, that there is more innovative open source in areas where the software is directly used to solve its programmer’s problems.

    Funding to implement freely available texts has a quirk that it doesn’t motivate maintenance of that text, errata etc. But it does create the skeleton structure on which other small layers could be built.

    I discuss this in the open source context in a working paper version of the work I pointed out to you before (a much more readable length ;):

    Click to access CollaborationThroughSuperposition-WorkingPaper.pdf

  • 5. James Howison  |  August 1, 2010 at 3:03 pm

    One other thing occurs to me. It is relatively easy to argue that contributions could be driven by the academic reputation/contribution economy (ie authors rewarded by publication (the book) and evidence of its use, usually citations but here perhaps use/distribution/downloads).

    My new research is on scientific software production. Here there is an issue in that when a scientist writes a software package and releases it they often write a paper describing the software’s contribution, then request users to cite it. This works well enough for the original author, but subsequent bug fixers/improvers etc aren’t authors on the original paper and citations to that paper aren’t great evidence of their contribution.

    This quirk seems to suggest over-production of non-collaborative projects in software and, I’d venture, also in textbooks.


    • 6. Mark Guzdial  |  August 1, 2010 at 7:39 pm

      One difference, James, is that someone is paying for the scientist’s time when they’re writing the software package. That reduces the effort of having one job to pay the bills and volunteering on one’s off-hours. The economics are different there.

      • 7. James Howison  |  August 2, 2010 at 11:24 am

        Thanks Mark.

        I was thinking of contributions to open source textbooks as part of a full-time academic’s publishing practice. So in that case their university would be paying for their time, just as they do for Journal articles or teaching cases.

        So it seems to me that the analogy stands: scientists writing software for scientific reputational credit are doing it in their paid time too, not in their off hours (although with salaried employees and particularly academics that’s a tenuous distinction anyway 😉

        Of course some data on what type of time is being used to contribution to open source case studies would be great (ie what is being sacrificed for this work). Does anyone know of studies of these text’s authors’ time usage or even their motivations?

  • […] This post was mentioned on Twitter by Cathy Anderson, Mark Guzdial. Mark Guzdial said: Can free and open-source textbooks work? The same as for-profit books? […]

  • 9. Gary Litvin  |  August 1, 2010 at 8:03 pm

    There is a big difference between textbooks and software: textbooks are visible to the end user, software code is not. As a result, quality standards for textbooks are higher than for software code. Open source textbook authors usually like to do the fun part — explaining the material they like. They rarely want to do the dirty work: exercises and solutions to them, assessment materials, etc. Once a book matures to the level of commercial textbooks, it is often picked up by a ciommercial publisher and is no longer open source. Commercial textbooks compete by and large on their merits, at least mine do: I have no sales force. When people pay for my textbook I can be sure they chose it over other books based on its merits.

    It is hard to say what is more important: a good textbook or a good teacher. In my experience, better teachers choose better books for their classes. I believe the quality of a textbook is crucial. It defines what is presented, in what sequence, and at what level, setting its own standard. A good textbook also helps novice teachers master the material they teach.
    Gary Litvin
    Skylight Publishing

    • 10. Daniel Martin  |  August 1, 2010 at 8:54 pm

      What open source textbooks projects do you know of that seem to be noticeably lacking in exercises and solutions?

      I ask in total ignorance of what open source textbooks are out there, but it would seem to me that “contribute questions for this chapter” is the kind of thing open source could do well: it’s something that can be done by someone not the original author of the chapter, there’s not much concern over maintaining a consistent “voice” over the different questions as there would be with a content edit, and it can be done in small bursts of work by people contributing when they have a few minutes to spare.

      That kind of “let’s harness tiny bits and pieces of spare time from lots of different people” is where open source generally excels.

      • 11. Gary Litvin  |  August 2, 2010 at 9:17 am

        Theoretically, it is possible that someone will take a reasonable text and add exercises and tests to it. I’ve never seen it happen. If it does, who is going to select and edit the questions and assure their quality and consistency? If someone can recommend a complete high-quality open-source math or CS textbook, please let me know — I’ll approach the author with an offer to publish it with generous royalties…

        I might be old-fashioned, but I believe a textbook should reflect the vision, approch, and style of its authors. The idea of an open source textbook makes as much sense to me as the idea of an open source painting or an open source movie. I also strongly believe textbook authors should be fairly compensated for their work. I have been writing, editing, and publishing textbooks full time for 14 years; I wouldn’t be able to do that as a volunteer. My “contribution to society,” I believe, is that many thousands of students studied by our books, and that our cometition was forced to raise their standards and slash their prices.

    • 12. gasstationwithoutpumps  |  August 1, 2010 at 8:58 pm

      There is a company providing some of the dirty work without the textbook (Turingscraft), but not for free.

      A good text can replace a good teacher for a motivated, literate student. A good teacher can replace a good text (sometimes). The best situation, of course, is to have both.

      I’ve had a lot of trouble finding good texts for my somewhat idiosyncratic courses. I once wrote one chapter for book, and it took me so long that I decided that writing a whole textbook was not likely to happen.

  • 13. owen astrachan  |  August 2, 2010 at 7:30 pm

    Several of the posters to this thread make conclusory statements. It’s a great word — it means making an assertion with no evidence. It’s easy to do that. Gary Litvin writes “Open source authors want to do the fun part, not the dirty work of exercises, solutions, etc.” [paraphrased]. Which open source authors does Gary mean? He doesn’t reference *any*, it’s easy to take potshots when you don’t have to supply evidence. In a subsequent post he writes “I’ve never seen a [reasonable open source text] to which someone has added exercises” therefore none can exist. Again, positing that there aren’t any, can’t be any, haven’t been any is easy, all the burden falls to someone else.

    There’s also a huge conflating of open-source, free, software and texts. An open-source book doesn’t have to mean “developed by a group” — it could be developed by one person with a vision and then distributed so that it could be modified, e.g., by one of several Creative Commons and related licenses. Open-source as defined for software isn’t the same as defined for books, so conflating as Mark does “open source software is derivative” with “and so any *open* text book would be derivative too” is just completely sloppy reasoning.

    Wikipedia is a successful “open” encyclopedia. Some reports
    (e.g., show that its error-rate is comparable to traditionally-edited encyclopedias. Can open texts succeed similarly? We’ll find out, but because they haven’t doesn’t mean they won’t.

    Gary: PS: you do have a sales force, it’s the crowd of folks that use your book and vouch for it online — a far better recommender system than someone coming to my office, for example, and extolling the new book on their list that they have to extoll because it’s their job.

    • 14. Mark Guzdial  |  August 2, 2010 at 7:41 pm

      Owen, I didn’t say “and so any *open* text book would be derivative too.” What I explicitly said was, ” Will open source textbooks fare better? Will the next innovations in textbooks (because we certainly need them!) arise from an open-source effort? I am skeptical that the economics will work. Innovations arise from people who know their stuff really well. People who know their stuff well get paid for that. “ I’ll stand by that — I’m raising the question, and I’m skeptical. I did confound open-source and free, agreed. I’m skeptical because I don’t see how authors get paid through free books. I don’t see how innovative authors, who really know their stuff and are creative on top of that, will work for free. Of course, it might still happen — as you say, “because they haven’t doesn’t mean they won’t.” I’m not saying that it can’t happen. I’m simply skeptical. If you’re going to accuse me of “sloppy reasoning,” please get my mistakes right. I’m sure that my reasoning is sloppy often enough that it should be easy to quote those mistakes with my real words.

      • 15. owen astrachan  |  August 2, 2010 at 10:07 pm


        First, as you note you either/both confounded and conflated open course and free. Gary L does this below when he talks about putting his books “in the public domain”. There are huge differences between public domain and creative commons in terms of licensing. There are different open learning licenses including the initiatives via ccLearn which has licenses used/embraced or something similar both by the Curriki initiative and MIT’s open courseware.

        You specifically call out htdp (how to design programs) whose auithors make their book available online for free (as in no cost to see/use it online). Open source software projects are NOT OPEN for anyone to contribute to the main branch/trunk/release. You earn the right to contribute to the main project via a results-oriented reputation. Of course you can take an open source code base and do something else with it, but what you do DEPENDS ON THE LICENSE. Something released with a GPL software license has very different restrictions on what you can do with the code base than something released with another license, e.g., Apache, Mozilla, etc.

        In Wikipedia editors earn the right to do more than “any person” for some articles. Not every article written in Wikipedia necessarily becomes widely available — there are criteria.

        So, licensing and development are separate. I personally pledge that the next book I write will be available via a ccLearn license. Even if you can buy it, you’ll be able to get it without paying for it. I don’t have a book in mind, so arguably my promise is vacuous — but it’s a real pledge.

        • 16. Mark Guzdial  |  August 2, 2010 at 10:39 pm

          Owen, I appreciate the distinctions you’re making between licensing and development. I’ve never heard of a “ccLearn license” before, so I’ll have to learn about that. Not knowing about something feels different to me than “sloppy reasoning.”

          I’ll try to be more specific here:

          • I do believe that criteria for contributing to open source textbooks would lead to higher quality. The criteria for contributing to Wikipedia and large open source projects (e.g., Apache) are different (re: Andrea Forte’s excellent work on Wikipedians, contrasting with Jim Herbsleb’s analyses of Apache development), as are different than the criteria for contributing to small open source software projects, which tend to accept all contributions (from my experience). I don’t know about the processes of open source (in the sense of open, group effort) books. Let’s see if that does result in quality textbooks.
          • My issue about innovation is not about licensing. It’s explicitly about book development through on-line social processes. Your point is well-taken that that’s neither about free nor open-source necessarily. As I mentioned, I am skeptical that many single authors will produce free textbooks (HTDP is neither single-authored, nor solely free — there is a for-profit hard-copy available), but as you mentioned (and I agree), just because it hasn’t happened, doesn’t mean it won’t.
          • The issue of “free” is only important to my concerns in the blog post because it suggests the lack of an economic incentive to create and maintain the books. You’re right that “free” doesn’t have to be “open source” nor does it have to be socially-authored. I do find it unlikely that no “open source” book will ever be commercially viable with traditional publishers, so “open source” books will likely be “free.”
          • I’m using “open source” to refer to socially developed content, but you’re right that it’s technically an issue of licensing. My issues about quality, innovation, and sustainability are specifically with respect to book projects whose output is offered for free (for whatever license) developed in a social forum.

          I hope I’m being clearer and less confounded/conflated.

    • 17. Gary Litvin  |  August 2, 2010 at 8:31 pm

      Gary Litvin writes “Open source authors want to do the fun part, not the dirty work of exercises, solutions, etc.” Which open source authors does Gary mean?

      Owen, sorry, I won’t bite at this. As you know better than others, it takes a lot of work to write a book. It is not my place to disparage other people’s books in a public forum. I am an author, not a book reviewer.

      Gary: PS: you do have a sales force, it’s the crowd of folks that use your book and vouch for it online

      Precisely. They do, and I’m grateful for that. They are my “sales force,” not my my sales force. Our books are sold to the K12 market. Big educational publishers have a sales rep at every school district, besiege teachers with their offerings, put up huge booths at every trade show, etc. And, BTW, they come up with unnecessary new editions every other year. We rely, as you say, on word of mouth. As you can tell, I am proud of being able to outsell big institutional publishers (occasionally all of them combined).

      I didn’t say open source is not possible. What I am saying is I haven’t seen an open source book yet that would meet my standards. I am also saying that the notion of someone modifying one of our books without authors active participation doesn’t appeal to me. Others may not feel that way. (I welcome it, of course, when people develop supplemental materials for our books.) If someone wants to adapt one of our books for their needs and has good ideas, my co-author, Maria Litvin, and will be glad to work with them. I will put our books into the public domain when I decide to retire (or in my will); but for the time being, no open source for me.

  • 18. gasstationwithoutpumps  |  August 2, 2010 at 11:02 pm

    You said ‘I do find it unlikely that no “open source” book will ever be commercially viable with traditional publishers, so “open source” books will likely be “free.” ‘

    I already gave you a pointer in my post at to an article about non-free open-source books.
    The books are being sold by Macmillan through, so at least one major publisher disagrees with you about the commercial viability of open-source. (Personally, I think you may be right, and dynamicbooks is likely to last only 2 years before the startup funding runs out.)

  • 19. Flipping Textbooks | doug woods  |  August 6, 2010 at 8:18 am

    […] Are all textbooks created equal? ( […]

  • […] online learning tools, and even larger challenge, all the content and curricula with those tools?  I argued in a previous post that open source, multi-author, free books will likely not innovate nor ….  There’s some support for that argument in a recent New Yorker article by Malcolm […]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trackback this post  |  Subscribe to the comments via RSS Feed

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 10,185 other subscribers


Recent Posts

Blog Stats

  • 2,060,405 hits
August 2010

CS Teaching Tips

%d bloggers like this: