The Evils of OSO

I have come to the conviction over the summer that the field of Philosophy, and academic publishing more broadly, faces a dire threat — one that, so far as I can find, hardly anyone has taken notice of. I refer to the evils of Oxford Scholarship Online.

In case you have had the good fortune not to have had any dealings with OSO, let me explain that this is OUP’s idea of how to make its books available in digital format to institutions and individuals who are willing to subscribe to such a service. Last year, my library (at CU/Boulder) made the decision to sign up, and this seemed like quite a good thing at the time, inasmuch as it means we get electronic access to all OUP books (at any rate, all of them published after the starting date of our subscription). Right away, however, we received the grim news that, with our OSO subscription in place, our library would no longer be purchasing OUP books (except by special request), meaning that our only access to OUP materials would be digitally, through OSO.

Still, we thought, perhaps this is on balance a good thing. It has become very clear, however, that the decision was disastrous.

Now I am no Luddite – indeed, for a great many purposes, I prefer to read material on screen, and I have accumulated the usual collection of programs and devices to facilitate that sort of thing. So my objection is not that OSO marks a prominent step on the path toward the end of books in academia. The problem is that what OSO offers, in place of OUP books, is, to be blunt, execrable.

The heart of the problem, as anyone who has used OSO will know, is that Oxford is unwilling to make images of their published books available. So instead of the usual .pdf file that we are all familiar with receiving in place of printed journal articles, OSO gives the reader something like an .html version of the book, one that looks nothing like the book itself, even if in principle it is a word-for-word duplicate. Their fear, presumably, is that if an exact digital version of the book were made available, it would soon become available everywhere for free. The worry is a reasonable one, but unfortunately their solution is to make their product so wretched that no one could possibly have any interest in circulating these OSO editions.

To make things more vivid, let me concentrate on the last two works that I read using OSO, a recent paper in epistemology by Jane Friedman and a recent book on Hume by Frederick Schmitt. I might instead have talked about anything I have ever read on OSO, since the problems that I will describe are endemic to the system. But these two will do.

Here is a screenshot of what the printed OUP version of Friedman’s article looks like (ignore the highlighting):

friedman oup

Now here is a screenshot of the OSO version:

friedman oso

The first thing to remark on is just how horribly ugly the OSO version is. There was a time when Oxford University Press cared deeply about the art of printing, maintained high standards of typesetting, and took pride in decisions about fonts and other aesthetic niceties. The printed version of Friedman’s article reflects those traditions in ways we have come to take for granted from serious presses. How is it, then, that one of the world’s great publishing houses can now care so little about all that as to give us something as wretched as what we see here from OSO? Are they, for instance, incapable of right-justifying the margins, or do they just not care? Did they deliberately choose the ugliest possible mixture of fonts? Why all that extraneous numerical and legal information? Why on earth are there distracting OUP “watermarks” stamped all over the page?

If this were just a handy, searchable supplement to the book, to be consulted only as a temporary stand-in for the real thing, then these complaints might not matter. But in my part of the world this is the main form in which OUP publications are available. And I am not alone. In the past, when I could not get an OUP book from my own library, I could almost always get it from a consortium of libraries in the Colorado area. But with respect to both of the books I have mentioned, a search of the consortium reveals that not a single library in the Colorado area owns a copy of these books. The only versions available are the electronic versions.

If only this were the end of the story. Matters get even worse when we look at more substantive matters. Notice, for one thing, how the book gives us footnotes, whereas in the OSO version these become endnotes. Apparently, footnotes are another thing that the technical whizzes at OSO are incapable of. But the difference matters. Careful authors find out whether their work is to published with footnotes or endnotes, and treat the two differently. (Moreover, OSO does not even bother to supply hyperlinks making it possible to flip back and forth from text to note, meaning that one must laboriously scroll forward and hunt down each note, then scroll back. Then repeat.)

Notice, too, how it is impossible to tell where Friedman’s article begins, in the OSO version. For reasons quite unclear to me, it is an obsession at OSO to collect abstracts and keywords not just of whole books, but even of individual chapters within books. Now you might suppose that, if an author thinks such information belongs at the start of her work, then she will supply that information herself. Not at OSO. Here every chapter to every book has stuck onto it at the start an abstract – maybe written by the author herself, or maybe not, maybe cranked out in two minutes before heading off to teach a class, based on vague memories of what the chapter says – and formatted in such a way that the innocent reader has no way of knowing that this is not in fact a part of the author’s text.

At least here we are given the author’s name. I have seen books where that has been omitted, and for some reason these OSO versions never seem to provide the number of the chapter, something that obviously matters when you are trying to read the chapters of a book in the correct sequence. It was in fact not easy to find Friedman’s chapter in the book, because for some reason the front page of the OSO version provides only the title of each chapter, not the authors’ names. Moreover, even stranger, the OSO version omits the table of contents where the authors’ names are listed in the book, so that a reader who wants to know who has written the thirteen chapters in this book has to hunt that information down chapter by chapter.

Friedman’s article contains a fair number of technical notations, and I was impressed with how much success OSO had at converting these into its proprietary ugly font. Or, at any rate, into some font or another. One thing that one sees over and over is that, although the OSO version is converted straight-over from the final printed version, the conversion process is often shaky at best. On page 65, for instance, Friedman gives us the formula 0 < x y < 1, which in the printed version looks as it should. But for some reason in the OSO version that comes out as this:


Similar infelicities abound, thoughout. Usually, the reader can make sense of what is going on, but is this really the best Oxford can do?

Alas, Friedman’s case looks like a glowing success next to Schmitt’s new book on Hume. As before, superfluous material has been added in front of each chapter. But in this case Schmitt had his own ideas about the sorts of overviews he wanted to prefix at various places, and so the book is very carefully split into three Divisions, with a brief “preview of the Divisions” and then brief “previews” of each individual division. These are to be found in the OSO version, but the reader is unlikely in fact to find them, because the first is printed as if it were part of Chapter 1, and the others are hidden away as separate files that the OSO-user is very unlikely to discover. Only someone in possession of the book itself will understand how Schmitt himself wanted to introduce its various chapters.

There’s worse. As before, footnotes have become endnotes, but the really amazing thing here is that the notes have been mangled in a way that makes them virtually unusable. In the very first chapter, the reader is presented with two endnotes assigned the number 1, and two more endnotes assigned the number 2. What’s going on? A reader fortunate enough to be able to consult the book itself (or someone less fortunate, like me, who teases this information out of the Google online preview) will learn that in fact these doubled notes have been somehow imported from other chapters. The first doubled note comes from Chapter 2. The doubled second note comes all the way from Chapter 7. This goes on and on – there are actually 3 versions of notes 22, 27, 31, 33, 34, and then at the end there’s a note 85, following directly after note 65, that somehow got imported from Chapter 2. Here’s just a taste of what it looks like:


Subsequent chapters are just as bad, and indeed for good measure there are 87(!) pages of random notes stuck onto the end of the book, after the index – don’t ask me where these came from!

How can things be so bad? How can the world’s leading academic press be churning out material of such shameful quality? The essence of the problem is that – as every author knows – it takes a great deal of hard work to create elegant, carefully produced volumes. I myself have published several volumes with OUP, and have always been pleased with how the books have come out, but it has happened only at the end of a very long process involving a great deal of hard work by many hands. With OSO, almost none of that happens. In particular, at least in my experience, authors are never asked to look at the OSO version of their books, and judging from Schmitt’s book, no one else looks closely at them either. (In my case, not only was I never asked to look at the OSO versions, but in fact I have no access to them, so I am blissfully ignorant about how bad they might be!)

The solution, of course, cannot be for authors to go through two production processes, one for the book and one for the OSO version. As published authors know, one such process is agony enough. So the ultimate problem is that OUP is trying to do something that just will not work – it is trying to produce two versions of a book at the cost and effort it takes to produce one. The results are predictably appalling. But we should keep in mind what I said at the outset, that this is, for OUP, ultimately the whole point. The OSO version of a book is supposed to be something that is so markedly inferior to the printed version that it will not pose any commercial threat to the sale of printed books. This is what scholars and librarians need to understand. And if we care at all about the production values of the books we love, we need to stop letting Oxford get away with this.

This entry was posted in Rants.

28 comments on “The Evils of OSO

  1. David E Luscombe says:

    Dear Bob,

    Powerful critique. Extremely worrying David Luscombe

  2. Sander de Boer says:

    Dear Bob,

    Well put! This this been a long-standing frustration of mine too. I would add that OPO curiously even failed to properly implement the one big benefit of E-books over printed volumes: being searchable. Using the ‘search in book’ option return a list of the chapters (sic!) in which the term appears, rather than a list of pages, making it almost useless.

  3. […] It sounds very frustrating, but Pasnau writes in such an amusing way as he substantiates his charges that it is quite fun to read about. His remarks about the lack of care with typography are spot on, and the part about the endnotes will make you cringe. You can read the whole post at his blog, In Medias Phil. […]

  4. Alan Nelson says:

    Thanks for this, Bob. I assign these to students and am happy to save them money, but I’ve noticed how bad they are when I try to project them on the screen in class.

  5. Anonymous says:

    It’s also disturbing that your library doesn’t own these books. You have access to them (and, presumably, to a quite large number of other books) only as long as the library continues to pay for the subscription. When this eventually becomes a significant portion of your library, OUP may well feel they are in a position to raise the price of the subscription to extortionate levels. (I’m not suggesting they would, but there are other publishers that would probably not hesitate to do so.)

  6. gentle reader says:

    I just print out the chapters I need. It shifts the cost of printing to the consumer but I get to mark up my pages and I rarely need more than one or two chapters of any given OUP book

  7. Also dismayed says:

    I completely agree. Here’s another, related problem. Oxford also does Oxford Handbooks online, which I think is a separate service, and their text — both html and pdf — doesn’t include even the page numbers of the printed versions. So there’s no way to cite anything you read in an Oxford Handbooks article you got online unless you can get your hands on the original printed version.

  8. Anthony Fisher says:

    I concur. OSO need a reader like and their books are free.

  9. Thank you for detailing all of this, which sounds dreadful. The horrible term ‘crippleware’ for hardware or software deliberately made to lack features (normally charged for separately) seems relevant, although in this case OUP seem to have been partly motivated (by fear of piracy) to make a product with such poor functionality. Perhaps, though, they are planning a future ‘premium’ service where the books don’t look awful, and are easily searched, etc. There has to be a better way.

  10. Abe says:

    Very well said. This is a dreadful state of affairs. Let’s hope that the powers that be at OSO take note of each of your points and make it a matter of priority to dramatically improve things. In the meantime, academics should forward this to their university librarians so that they avoid signing up to this scheme and continue to order OUP books.

  11. pascal engel says:

    I have, with my colleagues at Geneva, used OSO for several years now. It’s true it’s not easy to read, very cumbersome, both to use and to download, but :

    1) you have the book chapter by chapter, which is often useful , since very often one works on a single or two chapters of a book , not the full book

    2) unlike what happens with e-books ( which I find distratrous and impossible to work with, since you cannot copy and paste and navigation within them is very hard) I can copy and paste some passages on OSO for reference or quoting small passages, which I can’t with e-books or other media

    So overall, in spite of the drawbacks you point out, I find it much better than other online formats

  12. In response to Pascal Engel:
    There are programs available online which you can use to remove the DRM (an electronic ‘lock’) on any e-book. This opens many features, including easy copying and searching.
    Hope that helps!

  13. Peter Momtchiloff, OUP says:

    Hello Bob

    Thanks for your comments. We shall certainly learn from these. Here are some thoughts in response.

    First, Oxford Scholarship Online is designed to be read directly online. It appears that the version which you are focusing on is the pdf that users can generate from OSO and save. OSO is not primarily designed as a pdf-generating device. Nevertheless, given that we are offering users the facility to do so, we should aim to avoid problems with this secondary format, and we are working to improve the quality of the pdf.

    We are confident that the appearance of the primary OSO format is good, and it offers much better functionality (e.g. navigating, searching, linking) than a pdf (or indeed than a print book). When people read OSO online, I think they will find no problem with contents lists, abstracts, structure, connections. Everything is very clearly laid out, and a click will bring you the move you want to make, the information that you are seeking.


    Are there technical problems in OSO? As with most technology, problems arise, and we address them. The site is constantly evolving in response to user feedback and as technological changes/enhancements are available. You have identified some problems which certainly should be addressed. We have corrected the incorrect characters you noted. We are currently testing a fix for the problem of extraneous notes, and an enhancement to clarify the display of chapter authors: we plan to roll them out to live next week. We are changing the way we generate our pdfs, which will now allow us to improve this feature also.

    Is OSO evil? No. It’s an excellent resource. It significantly enhances the potential for dissemination of academic work. One of the reasons librarians are under such pressure to cut their print spending is that books in academic libraries don’t get used as much as we would hope. OSO offers new, swifter, and more sophisticated ways for readers to find book content that will be valuable to them.

    Is OSO deliberately flawed in order to undermine the move from print to online? No: the need for effective online publication of academic research is inescapable, and is something which OUP has embraced, having been a leading player in online publishing innovation for many years. We will work with authors, readers, and librarians to maximize its success and popularity.

    At the same time, print sales are proving pleasingly resilient and we expect print and online formats to continue flourishing side by side. Not everyone will have access to both: and librarians will have difficult decisions to make. But we are offering them good options.

    Best wishes

    • Stuart Ingham says:

      Hi Peter,

      The following two of your comments contradict each other:

      “Oxford Scholarship Online is designed to be read directly online.”

      “Is OSO deliberately flawed in order to undermine the move from print to online? No: the need for effective online publication of academic research is inescapable, and is something which OUP has embraced, having been a leading player in online publishing innovation for many years. We will work with authors, readers, and librarians to maximize its success and popularity.”

      The ability to print a document, take it around with you and annotate it is essential to academics. Digital content which discourages offline reading–by making bad or no pdf provision–is simply crippleware. OUP Philosophy continue to to publish wonderful books, its online provision is not to the same standard.

      • Peter Ohlin says:

        OSO does allow you to output a PDF, which you can save, print out, or annotate, in any way you wish.

  14. tbb says:

    Just yesterday I accessed the chapter “Perception and Attention” from the Oxford Handbook of Cognitive Psychology. I generated a PDF and printed it out. It was really easy and the printout–four book pages to a sheet, to save paper–looks great and is easy to read. It has that kooky “abstract” at the top, which begins “The relation between perception and attention is discussed” (gee, thanks), but other than everything is nice.

    Maybe it’s just that my uni gives free printing to grad students, or maybe this is just a thing about the age and tech experience of the user, but I find it far easier to deal with the Oxford books (especially the handbooks) in this way via the online system. In addition to what I did, I could have highlighted the text from the HTML version and pasted that into a text editor, then put the resulting document on my Kindle to read. Or I could put the PDF on my Kindle and read that. I couldn’t do any of those things with a hard copy of a book. Though the issues with the footnotes are real, I doubt they will take long to correct.

    I’m usually all for raging against the oxfordian machine, but in my opinion, the books online–even in their present, slightly defective form–are far easier to work with.

  15. […] “I have come to the conviction over the summer that the field of Philosophy, and academic publishing more broadly, faces a dire threat — one that, so far as I can find, hardly anyone has taken notice of. I refer to the evils of Oxford Scholarship Online …” (more) […]

  16. Girard Desargues says:

    Mr Momtchiloff, we understand that you have to do PR for your employers. However, perhaps you care to answer two points raised in Professor Pasnau’s original post?

    The first was about OUP’s terrible choice of fonts. OUP in the past was keenly aware of how important that choice is, for large masses of text. Based on three centuries of typography, they had wisely chosen faces like Stempel Garamond and Membo. Both are eminently readable and easy on the eyes — the key factors in reading a book. Inexplicably, OSO publications are now in a sans serif font, wholly unfit for large chunks of text. As I’m sure you know, sans serif faces are meant for titles only, not for entire paragraphs.

    The second is the abominable layout. Why is the text left-aligned, and why are the margins so narrow? I’m sure you too are aware that it makes everything a lot less legible, not just ugly to look at.

    These two elements, combined, make your new publications look cheap, dismal, and a headache to grapple with. Do they really save you a lot of money? And, if they don’t, why not try to do better? Right now, OSO titles resemble the stuff that Axel Springer’s outfits put out. (Enough said.) You may want to tell your employers that they’ve gone down the path of General Motors in the 1990s — making everything shoddier and uglier for the sake of dubious savings; but in the end making cars that nobody wants to buy.

  17. Michael Kremer says:

    The problem with multiple foot/endnotes occurs in many other OSO books. For example I have just run across the same problem in Richard Swinburne’s Faith and Reason.

    The OSO text may be “designed to be read directly online” but it is not a pleasant experience. And the foot/endnote problem occurs whether in the PDF version or the online version. While one can click on a note number to get to the note, when one clicks on note number 4 in chapter 4 of Swinburne one is confronted with five separate notes numbered 4 and has to figure out which of them belongs in chapter 4. And this problem is pervasive.

  18. european says:

    Peter Momtchiloff: why doesn’t OSO provide exact pdf replicas of the print editions of the books? Why spend money and worktime on typography and layout and then prevent customers from seeing the result of that effort? Exact replicas should be a simple process since you already digital proofs for printing.

    Has fear of piracy been a big factor in the design choices OSO has made? It seems so to me. For example it appears OSO have made it unneccesarily cumbersome to download pdf chapters by adding extra warnings and buttons to go through for each download. And then the formatting bugs, typography flaws and distracting watermarks. I doubt any of that curbs piracy much as those who really want to pirate the content will find a way to do so anyway. In other words I doubt there is be a big *difference* in economic losses due to piracy from (a) a system with exact book replica pdf files and (b) the system with substandard pdf versions. But there is a big difference in quality of service for the core userbase and thereby for the universities that pay for the service.

    • anonymous says:

      this was my immediate thought as well. it also seems like a good time to remind everyone that Oxford volunteered to be part of the Copyright Clearance Center lawsuit against Georgia State University, which they decided to pursue on appeal after a decision that wasn’t very favorable to the publishers.

      It’s clear to me that the motivations here are about frustrating faculty to encourage them to pressure libraries to purchase print as well (which, frankly, libraries just can’t afford) and also thwart assumed piracy issues.

  19. Michel X. says:

    Peter Momtchiloff,

    Viewing the text online does nothing to resolve the weird abstracts of unknown origin or the left justification, nor does it resolve the serious problem with page numbers which Bob neglected to mention (viz., they’re incredibly difficult to find because they’re in the body of the text, and one can’t tell at a glance where pages begin and end).

    I can’t speak to the other problems Bob noted, since my library’s access to those texts is not via OSO (phew!). From what I can see (I have a PDF and online chapter of another text open at the moment), however, the main difference is that online the “printed from OSO” note doesn’t appear. I guess that’s a plus, although being able to view a text online isn’t all that useful if one can’t annotate it (as one can do with a PDF, where that infernal boilerplate is reproduced and further confuses the pagination).

    One last point bears mention: for a single-author work OSO seems to count the table of contents, index, bibliography, endnotes, images, etc. as separate “sections”. So unless one is careful to consult these things via some other means first (e.g. a physical book, amazon/google previews, etc.), one can easily be prevented from downloading the material one set out to acquire because one has already hit one’s copyright limit (by accessing these extra materials first). The frustration is a small one, but added to all of the other little details, the interest is compounded.

    All in all, I regret to report that my own experiences with OSO have been infuriating. If piracy is the driving concern (which I know you deny, but still…), then I would prefer to have proper access to a PDF of the book itself with a built-in “return” date (i.e. a date past which it’s unusable and must be re-acquired through the library). Our library already does this for audio items, and it works well enough.

  20. Peter Momtchiloff, OUP says:

    Thank you all for your comments. You have all made points well worth noting. We always appreciate hearing from our readers, even when their views are critical, and we always bear their feedback in mind as we work to improve OSO.

    The concern that has been voiced most commonly here is about the readability of the pdf version of OSO content, and we have heard this clearly. We recognize that in some cases academics will want a pdf version of the content: as I mentioned, we’ve been working, and shall continue working, to improve the quality of that version. On the other major concern, that of the extraneous footnotes: we have a fix for this bug, and plan to begin rolling it out to affected titles next week.

    On some of the other specific comments:

    Yes, piracy of content is a very serious concern.

    The sans-serif font currently used on the generated pdfs offers maximum coverage of the uncommon characters that can occur in OSO content. On the site, all the content is presented in a serif font. We have put a lot of thought into the design of OSO online, with close attention to the different ways in which it will be used. I am confident that the design is good. Of course we shall continue to monitor reactions, and the design may evolve over time. Opinions are always going to differ strongly over typographical aesthetics.

    How we present the page number needs to be a balance between being clear and not obstructing the reading experience. It’s a compromise. The page-numbers on my screen stand out quite clearly. It is also possible to navigate directly to a print page number.

    I think there is nothing that prevents Michel X consulting the whole book online on OSO: the restrictions are only about saving content.

    We know that the reading experience is a central part of academics’ lives. It is essential that we work on new ways of presenting academic content for new ways of reading, alongside the traditional ways, but we shall continue to try to be sensitive to and responsive to readers’ reactions.

  21. A student says:

    No objection that these are live issues — but they are small compared to the huge benefits that OSO offers: free access to books, the opportunity to print and read on screen (no more heavy books to carry around), easy searching, etc.

  22. Girard Desargues says:

    Thank you for taking the time to address our concerns, Mr Momtchiloff. Unfortunately, I’m not sure that you gave real answers to our concerns.

    You say, “The sans-serif font currently used on the generated pdfs offers maximum coverage of the uncommon characters.” I’m afraid I doubt it. There’s a *very* readable, elegant serif font OUP used for their *printed* editions of the Oxford Handbooks. It’s Minion Pro. A much, much better choice than the terrible sans serif you in the PDFs of chapters from those handbooks. Why wouldn’t you make the PDFs looks just like their print equivalent? Or — gasp! – better? Here’s the most depressing things about the “new” OSO: it reeks of incompetence and cheapness. You hurried to roll out a shoddy, buggy product without enough thought and quality control put into it. OUP has been a world leader for many decades. It had a new opportunity to lead, as we enter the age of digital humanities. Instead, it lowered the bar to developing-world lows.

    You say, ” I am confident that the design is good.” No offense, but you’re an editor, not a typographer. Let’s hear it from the relevant experts.

    You say, “Opinions are always going to differ strongly over typographical aesthetics.” That’s a red herring. Opinions will differ, but what matters are the (admittedly varying) opinions of the best experts (which OUP used to rely on), not of everyone. Surely you didn’t just approach any Joe Schmoe on the street for advice on your (formerly outstanding) print books. You consulted the best experts — and it showed: re-read Professor Pasnau’s post above. Can’t you do that again?

    With OSO, OUP abdicated its tradition of excellence and leadership. You’ll see that, in the long run, it’ll leave a mark on your outfit, despite your chipper talk about resilient sales and stuff. Thanks to its rich universities, North America remains you biggest market, but Americans can smell a cheapskate from a country mile, and will run away from it. (Remember Microsoft? Where is it now, after Windows Vista and 8? They’re laying off people, that’s where.) I have two books I project to have ready over the next five years, and I won’t approach OUP with either. Not when there’s CUP, Stanford, Harvard and Princeton UP. I had agreed to write three contributions for your Oxford Handbooks, lured by the high quality of the print edition of those titles. Now that I’ve seen what you did with them online, I deeply regret my decision.

  23. […] always possible problems. Consider a recent examination of Oxford Scholarship Online (OSO) in this blog post by a philosopher at UC-Boulder that was discussed by Inside Higher Ed. He notes a number of issues […]

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