Two key conferences in digital media and game studies, Digital Arts and Culture (DAC) and the Digital Games Reserch Association (DiGRA) use an unexpected format for their papers: ACM, the format devised by the Association for Computing Machinery for publications in computer science.

I have nothing against computer science, but the use of ACM format is bad and dumb for venues like DAC and DIGRA. It may also be bad and dumb in general. Here are some reasons why.

ACM Format is Unnecessary for Rapid Publication

In computer science, the conference proceeding is considered a top-tier publishing venue, with the best conferences offering more prestige than journal publications. This state of affairs came about because progress moves quickly in science and engineering, and a publication can’t stand to wait years to find its way to readers. ACM format is one of several devised to allow the rapid preparation and publishing of papers by having their authors prepare “camera ready” versions of papers, formatted for direct insertion into a proceedings.

But timeliness is now very easy: papers can be published online, no matter their format. Indeed, ACM publications are put into the ACM Digital Library anyway, which is where most people find and read them.

In the cases of DAC and DiGRA, even though ACM format has been used, a print proceedings was never published for dissemination anyway (I believe some years of the DAC conferences are lost completely, with no publication having ever appeared in any format). Thus, the main purpose of ACM format is lost anyway.

ACM Format is too Discipline Specific

Standards like APA and Chicago are taught, used, and understood in many disciplines. Given the multidisciplinary nature of digital media and game studies, adopting a format from computer science is alienating. Ironically, few computer scientists do research in game studies anyway (a fact I lament).

Beyond putting a troll under the bridge, ACM’s indexing systems also prescribe specific fields and sub-fields (there is a huge hierarchical tree), most of which are inhospitable to humanistic, social scientific, and interdisciplinary work. The fact that conferences like DAC and DiGRA usually omit the ACM categories offers proof that they don’t work.

ACM Format is Illegibile

Because of the need to get papers quickly into print proceedings, and due to the cost of print publication, ACM format packs a ton of words onto one page, around 1,000, formatted in two columns. To accomplish this feat, ACM text is set in 9-point, making it completely illegible on screen and difficult to read in print too. Leading and other text formatting techniques used to increase legibility are likewise eschewed. The result are papers that are just plain hard to read.

The two-column format makes ACM papers hard to read even on screen. The small type requires zooming in, but the two columns require scrolling up and down the page. This isn’t the end of the world, but given the fact that many papers are read on monitors rather than dead trees, it seems silly to make it harder than it needs to be. (In the comments below, Mark Nelson reminds me that this also makes copying text out of ACM papers difficult due the the text selection behavior of PDF readers.)

ACM Format is Uninviting to Readers

ACM papers start with an in-line abstract and a set of arcane keywords instead of with the natural prose of an argument. This codes works as inaccessible. As a kind of media studies that hopes to clarify a form of popular media, in digital media and game studies it is foolish at best, irresponsible at worst, to alienate such a readership.

ACM Format Citations are Hard to Create

ACM format has a deeply unintuitive citation method that uses numerical references. One does not cite in order, as with foot- or endnotes, but one cites retroactively based on the position of a work in an alphabetized bibliography.

The format was designed for automated layout in LaTeX (computer scientists once insisted on programming their writing; some still do), in which case the process can be automated. It is less facile in traditional text editors or word processors. And even though bibliographical packages like Endnote can generate ACM formatted papers, an author should not be expected to have access to such tools to write a paper easily.

ACM Format is Length-Normative

In the ACM/CS world, papers usually come in two varieties, “short” and “long.” Short means four pages, long means ten (so, 4,000 and 10,000 words, respectively, minus any space taken up by figures and graphs). Length not only indicates the depth of an argument, but also the prestige of the publication: short papers are considered works in progress, tentative reports on new research, while long papers are considered mature reports of completed research.

While the journal article does imply a certian implicit length (usually 7,000-12,000 words), it is troubling to equate the length of a written work with its quality or importance. It’s particularly problematic to apply the short-long criterion to work that is not borne from laboratory-style research.

ACM Format Discourages Thorough Citation

The standard way to cite in ACM format is to off-handedly point to a source, which might be an entire book or article, without further clarification (e.g., “Earlier methods for automated gripe generation either fail under load [4, 14, 18] produce inelegant prose [1, 5, 17], or operate poorly on consumer-grade hardware [9].”) This convention arises partly due to the format’s provisions of length, but more so it underscores an ideology of the engineering world: research moves in a forward march toward progress and efficiency; research papers produce a single result that proves or disproves something.

As such, ACM citations rarely cite page numbers (there is a provision for it, but it is rarely used), nor do they clarify arguments with sufficient evidence. The lack of a tradition of “close reading” in computer science explains this conceit, as does the fact that most publications are considered simple reports on research findings, rather than arguments in themselves. It should be obvious why such an approach would be problematic in discussions of culture, history, and related matters.

I don’t wish to abolish ACM format, but I do wish that venues that publish work at the intersection of computing and the liberal arts would recognize that ACM format is unhelpful at best and dangerous at worst.

published August 5, 2009

Comments

  1. Mark J. Nelson

    I pretty much agree, but a few minor quibbles.

    I think the numbered-by-alphabetization referencing style actually predates LaTeX, and derives from physics journals that decided long reference lists would be less unwieldy if they were alphabetized. Of course, in that era one would submit a roughly formatted manuscript and the journal’s staff would deal with typesetting the references.

    The lack of page numbers I think is partly because many conference papers cite other conference papers, and it’s not usually necessary to point out a specific page number when referencing a 6-page article.

    I agree with the rest though. An additional problem with double-column format in general is that double-column PDFs seem to break most text-to-speech engines, ebook importers (for small-screen ebook readers), and even copy/pasting of text out of the PDFs.

    The short/long thing is a problem even in some technical areas. I’ve run across papers with a clever and reasonably important result, which could have made a punchy, to-the-point short paper. But if the result is sufficiently important, authors feel they should pad it out to long-paper length or else it’d be underselling their research to publish it in “only” a short paper.

  2. Ian Bogost

    Thanks for the clarification on LaTeX. I suppose I could point out that formats like Chicago and APA deal with this problem by providing an alphabetized bibliography separate from the references (whether they are noted parenthetically or with endnotes).

    Thanks too for pointing out the copy/paste problem with two-column PDFs. I meant to mention that. Ironically, this makes it harder to cite from ACM formatted papers, or at least less convenient, since a quote has to be retyped.

  3. Christian McCrea

    Well said. Its been used many times down here for various institutional reasons – and those are fine reasons of course – but it is also very content-prescriptive. Long-form arguments (say, for argument’s sake, for a piece on speculative realism and gaming nostalgia) are particularly ill-suited to a two-column, highly divided output.

    Though it must be said that the conference organisers in each case are fully aware of the limitations and have provided excellent flexibility, to their great credit.

    Though in the future, one hopes either DiGRA or DAC makes a break for it.

  4. Jill Walker Rettberg

    I agree with many of your points, and in general think DAC and DIGRA should use humanities formats, but I also really like some aspects of the ACM format.

    I think that the short vs long paper distinction is very useful – the point isn’t that short papers are inherently less important than long papers, but that you have a venue for presenting work-in-progress and new ideas that aren’t ready for the very thorough treatment of 7000-10000 words. I’ve thoroughly enjoyed both writing in and reading papers in this short format, which doesn’t really exist in the humanities disciplines – there’s usually no way of publishing a rigorous 1500 word paper as humanities disciplines want the 5000 word or 8000 or 10000 word standard, which obviously requires a lot more time and a much more completed project. There’s great value in getting ideas out fast even in the humanities.

    I also really like the abstracts. Many humanities journals do require abstracts, and I find them very useful in quickly figuring out whether or not I want to read a paper. I really don’t see how abstracts discourage readers. Keywords, well, they’re kind of haphazard, but can make finding articles easier.

    I agree about the page numbers in citations, and although I’m happy enough with the abbreviated references SO many people new to the format completely mess it up. Double columns suck on screens. Camera ready copy is quite useful though and certainly helps when print proceedings are made, which I think has been the case at about half of DAC conferences?

  5. Noah Wardrip-Fruin

    I think people may in part specify ACM format because they hope to print a proceedings and want to limit the cost (fewer pages). Hasn’t at least one DiGRA had printed proceedings? I’m pretty sure at least one DAC has.

    Also, the ACM has done work to provide templates and instructions that support people in different writing environments. Perhaps there are other groups, in other disciplines, that have done similar work? If so, we could urge digital media conferences to consider building on it…

  6. Ian Bogost

    @Jill

    I agree with you about the value of shorter lengths–what I’m bothered by is the idea that a short paper is inherently less valuable than a long one, and that 4,000 words is the correct length for such a paper.

    I also wasn’t knocking abstracts as such; just the way they are presented in ACM format and how that presentation reduces legibility. Compare the way a journal typesets an abstract to the way ACM does.

    @Christian

    The conferences have indeed provide good instruction, I agree. However, that doesn’t fix the underlying issues. Still, it’s not my intention to criticize the conference organizers.

    @Noah

    You’re right that DiGRA has had one printed proceedings, at least, but it was never disseminated beyond the attendees. I don’t believe that DAC has, but I could be wrong.

    As for templates and instructions, ACM provides Word files and the like, but that doesn’t reduce the punitive nature of citation. As for a different

  7. Lucas

    I agree with most of what you say, though I don’t think the 9 pt. typeface is all that bad.

    To me, the two-column article format and classification system are particularly awkward. It’s a shame that academic writing (at least, in computer science) still seems to be made to be printed and filed.

    In my experience, I’ve read most articles in PDF format, which is a big, slow, and inflexible format for reading on a screen and using on the web.

    Publications would be far more useful if they were designed with the reader and screen in mind, with the aim of being quoted and linked.

  8. Jose Zagal

    All the DiGRA conferences have had printed proceedings. In the case of DiGRA 2007, the proceedings were then re-published as an edited book (“Worlds in Play”) available for purchase beyond conference attendees. The proceedings are also available via DiGRA’s digital library.

    Rather than adopt a format from either the sciences or the humanities, we should come up with our own that uses the better aspects of each. Also, I would distinguish between citation format and style of articles…

  9. Ian Bogost

    @Jose

    Correct me if I’m wrong, but the DiGRA printed proceedings have only ever been distributed to attendees, not published to the larger community. A clarification: the Worlds in Play collection was a selection of the 2007 papers, not the entirety of them.

    I agree with you that citation style and article layout are different matters. Also, note that styles like Chicago and APA style are not really of one field; they’re pretty broadly used.

  10. Jose Zagal

    @Ian

    I believe the printed proceedings for the first DiGRA conference were also available for purchase after the conference. I remember thinking about buying them at some point. Thanks for clarifying about the 2007 proceedings.

    About Chicago and APA, agreed. Perhaps we should simply adopt those for citation style. I do think, however, that it’s more important to come up with a decent style (and template) for article layouts.

    Side note, another annoying thing about the 2-column layout is that it makes it more complicated to include images and tables (especially when they’re wider than the column of text).

    I wonder what it would take to get the ACM to change their layout style? I know that different ACM conferences have their own styles (CHI, SIGGRAPH, etc)…

  11. Ian Bogost

    Thanks for clarifying the availability of the DiGRA 2003 proceedings. I checked online and found them on the DiGRA site. It seems the other proceedings are available there too, which means I was wrong that the proceedings were never published. That said, the only way one could find them is by going to the DiGRA site or searching for a paper title, which is perhaps less than ideal, perhaps. In any case, I’ve updated the post above to clarify things.

    I’m intrigued at your original and follow-up question about an alternative template for articles. I suppose it would be important to think about where these articles would end up, whether in print or online, as a part of evaluating possible solutions.

  12. Christian McCrea

    One possibility that I’ve heard bandied about, and one that I have sympathy with, is the utilisation of the growth of blogs and academic vanity sites (I mean vanity in the historical, not pejorative sense, Ian!) to deliver articles already available elsewhere in these arcane databases, typeset in any manner the author sees fit. I know of at least four people who do this via academia.edu.

    The rationale for ignoring the copyright concerns of databases such as ACM is very strong, but I suspect in this publishing environment that it may be excellent scholarly practice for all involved to multiply rather than divide the locations of research. As an example, for all the noise that McKenzie Wark’s Gamer Theory generated, its easy to see the value of testing the foundations of work online and then when durability is required, the durable form is called on.

    This principle extends to multiple versions of the article online; one for the database, another with whatever additional illustrations for the world. It stands to reason that both the database and the vanity site or blog benefit; one purpose multiples the other, as scholars find the unofficial version first but then draw down and cite the other – which is precisely what I’ve done with those using academia.edu.

  13. Ian Bogost

    Christian, that does indeed make sense. The idea being, the article gets “unpublished” into the archive, and the “real” version exists on another site. That works as long as the other site has enough Google juice to become the definitive version of the article, or at least *a* definitive version. Still, it seems more like an interim more than a final answer, to me. Archival is important, after all.

  14. Erik

    Bad text selection is Adobe’s fault rather than ACM’s but there is a single column ACM format that some ACM journals use i.e. http://www.acm.org/publications/submissions

    why not for DiGRA? Changing _completely_ from ACM to Chicago or APA won’t encourage computer scientists. They have got used to the horror and the quiet chuckle from telling non-academics they deliver their papers in LaTeX.

    And Chicago has two main versions and many mutations that I find irksome.

    Secondly, I thought DiGRA 2009 was CD proceedings only? “All the DiGRA conferences have had printed proceedings.”-Jose.

    Instead of short versus long papers (you can have a very short long paper) why not papers versus research-in-progress papers?

    Finally, it is great to see keynotes online, I personally think conferences should ask for it from all keynotes. Or video them (DiGRA TED). Especially if keynotes are all on the last or second-last day! (mutter mutter).

  15. Ian Bogost

    Hey Erik, thanks for weighing in. Chicago or APA are really just citation styles, not formatting styles, and I think they are even supported in LaTeX if you want. Anyway, as a few people have correctly noted, a formatting style and citation style could be separately considered.

    DiGRA 2009 was CD proceedings, and not all of them are full papers. I think they will eventually go online in full form at the digra.org site.

    I agree that the length of a paper should be decoupled from the sort of function it serves.

    As for keynotes, do you mean live feeds, or the fact that some keynotes never materialize in written for, or something else?

  16. Elizabeth Losh

    To add to your list of irritations about ACM: not supported by Zotero.

    And, as one of the conference organizers, I can say that you are absolutely right about DAC 2009, but I have print proceedings from DAC 2005, which were published by ITU Copenhagen.

    Now back to dissatisfied fiddling with one of my own an ACM-formatted papers . . .

  17. Erik

    “Hey Erik, thanks for weighing in. Chicago or APA are really just citation styles, not formatting styles”-my initial answer disappeared but I have to say I think due to their publication origins they are both (for example, the Chicago Manual of Style and the APA Formatting and Style Guide) and even if used purely for citation, their usage could confuse some. I know I have been asked to write a document in Chicago style/format without being told if that meant 15a or 15b..and I also believe referencing format can affect the style of the paper.

    Keynotes: Since DiGRA (so far) has been every 2 years, and since the DiGRA keynotes I have seen varied wildly in style theme and content, I think some form of video capture would be an asset to the organization.

  18. Alex

    Thanks for the article Ian; I agree with many of your criticisms.

    One question though: what is the provision to cite page numbers in ACM that you mention? I would like to use it, but can’t find any information on the format elsewhere.

  19. Ian Bogost

    Alex, basically: [13, p. 45]. Or something like that.