Behind the Scenes at ApertureScience.com: Portal and Its Paratexts

Peritexts, Epitexts, and “The Book”

The introduction to this article cluster situates discussions of video game paratexts in broad terms. In this section I will examine terms inherited from Genette which need rethinking in the context of video games, and Portal in particular: namely, Genette’s distinction between peritexts (internal paratexts that are materially connected to the primary work, like a book’s cover) and epitexts (external paratexts at some degree of physical removal but nonetheless related to the work, like a critic’s review or author interview). Even in traditional book publishing, these lines between so-called external and internal paratexts can blur; for example, laudatory lines from a critic’s review may become a promotional blurb on the back cover of subsequent printings. Peritext has become a particularly troublesome sub-category in video game studies, with some researchers either lamenting its neglect (Dunne, 2016), or wrestling with its applicability (Carter, 2015, pp. 313–14, 337; Jones, 2008, pp. 25–26).

Conversely, the expanded definition of paratextuality associated with Consalvo has also attracted some recent criticism (Švelch, 2020; Dunne, 2016; Rockenberger, 2014; Jara, 2013), partly for the intimidating breadth of new materials it opens up for paratextual studies, and partly for its tendency to emphasize epitexts at the expense of peritexts. Although I do not join the critiques of Consalvo’s treatment of paratext, and see more advantages than disadvantages in its breadth of application to the rich world of video game epitexts, there is something to be learned by grappling instead with the more difficult sub-category of peritext as it travels—with some friction—from the domain of books to the domain of video games.

If video game studies has tended to neglect peritexts, the opposite has taken place in the adjacent discipline of book history, where a recent study laments the field’s neglect of epitexts (Murray, 2018, p. 171). To book historians like myself, paratext as a workaday term usually denotes physically connected things in the same bound book: prefatory letters, frontispieces, epistles to the reader, dedicatory poems, tables of contents, maps, indexes, and other texts usually bound (or re-bound) at the beginning or end of a book (also known in this field as preliminaries and postliminaries). These paratexts—or, as Genette would stipulate, peritexts—may or may not be printed at the same time as the primary text, or even by the same printer, but they always involve voices beyond the author’s.

The examples I have mentioned here tend to be found in early modern European books specifically, and book historians working in that field have produced many studies of peritexts as material sites where a book’s authorship, commodity status, and material and symbolic production (to use Bourdieu’s terms; 1993, p. 37 and passim) overlap and leave recoverable traces (Silva, 2020; Berger & Massai, 2014; Smith & Wilson, 2011). For early modern book historians, preliminaries are not just thresholds by which the reader enters the book, but also border zones where the book’s several relationships to the world are negotiated. Books and video games alike tend to begin with acts of calibration, whether the tweaking of brightness settings or of reader/player expectations. In this sense, early modern books may serve as strangely apt analogies for video games, in terms of their paratextual qualities.

Yet video games are not books, and their bibliographic qualities can be overemphasized when literary scholars negotiate their own crossing of disciplinary thresholds. For any discussion of video game paratexts that relies upon Genette, I would emphasize a caveat arising from the differences between books and software. Namely, it is deceptively easy to speak of books as supposedly straightforward containers for a literary work, whereas video games are recognizably more complex in their ontologies and boundaries as artifacts. Books are materially complex, too, but centuries of print culture have successfully built up the illusion that copies of books are exactly that: copies, defined by perfectly reproducible content capable of transcending differences in form, resulting in a self-sufficient object in the hands of a reader. With video games, the illusion of perfect equivalence between content and container is much harder to sustain. One can place a 1922 first edition of James Joyce’s Ulysses on a table in front of a class and plausibly declare, “this is James Joyce’s Ulysses”; it is a different matter to place an original Half-Life installation CD-ROM on the same table and claim the game itself sits before them. (And a smart class would immediately question what the word original means in this context with respect to versions of the game; see Altice, 2015, Appendix A, and Young, 2016.) Video games, in contrast to books, possess the distinct advantage of being highly resistant to illusions that they are straightforward objects.

A related caveat about relying too much on Genette’s understanding of paratext is that we should remember that he was working from a twentieth-century literary scholar’s understanding of books, in which the single-author novel becomes the paradigmatic form of “the book,” as the essentializing phrase goes. “The book” in this reductive sense is typically credited to a single author, with the contributions of editors and others usually in the background, whereas video games are more recognizably collaborative (especially in cases where there are credited writers, e.g. Half-Life’s Mark Laidlaw, mentioned above, or Portal’s Erik Wolpaw and Chet Falisek). ⁶ As Jonathan Gray has argued in the context of film and television paratexts, but with broader applicability, “authorship is a fractured process, spread out over time, not simply preceding the work or object in question” (2016, pp. 34–35). A broader exploration of video game paratextuality in relation to books, films, television, music, and other media forms would be worthwhile, but beyond the scope of this article. For now, my point is that video games embody the condition of fractured authorship more visibly than many other media forms, hence the need for more disciplinary crossover between video game studies and textual studies (at least of the non-book-centric kind). Paradoxically, video games may remain more visible as constructed artifacts in the early 21st century, whereas books are—supposedly—just books.

If peritexts have been more difficult to discuss in video game studies than epitexts, these material differences from books may explain why. The further we get from books, the less coherent Genette’s category of peritext becomes, and yet the more we need a concept like it which links the fractured nature of authorship (as Gray calls it) to the specifics of materiality. On this front, Genette’s work by itself offers little help. Ironically, for all its encyclopedic breadth and attenuated subcategories, Genette’s Paratexts was also a remarkably narrow study in historical, geographical, and cultural terms, confining itself mostly to 20th-century European books. The cautionary tale to be drawn for video game scholars, I would suggest, is that Genette’s original conception of paratext is less important than how the concept has morphed and mutated in other disciplines and among other kinds of scholars. ⁷ Indeed, as Genette himself implied about his own work (1997, p. 343, quoted in this article cluster’s introduction), the question to ask is not whether another video game researcher’s understanding of paratext is faithful to Genette. Rather, the question is what good the concept can do in the hands of those who take it up, and what forms of critical purchase it affords us in our interpretation of video games as cultural artifacts.

My approach, then, is not to re-ground the concept in its origins in Genette’s literary structuralism, but instead to draw from its more historically and materially nuanced takeup by scholars in textual studies—a field whose attention to the interplay of medium, materiality, and meaning parallels video game studies in many ways. In this regard I am echoing one of the other early importers of the idea of paratext in video game studies, Steven Jones, whose book The Meaning of Video Games acknowledges Genette but also ties the concept to book history and other forms of textual scholarship (i.e., the study of the production, transmission, and reception of texts in manuscript, printed, or digital form, closely related to the fields of book history, bibliography, and scholarly editing). As Jones points out, following his summary of Genette, “The paratext is thus about the material conditions of a book’s production, publication, and reception, how a book makes its way out into the world and comes to mean something to the public audience that receives it” (2008, p. 7). One could readily substitute “video game” for “book” in this assertion, and indeed one of the strengths of textual studies as a field is that its analytical and descriptive methods need not be confined to books, despite the field’s literary origins. Indeed, forms of textual scholarship have already been adapted surprisingly well to video games—taken together with the articles in this cluster, these studies point to the potential for interdisciplinary growth in both fields. ⁸

For all these reasons, it is important to understand the disciplinary currents that run through Genette’s work and its reception. If the study of epitext is valuable because it orients us toward social reception and the political economy of cultural industries, as demonstrated in the work of Murray in book history and Consalvo and others in video games studies, then the study of peritext is valuable because it orients us toward the particularities of the medium (Jara, 2013, p. 42). The next sections will pursue this second premise through the corridors, test chambers, neurotoxin storage rooms, and other backstage areas of Portal—a game which depends upon malleable thresholds as its central gameplay mechanic, and whose thematic and material qualities alike reward inquiry at its edges.

“You Really Shouldn’t Be Here”: Backstage at Aperture Science

    PLAYER: … We do on stage the things that are supposed to happen off.Which is a kind of integrity, if you look on every exit being an entrance somewhere else.

Portal is a game of thresholds. Its core gameplay mechanic allows the player to place linked blue and orange portals on various surfaces in the game world, in essence creating a human-sized wormhole between non-contiguous spaces. Moreover, the game’s realistic physics encourages players to experiment, combining the effects of gravity with strategic portal placement to reach seemingly inaccessible areas. Portal’s initial levels are essentially puzzle-solving exercises set in an environment reminiscent of the first Half-Life game, as the player guides their silent first-person protagonist through the concrete corridors of an apparently empty research facility. The Aperture Science Enrichment Center seems to be devoid of other humans, and the test-chamber windows where one would expect to see researchers recording observations are strangely empty. Yet in all of the chambers, and in each elevator ride that connects them, a computerized female voice makes periodic announcements over a loudspeaker, explaining (selectively) the conditions of the test and offering dubious encouragement—and all the while displaying a questionable grasp of research ethics protocols for human subjects. ⁹

As one progresses through Portal’s test chambers, the voice betrays an increasingly passive-aggressive attitude toward the player, inflected through the language of bureaucratic obfuscation, resulting in the kind of finely tuned contempt one might encounter when dealing with, say, a commercial airline or university administration. By the end of Portal, we learn that the computer voice belongs to a rogue AI named GLaDOS (for “Genetic Lifeform and Disk Operating System”). Some years prior to the game’s events, GLaDOS achieved sentience (on the company’s Bring Your Daughter to Work Day), then circumvented her creators’ safety protocols, took control of the massive underground facility, and flooded it with a fatal neurotoxin. GLaDOS then continued Aperture Science’s testing regimen, deep underground with no human oversight, using captive human subjects like the player. Portal shares the elements of this backstory slowly and selectively, and the player spends most of the game exploring a supposedly closed system (Wendler, 2014).

However, cracks in that system allow the player to access backstage areas of the testing facility, where test subjects are not supposed to venture. ¹⁰ The first appears in Test Chamber 16, where someone has used crates to block open one of the chamber’s moveable walls, granting access to a makeshift lair with scavenged supplies and graffiti (including the first instance of the game’s iconic catchphrase “the cake is a lie,” referencing GLaDOS’s dubious promise of cake as the reward for a successful testing cycle). Soon after, in Test Chamber 17, a stuck hydraulic arm lets the player teleport behind another moveable wall to explore a more elaborate lair (Figure 1). Again, this backstage space is empty but full of crazed scribblings on the wall. ¹¹ As the player discovers each lair, the revealing of a backstage area subverts any genre-based expectation that Portal is a pure puzzle game.OPEN IN VIEWER

The sneaking-backstage trope is used to its most dramatic effect in the moment first-time players could mistake for the game’s ending, at the completion of Test Chamber 19, as a moving platform carries the player not toward her promised cake but into an incinerator. A bleaker and less ambitious game might have ended at this point, but not Portal. At this crucial threshold, the nature of the game changes radically and unexpectedly, as the player uses the portal gun to escape GLaDOS’s trap and infiltrate spaces normally off-limits to test subjects. Working her way through previously inaccessible corridors, air ducts, and utility rooms, and using what she learned in the test chambers about the portal gun’s affordances, the player eventually reaches the final confrontation with GLaDOS. The player also learns more about Aperture Science along the way thanks to various environmental objects, including a slideshow left running in a conference room which depicts Aperture Science’s underdog relationship to the Black Mesa Research Facility (from Half-Life) in the competition for research funding.

Portal’s thematic and gameplay elements thus mirror one of the central questions of paratextuality: where are the boundaries of the system, and what authority controls who and what can cross those boundaries? The identity and character of Aperture Science is hinted at through environmental storytelling in Portal, but the game’s most pointed satire of corporate research appears via one of Portal’s chief paratexts, an official tie-in website created by Valve (2007) on the domain ApertureScience.com. As we shall see, this particular paratext gave determined players access to a backstage area of another kind.

ApertureScience.com and Storytelling Via Command Prompt

From Portal’s release in 2007 until 2010, visitors to the ostensible corporate website for Aperture Science Laboratories were greeted only by a black screen, command prompt, and blinking green cursor. ¹² In keeping with the game’s themes of minimalism, alienation, and control, the command prompt presented visitors with no options, no hints, and little help. ¹³ Indeed, entering the command “HELP” generates a message that places visitors in the position of a hapless Aperture Science employee at a workstation, trying to decipher a bureaucratic enumeration of possible crisis scenarios:

If this is an actual plea for help in response to a hazardous material spill, an explosion, a fire on your person, radiation poisoning, a choking gas of unknown origin, eye trauma resulting from the use of an emergency eye wash station on floors three, four, or eleven, an animal malfunction, or any other injurious experimental equipment failure, please remain at your workstation. A Crisis Response Team has already been mobilized to deliberate on a response to your crisis.

(There is, of course, no User Handbook.) The paratextual story that ApertureScience.com tells—even before the user begins to explore its hidden sections, as described below—thus begins with its command-line interface, which places the user in the position of an employee unprepared for the enigmatic task in front of them.

However, if the website visitor experiments with other commands typically found in command-line interfaces, and eventually types “LOGIN,” the system asks for a username and password. Any username will work provided the user also guesses that the password is “PORTAL” or “PORTALS.” The system then switches to a prompt that more closely resembles the MS-DOS operating system, created by one of Portal’s satirical corporate targets, Microsoft:

GLaDOS v1.07 (c) 1982 Aperture Science, Inc.

From here the user can enter standard DOS-style commands such as “DIR” and “HELP,” with the latter being more helpful than before, resulting in a full list of possible commands—including the ominous “INTERROGATE.” The “DIR” command, a standard DOS instruction to list the files on a disk, generates a response that will look familiar to those familiar with DOS’s command-line interface:

DISK VOLUME 255 [NEW EMPLOYEE WORKSTATION]

It is this (simulated) executable file “apply.exe” which constitutes one of Portal’s most interesting paratexts.

In this case, the text within “apply.exe” makes a difference for any interpretation that is sensitive to Portal’s satire of corporate culture, positivism, and research ethics. The “apply.exe” file, once executed, presents visitors with one of the standard instruments of social science research, a questionnaire, apparently designed to identify potential test subjects from among Aperture Science’s employees. For the sake of space I will not attempt to summarize every step of the questionnaire, though the opening screen is worth quoting in full as an example of satirical world-building through language alone:

Loaded: ENRICHMENT CENTER TEST SUBJECT APPLICATION PROCESS

This one paratextual screen does more than most of Portal’s actual gameplay to characterize Aperture Science as a labyrinth of bureaucracy, akin to those depicted in Franz Kafka’s novel The Trial (2009 [1925]), Terry Gillam’s film Brazil (1985), and more recent technology-focused series like Black Mirror (Brooker, 2011–2019) and the BBC satire W1A (Morton, 2014–2017).

The rest of the questionnaire continues the theme of authority sliding into absurdity. Questions such as “If given a choice, what would you like to be called?”, “What is your favorite number between 31 and 71?”, and “What is your favorite color?” (with 95 named options) combine arbitrariness, irrationality, and a dubious understanding of human responses. Some questions hint at an unpleasant experience ahead as a test subject in the Enrichment Center, such as the ominous inquiries, “Which interrogation technique do you think would be most effective on you?” and “If you disappeared tomorrow would anyone miss you?”. Other questions signal distrust or contempt toward the test subjects in the form of skill-testing questions that are impossible to answer. When the user is asked to correctly recall which of the 95 possible answers they gave for their favorite color, regardless of accuracy the questionnaire follows with “Lying about my favorite color makes me feel: 01] Sorry 02] Not Sorry.” Questions also begin to reveal a fixation with cake—GLaDOS’s promised reward for testing—and an ASCII-art image of a slice of cake briefly flashes on the screen at times. After the final question, the system prompts the user to complete the process by entering their 64-digit “UIN(+L)” number, which they were shown at the beginning of the test (in flashing characters) and asked to memorize. Even if the user somehow manages to enter the correct UIN(+L), they are told their response does not match and asked to “Please REMAIN AT YOUR WORKSTATION until a Computer-Aided-Enrichment Crisis Team arrives.” The cursor continues to blink but the prompt ceases to accept commands.

The impossibility of correct answers to the UIN(+L) and favorite color questions is confirmed by evidence from the Flash script that simulates the ApertureScience.com workstation interface. ¹⁴ These inbuilt flaws support the interpretation that Aperture Science—and, by extension, GLaDOS—conceal a basic irrationality beneath a veneer of managerialism. Even at the level of Flash source code we can also see one of Portal’s and Portal 2’s running themes: questioning the fairness of tests. Overall, the questionnaire places the supposed Enrichment Center applicant and Aperture Science employee in the position of being examined by a powerful yet irrational system, measured by impossible and inhumane standards, and expected to comply with the corporation’s cheerful disregard for human comfort and safety. The Enrichment Center application thus functions ironically, with a seemingly straight-faced institutional communique betraying more about the institution’s flaws, contradictions, and irrationalities than it intends. Indeed, the questionnaire reveals a pathology at the heart of this technocratic, positivist research process, even as it attempts to make the user the object of its epistemophilia with absurdly elaborate questions and multiple-choice answers. It is a pathology which, in the game’s story, originates with GLaDOS’s takeover of Aperture Science (her blinkered AI’s understanding of humans lies behind the questions) but is also indistinguishable from Aperture Science’s own corporate dysfunction and inhumanity.

The difference that “apply.exe” makes as a paratext, then, is that it provides considerably more detail than the game itself can about GLaDOS’s and Aperture Science’s entwined pathologies. It also does so textually, in the form of written screening questions and answers that place the reader/player/applicant in a different subject position than the first-person gameplay. In this sense, the experience of a game’s paratext can also involve a change in the medium of a game’s storytelling modes. Aperture Science is one thing when experienced as a three-dimensional space through which one navigates a puzzle-solving first-person avatar; it is another thing when experienced by a website visitor on the receiving end of a vaguely threatening application form, where there seem to be no good choices.

However, there is more to learn about Aperture Science’s nature and corporate history from a second (simulated) file, which may be accessed using information hidden as an Easter egg in the game. In the Test Chamber 17 lair mentioned above, behind some moveable debris the player may discover a username (“cjohnson”) and password (“tier3”) written on the wall, with “Trust Me” written next to it in red (Figure 2). A player aware of ApertureScience.com and its command line can use these credentials to log in, as described above, and access the system with a slightly different prompt:

GLaDOS v1.07a (c) 1982 Aperture Science, Inc.

OPEN IN VIEWER

Figure 2.

A username and password for use at ApertureScience.com, included as an Easter egg in the room shown in Figure 1. Screenshot taken by the author.

With the visitor now using an administrator account, the DIR (directory) command also produces a different result:

DISK VOLUME 255 [WORKSTATION CJOHNSON]

The “notes.exe” file, not previously visible with an ordinary login, appears as the player’s reward for following the Easter-egg trail into the Aperture Science website.

Once executed, “notes.exe” presents the user with Portal’s most revelatory paratext, a history of Aperture Science from its founding (which I quote here only as far as its first screen ¹⁵ ):

1953 – Aperture Science begins operations as a manufacturer of shower curtains. Early product line provides a very low-tech portal between the inside and outside of your shower. Very little science is actually involved. The name is chosen to make the shower curtains seem more hygienic.

1956 – Eisenhower administration awards Aperture a contract to provide shower curtains to all branches of the military except the Navy.

1957 – 1975 – Mostly shower curtains.

1978 – Aperture Founder and CEO, Cave Johnson, is exposed to mercury while secretly developing a dangerous mercury-injected rubber sheeting from which he plans to manufacture seven deadly shower curtains to be given as gifts to each member of the House Naval Appropriations committee.

The “notes.exe” file continues on in this vein, with the company’s founder, Cave Johnson, becoming increasingly eccentric in his decline. He conceives of a three-tiered research and development program to keep Aperture Science in business, with the tier 3 project (hence the password) resulting in government funding for a “man-sized ad hoc quantum tunnel through physical space with possible applications as a shower curtain.” In 1986, in response to competition from Black Mesa (of the Half-Life games), Aperture begins work on the artificially intelligent disk operating system that eventually becomes GLaDOS. The final entry in the file brings the story up to date:

Several Years Later – The untested AI is activated for the first time as one of the planned activities on Aperture’s first annual bring-your-daughter-to-work-day.

Presumably, Portal’s silent protagonist (named as Chell in the end credits) was one of the daughters brought to work on the day GLaDOS took control, and then kept in stasis until GLaDOS determines her ready to become a test subject, at which point the game itself begins.

Yet the supposedly self-evident meaning of the game itself is exactly what video game paratexts like these call into question. The history of Aperture Science related in “notes.exe” fills in many of the blanks that a gameplay-only experience of Portal would omit. For one, “notes.exe” introduces an entirely new character, Aperture founder Cave Johnson, prior to his portrayal by J.K. Simmons in Portal 2. The “notes.exe” paratext also fleshes out Aperture Science itself as a kind of corporate character, whose technocratic mania and obsessive detailism flow from Johnson’s own failures and paranoia. Overall, “notes.exe” provides a detailed timeline for Portal’s backstory where the game only offers oblique hints, and opens the mystery box that the game itself leaves closed.

It is a missed opportunity, then, that these materials from ApertureScience.com—which are both paratextual and canonical—have gone unmentioned in academic publications about the game, despite being germane to many of the studies cited in this article’s introductory section, especially the studies concerned with agency and power in Portal. Here we also have an opportunity, following Consalvo’s emphasis on video game Easter eggs (2007), to understand the malleability of the distinction between texts and paratexts (and peritexts and epitexts) through the connection between ApertureScience.com and the “cjohnson”/“tier3” login details shown in Figure 2. The experience of following the path leading from the login details hidden in Test Chamber 17 to reading “notes.exe” hidden on the ApertureScience.com website offers a lesson in paratextuality as a storytelling strategy—one that can imbue our experiences of complex digital materials with the pleasure of sneaking backstage. In the final section, I will follow this thread into Portal’s own files and digital assets to consider the implications of working with actual digital files to recover supposedly ephemeral (and well-hidden) paratexts.

Computational Memento Mori: Files, Ephemerality, and Evidence

Critics who focus on the materiality of video games and other digital artifacts often do so via questions about hardware and collector cultures (Guins, 2014), or software versions and ephemerality (Newman, 2012), or code-level traces of human labor and the forensic methods required to read them (Kirschenbaum, 2008), or the challenges such objects pose for digital preservation and archiving (McDonough et al., 2010; Winget & Murray, 2008). These are all important lines of inquiry. Yet the example of ApertureScience.com as a paratext raises a more basic, even mundane question about materiality in the practical sense: what are the materials we place on the table, so to speak, when we say we are studying a given video game? Paratextual theory of the kind discussed in this article can prompt video game scholars to answer this question differently, but as we have seen above, Portal itself points to an expanded definition of its own materiality via the Easter egg shown in Figure 2. What has not been much considered by those who argue for the expanded definition of paratextuality, however, is that paratextual analysis of video games as digital artifacts may also require an understanding of files, formats, and code-level materiality.

Just as ApertureScience.com called attention to operating systems, files, and command-line interfaces, so did another of Portal’s paratextual puzzles lead players to investigate a mystery in the game’s own installation and update files. In March, 2010, three years after Portal’s release and one year before the release of its sequel, Portal 2, players who had recently updated their game began to notice mysterious radios showing up in levels where they had not appeared before. Given the minimalist design and controlled spaces of Portal’s test chambers, the new radios stood out unsettlingly, like arriving home to discover one’s living room rearranged by some unseen presence. The relatively few radios that had appeared in the game previously would normally play a tinny jingle version of the game’s end-credits theme song, but some players discovered that moving the new radios to certain parts of the test chambers caused them to emit Morse code or sounds resembling dial-up modem connections. Suspecting more Easter eggs, some players dug into the update’s source code and located the sound files that matched what the mysterious radios were playing. There were 26 new sound files, all of them encoded in the standard WAV audio format, all curiously named “dinosaur1.wav” and so on, and all added via the March 1, 2010, software update. Digging deeper, players then discovered that several of the new sound files had images encoded into them using techniques connected with Slow-Scan Television (SSTV), an image-transmission technology which has existed since the 1950s (making it contemporaneous with Aperture Science itself), and has been used by NASA and ham radio enthusiasts alike to transmit images via radio signals. By 2010, consumer-grade software had become available for encoding and decoding SSTV images in analog signals, which is what Portal players used in 2010 to recover the hidden images from the radio transmission audio files. ¹⁶ Figure 3 shows the most unsettling of the image files, decoded from the file “dinosaur9.wav” using freely available software; other hidden images appear to depict more mundane objects within the Aperture Science laboratories (all documented thoroughly by Portal fans on discussion boards and wikis). As it transpired, these mysterious out-of-place digital artifacts were part of an elaborate paratext: an alternate reality game (ARG) created by the game’s developers to announce Portal 2. ¹⁷ OPEN IN VIEWER

More than any other example we’ve considered, SSTV images like the one in Figure 3 serve as a media allegory for the complex materiality of video game paratexts, especially ones that might be considered peritexts, to return to Genette’s term for materially connected paratexts. As internal yet partially hidden paratexts, this and the other SSTV-encoded images resonate in particular with Boluk and LeMieux’s chapter “Stretched Skulls: Anamorphic Games and the Memento Mortem Mortis,” which likens Portal to early modern anamorphic paintings like Hans Holbein’s The Ambassadors (1533; Figure 4). As Boluk and LeMieux put it, Portal and The Ambassadors alike “intentionally disrupt the immediate experience of space” (2017, p. 84). Their comparison is strangely apt, and although they do not mention Portal’s own skull image, shown in Figure 3, its hidden presence in the game files evokes the idea of anamorphic painting in code. Like the game, Holbein’s painting uses strategically placed objects and implicit environmental storytelling to present what at first seems like a perfectly ordered world, but on closer inspection harbors disorder and death, via small details like a broken lute string and larger anomalies like a non-diegetic blob which resolves into a skull when viewed from a certain angle (Greenblatt, 1980, pp. 17–21). The game and painting also happen to use the image of a skull—a symbol of death, but also of humanity laid bare behind its facades—as a paratextual object whose presence becomes apparent only through the decoding of an analog signal. With both the painting and the SSTV software, the act of decoding becomes the performance of the memento mori: as the skull resolves before our eyes, we realize it was there all along.OPEN IN VIEWER

As Boluk and LeMieux suggest, the Portal ARG also exemplifies the broader phenomenon of “massively collaborative play in the form of online flash mobs, digital scavenger hunts, encrypted codes, multimedia riddles, and adaptations of other ludic happenings and folk games” (2017, p. 318, n. 17). These analogies hold true for the kinds of social phenomena the Portal ARG generated, but we should also consider what Portal, as a digital artifact with material qualities of its own, became as it changed with the ARG updates in March, 2010. Analogies to help video game scholars rethink the materiality of peritexts might be found among the objects that textual scholars study, from palimpsests with their still-readable erasures (Carruthers, Chai-Elsholz, & Silec, 2011), to uninked type used as bracing by printers which sometimes left ghostly impressions on pages (Smyth, 2021; McLeod, 2000), to the recycled leaves from other books that one sometimes finds in old bindings (Schofield, 2021; Ryan, 2020; Smyth, 2018, pp. 137–174). Like anamorphic paintings, all of these material phenomena create situations where one artifact peeks out from between the cracks in another, so to speak.

Portal’s paratexts also complicate the idea of digital ephemerality, on the one hand pointing back to the obsolescence (or mortality) of DOS-era interfaces and file systems, while on the other challenging fan communities and video game scholars alike to recover texts and images supposedly lost or forgotten. ¹⁸ In a coincidence of metaphors, Neal Stephenson evokes the same early modern European artistic tradition as Boluk and LeMieux when, in his polemic on operating systems titled In the Beginning… Was the Command Line, he describes the white-on-black Linux command prompt as “a computational memento mori” (1999, p. 119). Today, the Aperture Science website serves as a double reminder of technological mortality, first as an echo of the bygone era of DOS computing, and second as a casualty of the more recent discontinuation of Flash (Fiadotau, 2020). As I have mentioned (see Notes 2 and 14), the ApertureScience.com website—at least in its earlier, interactive versions—is essentially a text-based Flash game in itself, and with the right software it remains playable as an SWF file downloadable from the Internet Archive. Flash may seem as dead as the owner of the skull in Figure 3, yet the “ApertureScience17.swf” files that comprised the website are still accessible thanks to the Internet Archive, and their source code can still answer questions about how the website worked when it was online. Recovering evidence from code in this manner, in digital paratexts that hover on the edge of ephemerality, highlights both the challenges and possibilities for studying video games as historical artifacts with complex materialities.

Conclusion

Just as a piece of software is a text composed of many interlocking texts, in the form of files, drivers, updates, libraries, APIs (Application Programming Interfaces), and such, so can a given video game be several objects at once—though, like Holbein’s anamorphic painting, it may require a shift in perspectives to see them. There is no question that one can play Portal without having visited ApertureScience.com; vice versa, one can conceivably experience the fictional website without ever playing the game. Likewise, discovering the Easter eggs leading to “notes.exe” would depend on a combination of playing style, chance, delight in complexity, and diligence (or willingness to look up the ApertureScience.com entry on one of the Portal wikis). By the same logic, the Flash file that makes any given version of ApertureScience.com work is computationally distinct from the files that make Portal playable as a game. One could extend that logic to the hidden SSTV images, too, given that images occupy distinct parts of the bitstreams within sound files which themselves are not strictly necessary to gameplay. Even at the level of the bitstream, and even within files that are stored within the subfolders of one’s local installation of Portal, the strict internal/external distinction fails to police phenomena into neat categories like epitext and peritext.

Nor should we seek to police those categories. In a recent article that revisits her ideas on video game paratexts, Consalvo closes with a call to focus less on “‘central’ texts and more … [on] the relatedness, interconnectedness, and contingent nature of many kinds of popular culture texts” (2017, p. 6). My focus here on ApertureScience.com departs somewhat from the spirit of Consalvo’s call—I am interested in the website precisely because of its canonicity and authorship—but the key word text in her call points to the relevance of textual scholarship for video game studies. For textual scholars, a text is not merely a sequence of alphanumeric characters, like the ASCII sterility of the command-line interface of ApertureScience.com. Rather, text denotes any form of materials that can be studied analytically and interpretively using the (still-evolving) tools of textual scholarship, from stone inscriptions, to Tom Stoppard plays, to films in different versions like Brazil (Mathews, 1987), to born-digital graphic comics like Portal 2: Lab Rat, to Jonathan Coulson’s songs, to anamorphic paintings like The Ambassadors, and to larger works of transmedia storytelling like Portal. ¹⁹ The text is that which we can study—in all its relatedness, interconnectedness, and contingency, as Consalvo says—and the paratext is that which hovers at its edges, leading into (and out of) what lies beyond.

What Portal calls us to consider, then, is not just the reconfigurations of relationships among works and texts—for example, fanfiction or mods supplementing or supplanting originals—but also the reconfigurations that happen within games as complex digital artifacts. There is value in the inevitable failure of a category like peritext when applied to video games: the mismatch between category and phenomenon forces us to pay attention to the distinct shapes of each. From the perspective of a field like textual studies, the inadequacy of the epitext/peritext distinction for video games thus creates a welcome challenge, not a problem. The most challenging thresholds Portal invites us to cross are found not only within its test chambers, but also within files, formats, and bitstreams—and perhaps within our own disciplinary domains, too, where every exit is an entrance somewhere else.