25

ZORK (1980): TEXT IMPS VERSUS GRAPHICS GRUES

It is only fitting that Infocom's Zork should be the final game discussed in this book. After all, it is the only game we've covered that relies totally on text for both input and output—it's the most book-like of all the games we've covered! Although modern critics might take its lack of graphics and sound as proof of its obsolescence, anyone with an appreciation for human language should consider it miraculously advanced. Zork represents something of a paradox: it's at the same time obsolete and cutting edge. On one hand, computers and consoles have made huge gains in audiovisual technology since text adventures like Zork were dominating best-seller lists. On the other hand, developers have made surprisingly little progress in the areas of natural language processing. It's easier for a game to let us realistically blow a man's brains out than engage him in conversation! Where Zork and its heirs triumph over modern games is precisely in presenting the illusion of consciousness. Playing Zork isn't about moving a disembodied shotgun through a 3D battleground in search of demons to perforate with buckshot (see Chapter 5, “Doom (1993): The First-Person Shooter Takes Control”). Instead, Zork is like reading a novel as it's being written, or starring in a play without knowing one's lines.

The goal is to improvise one's way through a labyrinth, intuiting the proper responses to a series of conundrums. It's also about getting to know the narrator, whose witty repertoire and scathing one-liners come most often at our own expense. Zork is funny, Zork is witty, and—above all—Zork is human. Where the game fails is precisely where we glimpse the machine posing as a person; the telltale blush response that proves our seductive companion is an automaton.

It's quite likely that no computer game in history has inspired as much prose as Zork, even if we omit the billions of commands inputted by legions of hunt-and-peckers. Zork and other text-based adventure games have long been the subjects of academics writing about games, such as Brenda Laurel, Janet Murray, and Espen Aarseth. The games were also widely discussed in popular media, where journalists were fascinated by their artificial intelligence and imaginative scenarios. Although these games lacked audiovisuals, they were still widely played and admired by pretty much anyone who played them.

No doubt, many of those early visitors to the Great Underground Empire felt like they were experiencing the future of the novel. Developers dreamed of a day when interactive fiction (IF) “novels” would stand proudly alongside Thomas Pynchon and Norman Mailer on the shelves of Borders and Barnes & Noble. Sadly, it hasn't happened—but why not?

Most gamers would respond simply that these games lost their appeal as technology progressed. Modern gamers demand sophisticated audiovisuals; text alone just won't cut it. However, although television and movies have been around for quite some time, people still buy and read books. A fan of novels might respond that the pleasure she gets from reading a good novel is simply different than watching a film; it's not necessarily a question of which one's better in some universal sense. There doesn't seem to be any practical reason why all games should have advanced audiovisuals; conceivably there could still be an audience for games consisting of well-written text. Nick Montfort, author of Twisty Little Passages (MIT Press, 2003), argues that interactive fiction will remain an “essential part of digital media and literature even if no one manages to sell it,” likening the genre to poetry.¹ Indeed, there are thriving niche communities of gamers who still enjoy playing (and making) interactive fiction, but this activity has sadly remained very much on the margins of the industry.² We'll now table this discussion and proceed to the story of Zork.

Zork began life in much the same way as many of the early computer games; that is, as a fruitful, informal collaboration by starry-eyed college students.³ Indeed, it's easy (if, perhaps, a bit misleading) to compare the development of Zork with that of another classic computer game, Spacewar! (see bonus chapter, “Spacewar! (1962): The Best Waste of Time in the History of the Universe”).

Zork was developed by four members of the MIT's Dynamic Modeling Group, whereas Spacewar! was developed by members of MIT's Tech Model Railroad Club. Both teams were excited about the possibility of computer games, and both were fueled by the adrenaline rush of hacking. However, the authors of Zork had a much different vision of the future of computer games than the hackers responsible for Spacewar!. For, although Spacewar! paved the way for graphical “twitch” games, Zork was a game for folks who preferred prose to pyrotechnics.

The authors of the mainframe Zork, Tim Anderson, Marc Blank, Bruce Daniels, and Dave Lebling, began writing the program in 1977 for the DEC PDP-10, the same computer used by Will Crowther and Don Woods to create Colossal Cave Adventure (Spacewar! was programmed on the earlier PDP-1). The PDP-10 was a mainframe computer that was much more powerful than any home computer of the time, but certainly too large and expensive for the consumer market.

Colossal Cave Adventure is a groundbreaking game that pioneered the adventure game. The first version was designed by Will Crowther and later substantially expanded by Don Woods. It is the progenitor of Zork and all later games in the genre. Screenshot from a PC conversion shown.

The few home computers that existed were so woefully under-powered compared to mainframes like the PDP-10 that most of the early game developers had little interest in trying to restrict or sell their software; if you had access to one of these behemoths, chances are you could easily acquire games like Colossal Cave Adventure for free using the ARPANET,⁴ the progenitor to the Internet.

The “imps,”⁵ as they would later style themselves, were enchanted with Colossal Cave Adventure, also known as Adventure or simply Advent, whose first appearance was in 1976. Colossal Cave is certainly a groundbreaking game, both in the figurative and literal sense—the original author, Crowther, and his wife, were dedicated cavers and he based much of the game on an actual cave system in Kentucky. Although many critics tend to overlook this caving connection, it's important if we want to fully understand the appeal of games like Colossal Cave and Zork. It's not simply a coincidence that both games are focused on the thrilling exploration one enjoys as a caver or urban explorer. These games are less “interactive novels” than “interactive maps” (or “interactive worlds” to use language popularized by Cyan of Myst fame). Another intriguing coincidence is that the first jigsaw puzzle ever sold was of a map.⁶ It seems that maps and puzzles have been associated from very early times!

Although exploration games can be rendered with graphics instead of text, this eliminates much of the freedom (or at least the illusion of such) allowed by text—a point we'll return to later. As for Crowther, his purported intention for creating the game was chiefly as a way to share some of his enjoyment of caving and the role-playing game Dungeons & Dragons (D&D) with his two daughters. It's easy for critics, perhaps avid D&D players themselves, to get so fixated on this fantasy role-playing connection that they overlook the influence of caving.

Colossal Cave established many of the conventions and principles upon which almost all subsequent text adventures are based, such as the familiar structure of rooms and objects, inventory, point system, and input structure (“OPEN DOOR,” “GO NORTH,” and so on). The game also introduced several elements inspired by J. R. R. Tolkien's famous works, including dwarves and magic (Tolkien was always drawing maps himself). To get through the game, players frequently sketched out their own graphical maps of the areas they explored. Colossal Cave also famously introduced the “non-Euclidean maze,” or a series of identically described rooms.

The only way players can navigate these mazes (besides cheating) is to drop breadcrumbs, or objects whose placement in a room will alter its description (thus allowing the player to retrace steps). The game also requires players not merely to collect treasures, but to deposit them in the proper location to earn points. Although Colossal Cave was certainly a breakthrough, it didn't take long for hackers to master it. Some hackers went a step beyond; they had sighted a new vista and wanted to explore its possibilities to the fullest.

Much like Crowther and Woods, the imps were initially inspired more by a desire to test their hacker skills than a singular desire for wealth. Indeed, in a famous 1979 article published in the scientific journal IEEE Computer, the authors promised to send anyone a copy of the game who sent them a magnetic tape and return postage.

This article, written by Dave Lebling, Marc Blank, and Tim Anderson, describes the game as a “computerized fantasy simulation,” and uses terminology familiar to anyone who remembers D&D: “In this type of game, the player interacts conversationally with an omniscient ‘Master of the Dungeon,’ who rules on each proposed action and relates the consequences.” However, like Colossal Cave, Zork is primarily a game about exploration, involving such activities as breaking into a supposedly abandoned house, rappelling down a steep cavern, and even floating across a river in an inflatable raft (watch your sword!). Along the way, the player is continuously confronted with puzzles and even a few fights (such as a troll to be dispatched with the sword). Most famously, though, the player must at all times be wary of the grue—a mysterious beast which lurks in total darkness, always hungry for adventurers.

On the surface, Zork appears to have much in common with its progenitor, Colossal Cave, and IF scholar Dennis Jerz has gone so far as to say that “whereas Adventure began as a simulation of a real cave, Zork began as a simulation of Adventure.”⁷

Zork added several key innovations, including a much more sophisticated parser capable of handling commands like “KILL TROLL WITH SWORD” and “PUT COFFIN, SCEPTRE, AND GOLD INTO CASE,” whereas Colossal Cave (and for several years, Infocom's commercial competition) was limited to commands of one or two words (“GET BOTTLE,” “PLUGH”). Zork also offered a more sophisticated antagonist, the famous thief, who roams about the world independently of the character and eventually plays an important role in solving the game. Montfort described the thief as a “real character with the functions of a character as seen in literature, not the mere anthropomorphic obstacle that was seen in Adventure.” Zork II also introduces a coherent plot to add some narrative coherence to the player's treasure hunting. Overall, though, the game was praised for its humor and excellent writing.

The mainframe Zork was not broken into a trilogy, but rather existed as a single massive game. After the imps founded Infocom and decided to commercially release the game for personal computers, they were faced with stiff memory limitations (and a wide variety of incompatible platforms). To get around the problem, they broke the game up into three parts, though not without some modifications and additions. It's also worthwhile to mention the brilliant design strategy they followed.

Rather than port the code to so many different platforms, Joel Berez and Marc Blank created a virtual platform called the “Z-Machine,” which was programmed using a LISP-like language called ZIL. Afterwards, all that was required to port the entire library to a new platform was to write (or have written) a “Z-Machine Interpreter,” or ZIP. Scott Cutler took on the task of creating the first commercial ZIP, which was written for Tandy's TRS-80.⁸ Indeed, one of Infocom's key assets as a text adventure publisher was the ease with which they could offer their games on a tremendous number of platforms; graphical games were much harder to port.

Early on, Infocom's popular software came in packaging of all different shapes and sizes. Some of the rarer types, like Starcross, which came in a plastic flying saucer, and Suspended, which came in a plastic mask, are particularly valuable collector's items today—sometimes selling for hundreds of dollars.

Indeed, for many of the more obscure platforms, Infocom's lineup was the best (if not the only) games available. It seemed like no matter what type of computer you had, you could always buy a copy of Zork. This fact no doubt offered them considerable leverage in the terrifically diversified home computer market of the early 1980s, when consumers could pick from many different machines, each with its own advantages and disadvantages (cost, speed, memory, ease of use, expandability, software library, and so on). However, Lebling notes that there were some “negative impacts, especially as the newer machines began to have more memory and better graphics. We had to write to the lowest common denominator, or spend time on each game fitting it to the different platforms.” As the latter approach was cost-prohibitive, Infocom's games were identical on every platform. It's notable that later adventure game publishers followed Infocom's example, including Sierra (AGI) and LucasArts (SCUMM; see Chapter 11, “King's Quest: Quest for the Crown (1984): Perilous Puzzles, Thorny Thrones”). In each case, the idea was to separate the creative assets (script, graphics, and so on) from the machine-level programming.

Infocom's games were known for their “feelies,” or extra items inside the box. Some, like letters and maps, provided vital clues or information for the game, while others—like Wishbringer's glow-in-the-dark magic stone, The Hitchhiker's Guide to the Galaxy's peril-sensitive sunglasses, and Sherlock's key fob—were just for fun.

Infocom also lured gamers with innovative packaging and “feelies,” or small items included with the disks and manuals. Usually these items were added to complement the game's theme, such as the peril-sensitive sunglasses included with The Hitchhiker's Guide to the Galaxy (1984). Other feelies served to curtail illegal distribution. For instance, without the “QX-17-T Assignment Completion Form,” players would not be able to input the coordinates needed to access the space station in Stationfall. Sadly, the ambiance achieved by the clever packaging and feelies are incapable of being “emulated,” and players really wanting to get the full experience would do well to own an original boxed copy.

The three commercially released Zork games are Zork I: The Underground Empire (1980), Zork II: The Wizard of Frobozz (1981), and Zork III: The Dungeon Master (1982). Although these games are based on parts of the massive mainframe version, the imps worked hard to make each game more coherent, such as the plot structure of The Wizard of Frobozz. Now, players had to do more than just find all the treasures—they had to find a way to bring the story to its natural resolution.

An early scene from Zork III: The Dungeon Master on the Apple II.

No doubt to the angst of many parents worried about the “Satanism” of so much fantasy role-playing, this story culminates in giving ten treasures to a demon, who takes them as payment for performing one critical task. The game is also noted for two infamously difficult puzzles called the “Bank of Zork Vault” and the “Oddly-Angled Room.” The final game in the trilogy, The Dungeon Master, takes leave of much of the humor and opts for a more solemn and gloomy tone; one reviewer calls it “brooding.” Instead of merely hunting for loot, the player must find items that allow him (or her) to take on the role of Dungeon Master.

In 1983, Infocom released Enchanter, the first of another trilogy of games set in the Zork universe. These games were much more focused on magic and spellcasting than Zork, but retained much of the humor and excellent writing. Sorcerer (1984) and Spellbreaker (1985) round out the series. Each entry in the series is increasingly difficult, to the point that some critics complained that Spellbreaker was a contrived effort to boost sales of Infocom's InvisiClues hint books.

Infocom's Zork and Enchanter trilogies were fabulous successes, and the company followed up with several other classics. To make a long story short, Infocom's business was booming, and its superior interactive fiction titles earned them enough “zork-mids” to build their own empire, not to mention throw incredible promotional parties, including the legendary “murder mystery party” thrown at the 1985 Consumer Electronics Show (CES) for Suspect (1984). Infocom even hired a troupe of actors and let participants indulge in some “live-action role-playing” to solve the murder. Infocom was at their zenith.

Unfortunately, the company would soon flounder. From the beginning, Infocom was not intended solely to develop and publish games (one thinks of countless rock and pop bands dreaming of producing “serious” music with the London Philharmonic Orchestra). Although their text adventure games had sold amazingly well, Infocom wasn't satisfied—they were convinced they were destined for bigger and better things. The albatross flopping around Infocom's corridors was a relational database called Cornerstone (1985). Cornerstone sounded like a brilliant idea—everyone knew that database software had revolutionary potential for business, but the current offerings were far too complex for the average user. Infocom saw an opportunity, and felt that the same virtual machine strategy they used for Zork would work well for Cornerstone. But it didn't.

Box back from the Coleco Adam's 2010: The Text Adventure Game from 1984. Although still considered a text adventure, the game was played entirely with the Adam computer's arrow keys and SmartKeys (function keys), which eliminated the need for a traditional parser. Although the only real graphics and music were at the title screen, players would hear their own breathing (which would slow and quicken depending upon the situation), as if they were playing the game in a spacesuit.

Micro-Fun's Death in the Caribbean from 1985, Apple II version shown, was advertised for its visuals rather than its parser, which was severely limited in comparison to Infocom's text-based games.

By the time copies of Cornerstone began lining up on store shelves, the IBM PC was the overwhelmingly dominant platform for business; portability was no longer an issue. Furthermore, the virtual machine setup reduced its speed, and it lacked several of the advanced features that made its rival database programs worth learning in the first place. The program was not a success, and several critics remarked that its name was apt—it sat on store shelves like a stone. Infocom had foolishly invested so heavily in the product, however, that they were unable to recover, and in 1986 the company was acquired by Activision.

What happens next is a rather dismal story indeed. Activision seemed uninterested in publishing text games, preferring instead to exploit the popularity of games like Zork in graphical adventure games, starting with Beyond Zork, in 1987, by Brian Moriarty (Wishbringer, Trinity, and others).

Screenshot from the Commodore 64 version of Datasoft's The Dallas Quest (1984). Despite having only superficial similarities to the popular television series, this text and graphics adventure was part of a unique group of surprisingly good licensed games in the format from the 1980s, which includes the text-only Rambo: First Blood Part II (Mindscape, 1985) and Star Trek: The Kobayashi Alternative (Simon & Schuster Interactive, 1985), among others.

Screenshot from the Apple Ilgs version of Activision's quirky Tass Times in Tone Town (1986), a text and graphics adventure that can be considered a transitional type of product—the game could be played using either traditional text input or by clicking on the various icons, either in combination with or exclusive of each other.

Beyond Zork offered players a crude automap and several random and RPG elements to theoretically enhance the game's replayability. Replayability is always an issue with most adventure games: once the player figures out the puzzles and solves the game, there is little reason to play it through again—though a few years may be sufficient time to forget enough of the details to make it fun again.

Steve Meretzky (Planetfall, A Mind Forever Voyaging, and others) got in on the act with Zork Zero, another graphically enhanced game published in 1988. Zork Zero is a prequel to the trilogy, and offers several nice features like in-game hints, menus, and an interactive map.

Screenshot from the PC version of Infocom's Arthur: The Quest for Excalibur (1989), one of a handful of later releases from the company that would go the text-and-graphics route.

The last game to be published under the Infocom label was Return to Zork, a 1993 game released for Apple Macintosh and PC. Developed by Activision, Return to Zork is quite a different animal than the previous Zork games, even the graphically enhanced games described previously. Return to Zork will no doubt remind most gamers of the far more popular Myst (see chapter 12), which was released a few months afterwards. The parser is gone, replaced by a purely graphical interface that is surprisingly complex and multifaceted.

Infocom's games were often more varied in theme than other companies. Take for instance, Trinity, pictured to the left, which blends history and fiction as part of a prose poem regarding the destructive power of the atomic bomb and the nature of war in the modern era; and Plundered Hearts, pictured to the right, which is the equivalent of a romance novel and casts the player as a young female in the late seventeenth century.

The Lost Treasures of Infocom, Commodore Amiga version box front and back shown, was one of several collections of Infocom's classic games released over the years.

The game also offers live action sequences, including performances by a few recognizable but second-tier actors. Contemporary reviewers seemed to mostly enjoy the game, though Zork aficionados were (and are) divided over whether to include the game as part of the Zork canon. Very few of the original characters show up in the game, and there will always be the issue of whether any graphical adventure game could truly compare to the great text-based classics. Jay Kee, who reviewed the game for Compute! magazine, wrote that “people accustomed to the speed and flexibility of a text-only parser are going to feel handcuffed.”⁹

Activision released two more Zork-themed graphical adventures: Zork Nemesis: The Forbidden Lands (1996) and Zork: Grand Inquisitor (1997), quietly dropping the name “Infocom.” Nemesis offers a much simplified graphical interface and a much darker atmosphere than previous games. Like Return to Zork, Nemesis was loaded with live action sequences—to the point that the game shipped on three CD-ROMs.

Most reviewers noticed the game's intense gore, including a puzzle requiring the player to behead a corpse with a guillotine. However, Grand Inquisitor did try to bring back much of the humor missing in Zork Nemesis, and seemed to pay more homage to the series than the previous two games. Perhaps more significantly, Activision released Zork: The Undiscovered Underground for free, a text adventure by Marc Blank and Michael Berlyn. The Undiscovered Underground no doubt eased some of the bitterness that dyed-in-the-wool Zork fans felt toward Activision, who some viewed as merely exploiting the franchise to turn a quick buck.

Unfortunately, even a new text adventure was not enough to save Zork; Grand Inquisitor did not sell as many copies as Activision hoped. To date, there have been no more official Zork titles, though there have been several anthologies.

It is important to consider that even without Infocom's failure with Cornerstone, the desire for graphics would have likely pushed the pure text adventure to the side anyway. Many other companies had tried text and graphics adventures before, like On-Line Systems/Sierra with Mystery House (1980; Apple II and others), or text adventure pioneer Scott Adams's¹⁰ Adventure International with Return to Pirate's Isle (1983; Texas Instruments TI-99/4a), though the parsers in such games rarely went beyond simplistic, two-word input. Even when companies like Trillium/Telarium with Amazon (1984; Apple II, Commodore 64, and others) from author Michael Crichton, or Magnetic Scrolls with The Pawn (1986; Atari ST, Commodore Amiga, and others), had strong combinations of graphics and parser technology, the push always seemed to be to automate the process away from natural language interaction to point-and-click simplicity. Even Legend Entertainment, founded in 1989 by Bob Bates and Mike Verdu after the official end of Infocom, eventually changed from producing text and graphics adventures with natural language input to purely mouse- and menu-driven graphics adventures,¹¹ until finally settling on action games before their shutdown in 2004. The market had spoken.¹²

Scott Adams and his Adventure International Company, started in 1978, preceded Infocom's philosophy of releasing games—particularly text adventures—on as many platforms as possible. The Scott Adams games had modest parsers and complexity, which—unlike Infocom's—allowed them to work on systems with limited processing power and memory, and ship on cartridge, tape, or disk. Despite moving to text and graphics adventures and having popular licenses like Marvel Comics, Adventure International went bankrupt in 1985.

Nevertheless, to say that Zork is not an influential adventure game is like saying the Iliad is not an influential poem. At some point, the question is not so much one of “influences” but rather of laying foundations. Although the game's mechanics have no doubt been surpassed by later parsers and arguably, interfaces, no one can deny the incalculable influence Zork has extended across a broad spectrum of games and genres. Could we have Myst without Zork? What about Doom? All of these games borrow and pay homage, whether directly or indirectly, acknowledged or not, from the type of gameplay found in Zork.

Box back for the PC version of text and graphics adventure, Spellcasting 101: Sorcerers Get All the Girls, from Steve Meretzky and Legend Entertainment Company. Note the menu-driven interface option in the screenshot on the upper right of the box.

The player is still exploring spaces, uncovering possibilities, and overcoming obstacles. The only crucial differences are the ways these activities are represented on the screen, and the way they are selected by the player. In the first case, Myst, rooms and actions are described via graphics rather than text. In the second, Doom, players use arrow keys or mouse clicks rather than typing commands in the form of words and sentences. For example, if the player “goes north” in Doom by pressing the up arrow key, players in Zork type “GO NORTH” or simply “N.” To say that the former method is objectively superior or more “immersive” than the latter seems foolhardy at best.

What Zork seemed to contribute more than anything was the idea that the computer could simulate a rich virtual environment—much, much larger and nuanced than the playing fields seen in games like Spacewar! or Pac-Man (Chapter 13, “Pac-Man (1980): Japanese Gumption, American Consumption”). Furthermore, the game demonstrated the literary potential of the computer. Thousands upon thousands of gamers have been charmed by the wit and elegance of Zork's many descriptions. Perhaps more than anything, though, these games offered players the illusion of total freedom. Instead of merely selecting a few set commands from a menu, Zork encouraged players to imagine infinite possibilities.

For most players, a great deal of the fun was simply experimenting with strange commands to see whether the developers had anticipated them. For example, typing “HELLO” results in, “Nice weather we're having lately.” or “Good day.” Type “JUMP”, and you're told, “Very good. Now you can go to the second grade.” On the other hand, typing “HELP” results in “I don't know the word ‘help,’” a response that seems to have unintended significance. You can try out the results of curse words and more sinister actions yourself.

There have been many claims made over the years (particularly by disgruntled fans of interactive fiction) that their games are simply more intellectually challenging, and that the reason so many modern gamers don't like them is that they simply aren't intelligent or refined enough to appreciate them. On the other hand, we might question whether a textual description really requires more imagination than an image. Perhaps a similar sort of thing is going on in our heads whether we see the word “mailbox” or see an image of one on the screen. To make sense of either, we have to have some sort of familiarity with the concept of mailboxes, and imagine the possible reasons why the mailbox is there and what role it could play in the game. Either way, we have to use our brains to make sense of it.

Synapse's short-lived “Electronic Novel” series came packaged as hardcover books in slipcases. The parser for these products were considered comparable in most cases to the best that Infocom could offer at the time and even featured some unique capabilities. Unfortunately for Synapse, upon the games’ release in 1985, the market was shifting away from pure text adventures.

What, then, is the true advantage of a text adventure over a graphical one? The answer to this question seems to be the perceived freedom and intelligence of the parser. It's nice to be able to interact with a game in such a thoroughly compelling manner, and it's here that we may see the future of Zork, or the future of any text-based interactive narrative. The key is an increasingly sophisticated parser, with enough artificial intelligence to make convincing responses to anything the player might type; it would be as though there was an actual person or “dungeon master” on the other side of the screen. This technology could also be useful in graphical games, where it could complement the icon-based and context-driven interfaces of modern graphical adventure games. An intriguing and important example is Procedural Arts’ Facade (2005), an experimental videogame that had players using textual input to interact with an estranged couple. Though still clearly in its early stages, these technologies could easily lead to a true breakthrough for gaming, greatly enriching the possibilities for conversations with and between computer-controlled characters. Imagine, for instance, a role-playing game or first-person shooter in which players could talk to characters about practically anything—rather than simply clicking on a small set of options in a dialog menu. The dramatic possibilities of such a feat are practically endless.

There have actually been three major The Hitchhiker's Guide to the Galaxy text-adventure adaptations of Douglas Adams legendary radio, book, and television series (among others). The first, released in the early 1980s by Supersoft for the Commodore PET and 64, was quickly pulled from the market over publishing rights. The first official translation came from Infocom in 1984 with help from Adams and proved to be one of the company's best-sellers. In 2004, the game was revived by the BBC for free play in a Web browser, shown here in its second edition. The BBC adaptation added illustrations and a visual interface to the previously text-only game.

Although such technology is far beyond what we currently have available, consider how far graphics technology has come since 1980. What if the same level of exponential growth had occurred in artificial intelligence and natural language processing? “The things that interest me,” writes Montfort, “are advancing the state of the art, tackling simulation and language in new ways, and doing important work within our culture.” Though we've yet to see a resurgence of text adventures in the commercial sector, there's no doubt that it's easier than ever to create them. Powerful tools like TADS¹³ and Inform¹⁴ offer would-be authors with minimal technical knowledge the chance to design their own professional-quality interactive fiction. Dennis Jerz, a scholar who frequently writes about interactive fiction, notes that these tools have contributed to the “literary feel of most of the prize-winning IF games these days.” These tools and the Internet have helped foster a small but growing community of IF developers and enthusiasts, who continue to push the boundaries of the genre, and there have been recent efforts to reintroduce it commercially. Perhaps one day we will find the latest interactive fiction titles alongside other best-sellers at our local bookstore—but we may also find parsers cropping up in the latest triple-A titles, where they could greatly enrich our conversations with computer-controlled characters.

¹Unless otherwise noted, all quotations in this chapter are from private email correspondence with the authors.

²Today, dedicated and talented IF authors continue to advance the art of interactive storytelling as part of a mostly noncommercial effort. Unfortunately, for all of the innovation in story constructs and styles, the technology has more or less remained the same as that originally found in Zork. On the plus side, the modern interactive fiction creation languages are still highly portable and appear on nearly every possible platform, including directly within Web browsers. See http://www.ifarchive.org for more on modern interactive fiction.

³To be specific, three of Zork's developers were college students at MIT (Blank, Anderson, and Daniels). The fourth, Dave Lebling, was a member of MIT's staff.

⁴The Advanced Research Projects Agency Network of the United States Department of Defense was the world's first operational packet switching network.

⁵Short for “implementers.”

⁶See Daniel McAdam's “History of Jigsaw Puzzles” at http://www.jigsaw-puzzle.org/jigsaw-puzzle-history.html.

⁷See http://jerz.setonhill.edu/if/canon/Zork.htm.

⁸These first commercial versions of Zork for the TRS-80 (late 1980) and Apple II (early 1981) were published by Personal Software (known for the 1979 Apple II “killer app,” VisiCalc, one of the first spreadsheet programs) with poorly matched cover art that featured a crazed, sword-wielding warrior standing over a cowering troll. Infocom would soon take over publishing for Zork themselves, complete with new cover art. See http://home.grandecom.net/~maher/if-book/if4.htm for more on the story and to see the original Personal Software cover art. Interestingly, while the Personal Software releases are generally considered the first true commercial versions of Zork, an early form of the Infocom company sold a crudely packaged PDP-11 version of Zork on 8” disk that is said to predate them. See http://inventory.getlamp.com/2008/11/30/pdp-11-zork-manual-save-234831/ for more information.

⁹See the September 1994 issue of COMPUTE! magazine.

¹⁰No relation to the author of Dilbert. Scott Adams began releasing often-difficult commercial text adventures with simple two-word parsers as far back as 1978.

¹¹Legend Entertainment's last game to feature a text input option was 1993’s Gateway 2: Homeworld. Early on, the company published several titles from Steve Meretzky.

¹²In recent years, a smattering of mainstream games have tried to incorporate text input into their gameplay, but this has resulted in little to no impact on future developments. For instance, 1998’s Starship Titanic adventure game from Douglas Adams's The Digital Village for Apple Macintosh and PC featured a full-sentence text parser in its conversation engine, and Konami's futuristic survival horror game Lifeline from 2003 for the Sony PlayStation 2 allowed the player to speak full-sentence commands to the in-game protagonist.

¹³Standing for “Text Adventure Development System.” See http://tinyurl.com/4l6zxk for top titles created with the language.

¹⁴See http://tinyurl.com/4ceoxf for top titles created with the language.