February 4, 2008

EP 3.1: Digital Fictions and the Eliza Effect

by Noah Wardrip-Fruin · , 7:19 am

Imagine you’ve checked out your books and walked, through the damp twilight, to the bus shelter across the street. Its fluorescent tubes have flickered on — you can read the schedules and advertisements behind the plexiglass, as well as the stickers scattered over them.

One rectangular sticker catches your eye. It has the name of no band, the number of no locksmith, the logo of no corporation, and no image of Andre the Giant. It’s just a block of text. The first words read, “Why bomb libraries?”

The text is a passage from Implementation (2004), a novel written in small chunks formatted to fit on mailing labels that can be fed through a standard laser printer. Implementation is a sticker novel, in one sense, and also a kind of digital novel. Its authors, Nick Montfort and Scott Rettberg, don’t primarily distribute stickers. Instead, they distribute digital files, which others print on stickers and post in provocative locations. Readers sometimes encounter these stickers. But, like most stickers, they tend to be removed pretty quickly from the benches, doors, bathroom stalls, statue plaques, lampposts, bumpers, and other locations where they’re placed. What lasts longer are the photographs people take of Implementation stickers in interesting positions, and it is through these images that most people experience Implementation, the files for these images dwarfing the other contents of the Implementation website.

Encountering Implementation can be a mysterious experience. Finding a disconnected sticker of text, especially if well placed, can border on disconcerting. And the images of placed stickers can offer up their own mysteries: Where is that? What is that? But, despite the importance of digital technology for Implementation, the mystery is never about software. Implementation depends on digital processes — but ones we use every day, such that the processes of file downloading, laser printing, and photo sharing no longer attract our attention.

One might even argue that Implementation is not appropriate to discuss as a “digital” novel, but rather one that is distributed and documented digitally. Regardless, in its use of digital processes, Implementation stands as a relatively extreme example of something true of many digital fictions: they avoid the dilemma of the Eliza effect by employing processes that are conceptually simple and familiar, which are clearly exposed to the audience.

Different fictions approach this general strategy in different ways. Some, in a manner relatively close to Implementation, embed themselves in familiar digital contexts. An email novel, such as Blue Company (2001, 2002) by Rob Wittig, depends on processes of email transmission and reading — but, like the processes of laser printing, these have become completely naturalized for many with computer access. Similarly, the familiar link-following functions of web browsers and web servers makes it possible to construct fictions from interconnected networks of web pages, as with The Unknown (1999) by William Gillespie, Scott Rettberg, Dirk Stratton, and Frank Marquardt. Familiar processes are also the primary digital components of hybrid works such as the alternate reality game I Love Bees (2004) by Elan Lee, Sean Stewart, Jim Stewartson, and Jane McGonigal, which brought players into a fictional world composed of web pages, email messages, phone calls, physical settings and elements, and live performance — but defined its novel processes as rules to be carried out by human participants (rather than by digital computation).

However, another approach to digital fiction is more common than this sort of piggybacking on the processes of everyday information life. In this more common approach, digital fictions define their own versions of digital media processes in widespread use — or employ versions of these processes defined for use with digital media authoring tools. Such authoring tools range from Adobe’s Flash software (for interactive animations) to game engines (used repeatedly by the same developer or made available for commercial license) to tools identified with particular artistic communities (such as the Storyspace hypertext authoring system).

The most widespread fictions taking this approach are computer games. They tend to avoid the Eliza effect by employing versions of simple processes, familiar to those who play computer games within the same genre, and following conventions to expose the structure and actions of the underlying processes to their audiences. This chapter will look at two particularly well-crafted examples of computer game fiction — Star Wars: Knights of the Old Republic (Falkner, Gilmour, Hudson, Karpyshyn, Ohlen, Watamaniuk, and Watts, 2003) and Prince of Persia: The Sands of Time (Mechner, Mallat, Désilets, Langlais, Châteauneuf, and Lacoste, 2003) — as well as two commonly-employed operational logics that enable the fictions of computer role-playing games (RPGs). But, to understand all this, it is important to begin with a wider view of RPGs.

25 Responses to “EP 3.1: Digital Fictions and the Eliza Effect”

  1. scott Says:

    This chapter rocks!

  2. Dennis G. Jerz Says:

    I was hoping you’d maybe address the recent use of “interactive fiction” to mean “alternate reality” games, or the general category of such hybrid works. Nick has elsewhere discussed the turf war between critics of literary hypertext (who briefly claimed “interactive fiction” in the 90s), and I’ve recently noticed the term popping up to describe what you call “hybrid works.”

  3. noah Says:

    Interesting. I don’t necessarily want to wade into a terminological turf war, but I’m intrigued by how these things shift. Can you share a couple examples?

  4. noah Says:

    Glad you like it :-)

  5. scott Says:

    In some ways the term “interactive fiction” is an unfortunate one, in that the form is not the only form of fiction for the computer that is interactive. I don’t contest that “interactive fiction” generally denotes a particular form of computer literature/game with a conversational interface and a text parser, but I always have to give a two or three minute explanation that includes the phrase “You remember Zork?” when I meet someone new to the term.

  6. Dennis G. Jerz Says:

    Right. Among geeks, I say “text adventure,” but among literature types I say “interactive fiction.”

    I put a few examples of alternate uses of “interactive fiction” in a footnote here (the page takes a long time to download, and as each image loads it pushes the footnotes farther down the page, so it’s a bit awkward to link to)


    That includes a reference to McGonigal introducing “I Love Bees” as “a Web-based interactive fiction.”

    Scott, you are right “interactive fiction” isn’t a perfect term. But since much of my work in this area has been historical, when I use the term I always specify that I’m using it in this narrow context (i.e., to mean games like Adventure and Zork). Alternatives such as “adventure game” or “text adventure” or “adventure game” are too narrow (the content does not always depict an adventure, and the contemporary IF community continues to embrace the term “interactive fiction.”

    I’m not so much advocating for any particular use of the term, but it does seem that moving from Eliza to MMORGs is a bit abrupt without at least a brief detour through Adventure, Zork/Dungeon, and MUDs. The Sierra text/graphics hybrids are a good early example of early hybrid development, and the Infocom “feelies” are at least an embryonic example of creating a game that interacted with artifacts that exist in the real world (even if those artifacts only exist for copy-protection or fetish value).

    Maybe that will be part of your promised “wider view of RPGs”?

  7. J. Robinson Wheeler Says:

    I don’t know what the dilemma of the Eliza effect is (or just the “Eliza effect” itself, though I know what Eliza is), and I was hoping the article would tell me. Without that I’m just guessing at what the whole piece means.

  8. noah Says:

    J, that’s no problem. Hopefully reading these three posts will provide what you seek: 1, 2, 3.

  9. noah Says:

    The book definitely isn’t a straight shot historically, so hopefully these transitions aren’t too disruptive in their abruptness. After this chapter (in which the two main examples are from 2003) the next chapter goes back to the 1950s and 1960s, the following one focuses on an example from the 1970s, the next chapter opens with an example from this decade, and so on.

    There’s no part of the book that traces a historical chronology explicitly — and most of the things that are well covered in other areas of the digital literature field (like hypertext fiction and interactive fiction) are cited, rather than discussed at length. It might have been good to put in more about these things, but I wanted to organize the book around the argument (rather than the chronology) and I wanted to explore a small number of examples in depth (rather than do an exhaustive survey).

    That said, I appreciate the pointer to other uses of “interactive fiction” as a term. It might be good for me to put in a clarifying footnote (I’m using the term the way you do, and probably need to provide the kind of explanation Scott mentions) and note the alternative uses.

  10. josh g. Says:

    I’m curious, are the names listed next to the two game titles here chosen as part of a common citation format? ie. does the format dictate which team roles are highlighted? I haven’t seen a game citation include developer’s names at all before (but I haven’t caught up on nearly as much academic reading as I’d like so this may just be a confession of ignorance).

  11. noah Says:

    I think I’ll probably end up truncating these in the final text, so that they look like “(Falkner et al, 2003)” — with the rest of the information in the bibliography. The bibliography entry will probably look like this:

    Falkner, David, Steven Gilmour, Casey Hudson, Drew Karpyshyn, James Ohlen, Preston Watamaniuk, and Derek Watts. 2003. Star Wars: Knights of the Old Republic. BioWare, LucasArts. Multiple platforms.

    This isn’t what I’ve seen from other game scholars. For example, Game Studies gives citation guidelines that just include the studio and date. That doesn’t sit well with me, so I tried to do something more like citing other forms of collaborative media.

    But as for which team roles to highlight, I basically played it by ear. Any thoughts on how to make such decisions?

  12. Dennis G. Jerz Says:

    If you are eager to get games scholars to start focusing on the names of the game designers, maybe you’ll have to do your analysis on a much finer resolution.

    It’s not likely that a given reference to SW:KotOR will refer equally to the contributions of all the names you listed.

    In film and theater, you can discuss whether, for instance, a particular interpretation of A Streetcar Named Desire is due to Brando’s performance in the original Broadway show, Williams’s script, Kazan’s direction, or the set design, incidental music, lighting, etc. (contributed by artists whose names I’d have to look up).

  13. josh g. Says:

    It’s a tricky question. Team sizes are so huge and growing that even including just the leads from each area could start escalating to a dozen names per citation. And how do you standardize which roles to cite when different studios use the role names in different ways? eg. “Producer” meaning the equivalent of Creative Director in some studios, whereas in others it’s a project management role.

    But as far as your picks, looking at the MobyGames listings for KotOR and PoP:SoT, your choices look good, covering all the major team leads. So maybe my theoretical concerns above aren’t really that big a deal in most cases.

    Lately I’ve been learning from an artist who frequently works with sound art, so I can’t help but wonder if the Sound Designer should be included as well … but probably audio design gets the short end of the stick in film and other multi media as well, so oh well. =)

  14. noah Says:

    Dennis, it’s true, it isn’t until late in my discussion of KotOR that I mention Drew Karpyshyn specifically as the lead writer. But I do begin my discussion of PoP:SoT with Jordan Mechner. Probably the difference is exactly the one you suggest — I know more (and talk more) about Mechner’s specific contributions to his project, whereas I know less about how Karpyshyn functions in his collaborative work. Maybe as game scholarship begins to pay more attention to the contributions of the individuals involved in game creation our citation methods will change. And perhaps changes in game citation can help suggest to scholars that we should pay more attention to individual contributions.

    Josh, I think it’s interesting that you and I seem to have hit upon the same basic idea: listing the leads of the major creative teams. But that still leaves us with the question of name ordering, which is made more difficult by what you mention: the titles don’t mean the same things at different studios. Presumably we want to have the creative lead as the first name, but is that the designer, producer, creative director, or…? I have no problem knowing that Will Wright goes first when I cite SimCity, but who goes first for a WoW expansion? We could default to the credit ordering on the package, but that could be determined by an in-company hierarchy we’d rather disrupt.

    This might be one of the less-sexy issues that “software studies” needs to address. I know it sounds awfully boring to be on a committee to consider how to make academic citations. But a timely recommendation on these issues, backed by some smart thinking, could be quite helpful to the field.

  15. Dennis G. Jerz Says:

    Here’s a useful starting point, created for games journalists.


    I had considered writing a review of that book, for pretty much the same reasons you mentioned — setting new standards for citing the individual contributions of game designers might help translate, for a more general academic audience some of the sticky issues that go along with games studies. But your suggestion of a committee sounds like a better idea.

  16. noah Says:

    One place to start might be collecting a few suggestions from different groups and individuals of citation practices for games (as well as for other kinds of software and other forms of collaboratively-authored media). Does the guide above have any citation suggestions? I took a quick look at the wiki but didn’t find any.

  17. nick Says:

    Noah, it seems to me that the author of a WoW expansion pack is Blizzard Entertainment. When I try to determine an author for purposes of citing a work, I look to see who is put forth as the author on that work itself, as it is published. For commercial computer games, that almost always seems to be a corporation rather than a person. Early Activision games for the Atari VCS are a prominent exception – they do actually identify particular people as authors. But Jordan Mechner is not credited anywhere on the case or in the manual of Prince of Persia: The Sands of Time as having authored the game, while “Ubisoft” is clearly visible on the front and back. The only mentions of Mechner are in tiny type and indicate that he (1) owns trademarks, and (2) created Prince of Persia, which this game is based upon. If these notices have any implication to me about authorship, they are that he didn’t create this game; if he did, it seems like they’d mention that, too.

    Figuring out the contributions of various members of a team is certainly important to scholarship, but I don’t think such determinations would influence where I put something in a bibliography when a work presents itself as corporate-authored. Perhaps we have different ideas of what bibliographies and citations are supposed to do, though.

  18. noah Says:

    Nick, does that mean you cite Activision and Atari titles for the VCS differently — one set with the names of authors, the other just with the name of the company? To me, such an approach leaves the decision in the hands of the company, and their agenda isn’t necessarily mine. As you say, many companies seem to have the agenda of hiding authorship — like Hollywood under the studio system. When we participate in this I think we’re helping to reinforce one of the bad aspects of the current game industry.

  19. nick Says:

    I do actually cite the authors of Atari and Activision works differently, even when I know the Atari programmer, since these are presented as having a corporate author in the former case and an individual author in the latter. I would also cite Mark Twain as being the author of Huckleberry Finn, even though I know the person who wrote that work was actually named Samuel Clemens. I guess my feeling is that by using the claimed author, I’m making it easier for people to locate a work and look a work up on the bibliography, which to me is the main task of this part of a book – more than agitating for change in the way that people are credited, as noble a goal as this may be.

  20. nick Says:

    I should note that I do include the programmer’s name in a bibliographic entry which has “Atari” as its author, if I know who the programmer is. But the entry will end up being grouped with all other Atari carts as having a common corporate author.

  21. noah Says:

    Sounds like, rather than starting with a committee, it might be informative for the community to begin with a debate. Of course, it might be an odd debate, in that we’d probably suggest that people use whatever method suits their goals, rather than adopt a “right” method.

    Which is to say, I think you’re correct that it comes down to a question of what we hope to accomplish — both in the text and the bibliography. I’ve talked about my goals in the text, but an example bibliographic goal is for people to be able to find Peter Molyneux’s games together in a bibliography, even if some were produced while at Lionhead and others while at Bullfrog. I’d like the same to happen for David Crane’s games, though some were produced at Atari and others at Activision.

  22. Dennis G. Jerz Says:

    A bibliography page that lists all the Atari games together in one place, and all the Infocom games together in another place, is probably the most useful to me if I’m quickly scanning a bibliography to get a sense of what sources the author is using. I’m sympathetic to Noah’s interest in identifying the individual contributors, but I wonder if that sort of thing is better done via in-text references and the index.

  23. Paul B. Says:

    I would agree that for the sake of readability I’d truncate the in-text citations, as I found the long list of names cluttered and unhelpful in terms of the sort of quick identification for which the in-text citation is intended.

    As for the bibliography entries: I agree with Dennis that grouping solely by studio seems as if it would produce the most easily usable bibliography. I’m sympathetic in principle to Nick’s system of using the work’s claimed author, but in practice it seems to me to me that this produces the least useful bibliography for any purposes beyond verifying the book’s references, because it results in non-parallel citations: a person listed as the author of one project might not be listed as such in the same role on a different project with another studio, making such a bibliography extremely cumbersome for attempting to look at individual contributors, and it doesn’t produce any potentially useful lists of game by studio. Since I think a good bibliography has research use beyond simply providing references for the book to which it is attached, that system seems to me to lose a lot. While I understand your desire to recognize and group by individual contributors, Noah, as you have mentioned, who exactly should be the lead author is not clear on many projects, which is why I imagine grouping by studio to be the most consistent and useful choice.

    Whatever the format adopted, I would certainly go ahead and list what roles you think significant enough to merit listing in the entry, because that is useful information. But I would encourage you to list, perhaps parenthetically, what role each person mentioned in your bibliography entry performs. For any use of the bibliography beyond simple identification of your citations, knowing what roles the people involved performed with regards to a particular game is much more helpful than a simple list of names.

    Also, I know this isn’t entirely satisfactory concerning the problems of building a bibliography, but I wonder whether you might find a compromise in supplementing whatever system you choose for the book with a cross-reference online version. For instance, if you elect to construct your bibliography according to studio, you could have an online version that the user could elect to view by lead programmer, by producer, etc., or in which each name appearing in more than one entry was a hyperlink leading to a list of all the projects in which this person was involved. While I realize that what actually shows up in print is important, this seems as if it might provide a way for your readers to have it multiple ways, with all the advantages entailed by each. Or, for that matter, could you, say, include in your bibliography two separate lists, one arranged by studio and one arranged according to names as you desire, and simply specify up front which system takes priority in terms of in-text citations?

  24. noah Says:

    I certainly appreciate your position, but we should also remember that not all games are made by studios — while all games are made by people. The lack of studio is present for some of the most interesting games, from historical landmarks (Spacewar, Adventure) to today’s indie games.

  25. josh g. Says:

    Jumping back to this a bit late, I’ll add that one of the reasons this really stood out to me to begin with was that the long list of names is kind of awkward for a citation. But I can appreciate the motivation behind it, so I’m not sure if I’d prefer just listing a studio name or not.

    Also, just to throw out another monkey-wrench example, Valve Software lists all of their contributors to a project in alphabetical order. This might not be just a matter of being egalitarian, either. They’re one of many studios who are adopting more flexible team structures in which creative input isn’t funneled through a team lead. Projects using some variation of agile methods such as Scrum may change up who is leading which particular creative feature as they form up small groups to solve specific problems, and then change those groups around in a month or two. Or you can have groups like Portal’s development team, who claim that they simply make all major creative decisions democratically (if I remember correctly).

Powered by WordPress