"Chance (re)Collections: Twine Games and Preservation on the Internet"
by Tobias I. Paul | Xchanges 15.2, Fall 2020
Contents
Introduction
When Winter Lake’s digital game rat chaos was reuploaded after its host site lapsed, one journalist offered this imperative: “I highly recommend [the game], not least because it’s now playable again after vanishing for a few years, thanks to the hosting of Robert Yang” (Caldwell, 2017). The reuploaded game took on a kind of preciousness following its initial disappearance—after players were made painfully aware of its precarious status (DeNiro, 2019).
Examinations of text often focus on the incarnation of a work—the acts of composition, dissemination, and reception that make up an artifact’s lifespan. What may go unnoticed, then, is the afterlife—the decay, remembrance, and disappearance of text that occurs all-too-often in the digital sphere. While a physical text may enjoy some degree of stability through its medium, the longevity of a born-digital artifact is far more tenuous, relying on unstable platforms and largely arbitrary systems of preservation for survival.
In this paper, I explore how communities and creators are represented and forgotten through digital preservation. In the process, I discuss existing scholarship and media on digital impermanence, Internet archiving, hypertext fiction, and the ethics of preservation. Then, I analyze a set of 20 Twine games in various states of availability, focusing on several representative examples within the dataset. Through this analysis, I observe the quiet, ongoing crisis of Twine games’ decay, and find that these creations take on a kind of preciousness as a result of their ephemerality. This preciousness, if acknowledged and harnessed, activates a social impulse toward preservation—an impulse that content on the digital margins relies upon for its continued existence.
(Hyper)textual Renaissance
rat chaos is one of a growing number of digital games on the cusp of disappearance that were produced through the software Twine. Described on its website as “an open-source tool for telling interactive, nonlinear stories” (“Twine,” n.d.), Twine provides an accessible space for the creation of hypertext fiction—interactive, text-dominant creations (referred to in this paper as games, although terminology varies) navigated through linked words and phrases, which create an associative web of interconnected passages that “straddle the very threshold between game and literature” (Underwood, 2016). One of the software’s biggest appeals—aside from being free—is that creators need little to no knowledge of programming languages or videogame design to create with it. As a result, Twine lowers many of the barriers to entry of the traditional gaming industry: education, time, costs of development/publishing/marketing, and more. Further, Twine games are stored as HTML files and can thus be shared and played with near-universal ease. In the words of Twine creator merritt k (2014): “if you can navigate a website, you can play a Twine game.” This overall accessibility has led some to dub Twine “the videogame technology for all” (Hudson, 2014).
In some ways, Twine’s radical potential echoes early examinations of hypertext by digital humanities scholars, who touted hypertext’s capacity to “free the words from the page, the text from the line, the writer and reader from their separateness” (Johnson-Eilola & Kimme Hea, 2003, p. 417). At its core, hypertext is simply “a text which contains links to other texts” (“What is HyperText?,” n.d.) through hyperlinks, which can be used to bring the reader to a new section, document, or site. Early hypertext, however, challenged conventional relationships between writing, writer, and reader, by allowing readers to participate in the text; they could choose the order in which to navigate links, and in some cases even exert limited influence over the narrative. In hypertext, the role of the reader “lies halfway between the customary roles of the author and reader in the medium of print” (Bolter, 2001, p. 173).
In the end, hypertext fell short of this initial anticipation. The interactive electronic fiction created through hypertext remained obscure, with even its most notable pieces like Judy Malloy’s 1986 work Uncle Roger and Michael Joyce’s 1987 work afternoon, a story largely relegated to digital humanities scholarship, outside of mainstream attention. This stands in stark contrast to expectations and speculation that, over time, hypertext would “probably change our way of thinking . . . Perhaps, too, the notion that one can ever finish a ‘book’ may disappear” (Bevilacqua, 1989, p. 162). In the early days of the medium, many believed that the primary impact of hypertext would indeed be on storytelling—that through this new medium, “people would explore the story, not read it” (Johnson, 2013).
Rather than turning literary hierarchies on their head, hypertext became an instrument for exploring digital media that was otherwise conventionally constructed, with more traditional, linear forms of writing holding strong in social practice. The use of hypertext navigation became increasingly popular in the structure of the Internet, which can be seen today: “Someone tweets a link to a news article, which links to a blog commentary, which links to a Wikipedia entry. Each landing point along that itinerary is a linear piece, designed to be read from start to finish. But the constellation they form is something else” (Johnson, 2013). This practice domesticated hypertext—making it a central and conventional aspect of Internet navigation—while also “evacuating it of the revolutionary potentials it once held” (Johnson-Eilola & Kimme Hea, 2003, p. 417). Still, hypertext presented—and presents—something desirable, and indeed, valuable; even now, “the best hypertext still has to offer us is its complexity and openness” (Johnson-Eilola & Kimme Hea, 2003, p. 419).
Gaming on the Margins: Precarity, Value, and Loss
Perhaps as a result of these features, Twine has often served as a storytelling medium for marginalized voices—like that of Anna Anthropy, a prominent Twine creator whose book Rise of the Videogame Zinesters: How Freaks, Normals, Amateurs, Artists, Dreamers, Drop-outs, Queers, Housewives, and People Like You Are Taking Back an Art Form is widely credited with putting Twine on the map in 2012. Anthropy, like a number of the software’s most well-known creators, is a transgender1 woman, and she characterizes Twine as “fertile territory for marginalized voices to grow” (as cited in Harvey, 2014, p. 99). Twine games are known in particular for their prominence in the queer games scene (Pozo, 2018), and more generally for stories that explore unconventional, emotionally-charged topics that place the player in a highly interactive role—stories that “wouldn’t be the same in a static form” (k, 2014). The queer subculture that developed around Twine emerges as a kind of digital counterpublic space, seeking to “question or even . . . subvert some of the dominant practices of the public sphere” (Alexander & Rhodes, 2012). In this sense, Twine queers game design by challenging the so-called hegemony of play that dominates the mainstream gaming industry and constructs an exclusionary and overwhelmingly white, heterosexual, and cisgender male constituency of both creators and players (Harvey, 2014, p. 96). Further, Twine queers narrative itself, as the medium “questions assumptions of book-bound and print-imitative composition . . . [and] more complexly allows users to shape text itself by turning it into a haptic mechanism of progression: you must click to proceed” (Milligan, 2017).
However, Twine’s location on the margins of gaming, and culture more broadly, comes at a price. Twine games’ creation tends to be “contingent on goodwill, just enough, [and] temporary measures, a reality that begs the question of sustainability and the livelihoods of these game designers” (Harvey, 2014, p. 104). In a similar manner, the sharing and preservation of Twine games is dependent on Twine’s creators and players, as well as on the stability of the Internet—a precarious position for a community’s digital cultural memory. This precarity has already taken effect, with untold numbers of games becoming unavailable over time.
Between Twine’s creation in 2009 and the peak of its popularity in 2014, over 1,500 games were created (Arnott, 2014). Of the games I examined—nearly all of which were released within that same timeframe—roughly a third were still available on their original sites until sometime between 2016 and 2019 (see Appendix B). Only in recent years have we begun to observe this phenomenon of ephemerality in full, as more and more Twine games fall out of circulation. The personal archive of Twine developer Leon Arnott—singlehandedly responsible for the majority of archived Twine games in this dataset—was last updated January 21, 2014. Internet archiving tools offer expansive but noncomprehensive opportunities for preservation and often depend on the goodwill or passivity of the games’ creators. People are still creating and sharing Twine games today—but what happens when those, too, start to disappear?
There is perhaps something cogent in Twine as a transient medium. In some sense, attempts at preservation may derail or otherwise interrupt this—so perhaps we ought to be more comfortable with impermanence. However, larger ramifications lurk beyond these margins. At its most basic level, the loss of digital games means “we will lose access to the history and culture of contemporary games and find it impossible to trace the influence of interactive digital games and simulations on other forms of play, leisure, entertainment, communication, learning, and work” (Lowood et al., 2009, p. 140). Further, when communities lose access to a Twine creation, they often lose more than just a game. By the nature of the platform and its users, many Twine games tell diverse stories—especially from LGBTQ voices, a community that has long been underrepresented in mainstream media (Townsend & Deerwater, 2019). Twine preservation, then, preserves not just a digital experience, but a cultural artifact.
The Internet Is(n’t) Forever
Contrary to common belief, information on the Internet is ephemeral: “once a [website] moves, is taken down, or is updated, it can be impossible to retrieve the original version” (Pittman, 2018, p. 53). Most people have encountered the impermanence and instability of the Internet in some form or another. YouTube videos are taken down; webpages are removed, their URLs yielding only 404 errors (a phenomenon known as link rot); blank boxes appear where images used to be. The Internet is inherently unstable, retaining and erasing information in unpredictable manners at unpredictable times (De Kosnik, 2016, p. 48). Rather than an archive or repository of information, it acts as a data-sharing hub that relies on individual human maintenance and upkeep to function.
This presents a particular issue when faced with increasingly born-digital content, which offers no physical, savable backup for digital media. Instead, this data “resides in undisclosed locations inside the enclaves of corporate server farms, on disk arrays we will never even see or know the whereabouts of” (Kirschenbaum, 2013, para. 6), thus denying a hands-on means of preservation. Some even worry that society is headed for a digital dark age as a result, wherein “only a tiny portion of the digital record we’re creating right now will be preserved or even readable by future generations” (Wood, 2019).
In the absence of widespread, mainstream archiving efforts, the preservation of digital media is dependent upon the stewardship of its digital communities. One tool in the arsenal of amateur and professional preservationists alike is the Internet Archive’s Wayback Machine (WM). Developed in 1996, the WM serves as an extension of the Internet Archive’s mission: “Universal Access to All Knowledge” (“About the Internet Archive,” n.d.). The WM is a software which accesses publicly available websites through automated bots called crawlers (a process known as crawling) and takes interactive snapshots of what particular webpages look like at the time of access. That information is stored and preserved in the WM, offering a public record of what the sites looked like at different points in time. This functions as a kind of version control—a longitudinal record of Internet history.
While it may sound like exactly the solution the Internet needs, the WM is not without its limitations; at times, the site’s own policies work against its purpose (De Kosnik, 2016, p. 50; Lepore, 2015). Most notably, the WM allows website owners to block crawlers through the inclusion of a simple text file, robots.txt, in the site’s root code (“Using the Wayback Machine,” 2018). The policy exists to address potential objections from website and content creators, who may wish for their work to be excluded from these automated crawls, and it potentially shields the WM from some degree of legal/ethical ire. Not only will the addition of a robots.txt prevent future crawls; when a crawler encounters a robots.txt for the first time, it will also remove all previously saved versions of the site. This policy of total erasure has earned robots.txt a reputation as “the [Internet] archive’s kryptonite” (as cited in De Kosnik, 2016, p. 50).
Unlike traditional memory institutions—brick-and-mortar archives, libraries, museums, and so on—the digital archive is not a centralized one. The prevalence of non-professionals engaging in digital preservation has led scholars to describe the field of digital cultural memory as an “unregulated, decentralized and multi-voiced character [that] plays havoc with our inherited routines and established protocols” (as cited in De Kosnik, 2016, p. 27). Rather than operating from “core archival concepts, including appraisal, original order, provenance, and the very nature of the record” (Kirschenbaum, 2013, para. 8), non-professional archivists work together to collect and preserve digital media, often in accordance with their own values—seeking to preserve artifacts which they deem meaningful or significant. Digital-based efforts toward preservation serve to illustrate that, while the Internet is itself an inadequate archival technology, it can be harnessed to serve archival purposes in what Matthew Kirschenbaum defines as an essentially social effort (as cited in De Kosnik, 2016, p. 51).
Paradoxically, this social reliance can have an isolating effect on digital media with more limited online followings, as is the case for most Twine games. When awareness and enthusiasm for digital content is anything other than widespread, preservation becomes a patchwork endeavor; individuals save and index media by way of their own limited time, interest, and resources, or even by sheer chance. Even more established organizations and websites, like the Interactive Fiction Database (n. d.), rely on manual input and individual contributions in their indexing of Twine and other texts. In this cobbled-together form of preservation, Twine games can disappear without any indication they existed in the first place. If this aggregation and preservation is non-centralized and reliant on individual efforts, what does this mean for Twine games as a genre, and the people who play them? What gets preserved, by whom, and why?
[1] Hereafter, trans.