You are standing at the end of a road before a small brick building.
Around you is a forest. A small stream flows out of the building and
down a gully.
-- opening text in Adventure (Crowther and Woods, 1976)
In 1966, Joseph Weizenbaum, a researcher and computer programmer at The Massachusetts Institute of Technology, released an unusual program into the then rather small world of computing. Entitled Eliza, it was billed as a “simulation” of a session between a Rogerian psychotherapist (the titular character) and her patient (the user). A typical session with Eliza might look something like this:
How do you do. Please state your problem.
>> I spend too much time with computers.
Why do you mention computers ?
>> You are one.
Does it please you to believe I am one ?
You seem to be quite positive.
>> I suppose I am.
How long have you been ?
>> 31 years.
I'm not sure I understand you fully.
>> Join the club.
Please go on.
>> You don't make too much sense yourself.
We were discussing you -- not me.
If you have experienced IF before, or you have spent time with the sample transcript I provided in Chapter 1, you have probably noticed that Eliza’s mode of communication is basically identical to that of IF. Eliza is most emphatically not true IF, however. It simulates no universe, and does not truly understand the player’s input in any meaningful sense. When we actually peer under the hood, we find that Eliza is really little more than a clever parlor trick. It scans the user’s input for a few dozen key words and parrots back a set of stock responses based upon what it finds. It does not take long for the player to realize Eliza’s limitations, as responses that at first seem rather shockingly apt quickly begin to repeat, and eventually enough complete non-sequiturs appear to destroy the game’s mimesis entirely. (At this point some less charitable souls might suggest that Eliza actually provides quite a good model of interaction with many psychotherapists, but we shall avoid such controversy here.) The modern player might wonder how Eliza ever managed to cause the stir it did in the computing world when its limitations are so quickly obvious. Such a player much remember, however, that the idea of any sort of interactive computer program seemed quite revolutionary in the batch-processing world of 1966.
Eliza is of great significance in the history of IF, for it represents the first use of the general IF model of interaction as a textual dialogue between the player and the game. When later developers began to design true IF, they would borrow the interface conventions used by Eliza and the legion of so-called “chatterbots” that would follow it. Many early IF games in fact seem to see the player as issuing instructions to another who inhabits the game world, rather than the player inhabiting the world herself. Such games often end each turn by asking the player, “What do I do now?” and will respond to a pointless or non-understandable input by saying something along the lines of, “I can’t do that right now.” This mode of interaction, although rarely seen in IF today, probably stems from IF’s roots in Eliza and her successors.
The ideal of simulating a conversation with a sentient entity is still very much alive in modern IF circles. Parlor tricks are no longer acceptable, however, and managing to create believable interaction honestly is a tall order indeed. Perhaps the most successful attempt so far, and a game which could be considered a distant spiritual cousin of Eliza, is Emily Short’s Galatea, a work we will return to in a much later chapter of this essay.
Perhaps the most significant foundational work for IF arrived in 1972 in the form of a unique little game called Hunt the Wumpus. This game sets the intrepid adventurer lose in a network of caves, all rendered with textual descriptions, and tasks him with hunting down and killing a fearsome creature known as the wumpus, a monster that is “rather vague in physical detail” (Yob 247), as the game’s creator Gregory Yob himself admitted, but that is deadly if encountered directly. The caves are rendered, in classic IF fashion, as a series of discrete nodes or “rooms,” connected to one another by passageways. The standard IF convention of moving about through the use of compass directions is not in place, however. The player is rather told which room number she is currently in and to which rooms she can travel from there. There is also no parser in place. On any given turn, the player has only two choices: to move or to shoot an arrow. A typical session with the game might look something like this:
You are in room 14
Tunnels lead to 4 13 15
Shoot, Move or Quit (S-M-Q)? m
Where to? 4
You are in room 4
Tunnels lead to 3 5 14
Shoot, Move or Quit (S-M-Q)? m
Where to? 5
You are in room 5
Tunnels lead to 1 4 6
Shoot, Move or Quit (S-M-Q)? s
No. of rooms (0-5)? 4
Room #? 1
Room #? 2
Room #? 3
Room #? 4
You are in room 5
Tunnels lead to 1 4 6
Shoot, Move or Quit (S-M-Q)?
As you can see, the game’s mode of communication is decidedly simplistic, not to say cryptic. The challenge of Hunt the Wumpus arises from the fact that the player’s character does not have a prayer if she gets into a fair fight with the wumpus. She must use a more devious method to hunt down her prey. Luckily, she is equipped with a rather high-powered bow that can fire arrows through up to five adjacent rooms. Winning a game of Hunt the Wumpus involves first carefully exploring and mapping the cave system and all of its twisted interconnections, being careful to avoid the wumpus while doing so. The game will warn the player when she is in a room adjacent to the beast by saying, “I smell a wumpus!” Note the use of first person here, possibly an artifact of games like Eliza. To make the process more difficult, bottomless pits and bats are also scattered about the cave system. Stepping onto a pit means instant death, while bats pick up the player and randomly drop her at some other location in the maze. Finally, the wumpus itself occasionally wakes and moves about under certain circumstances.
Hunt the Wumpis is really just an elaborate logic puzzle. Once the player has mapped out the cave system, she must maneuver her character into position to kill the wumpus with a shot from her bow. She has limited room for error here, for she carries only five arrows with her. Further, if she attempts to fire an arrow into a non-connected room it will ricochet unpredictably and may even fly back to kill the player.
Primitive as it appears, Hunt the Wumpus is nonetheless a much more sophisticated piece of programming than Eliza, and much closer to our modern conception of IF. Whereas Eliza merely spouts back stock responses based upon the player’s last input, Hunt the Wumpus actually simulates a universe in the computer’s memory, albeit a very simple one containing only a few objects. The player’s actions directly affect the state of that universe. One could even say that a winning game of Hunt the Wumpus has a narrative arc of sorts, beginning with exploration and mapping, leading to planning and positioning, and finally climaxing with the kill shot that ends the game in victory. The building blocks of IF are all here. All that is needed now is a true parser.
Hunt the Wumpus has had enormous influence on IF over the years, for both good and ill. It is probably not a coincidence that the first work of true IF, Adventure, was also set largely within a network of caves, or that so many of the games that followed Adventure chose similar settings. Hunt the Wumpus is essentially an exercise in mapping, and many IF games have taken this to heart by making their geography a huge part of their challenge. For years, almost every IF game felt obligated to include a maze section. These sections – the “twisty little passages” famously described in Infocom’s Zork I – were made up of many non-descript rooms all exactly alike in their descriptions, and connected to one another in hopelessly jumbled ways. In other words, going north within a classic IF maze and then going south rarely returns the player to her starting position. Everyone who has been playing IF for any length of time knows how to solve these mazes, and doing so does not require intellectual brilliance on the part of the player, merely patience, care, and a good supply of graph paper for mapping. For that reason, mazes – at least ones that are not executed with some particularly clever twist or other – are considered annoying busy-work by most and are almost universally pointed to as examples of what not to do when designing a modern work of IF.
Nostalgia being what it is, however, Hunt the Wumpus is generally held in much higher esteem today than those games that imitate its more annoying aspects. Occasional revivals of the game even appear on the IF scene, notably Andrew Plotkin’s Hunter, in Darkness and Muffy St. Bernard’s Wumpus 2000, both of which took the bare stub of a story upon which the original is based and expanded it into full-fledged IF games. Many years earlier, two other hackers named Will Crowther and Don Woods probably spent considerable time with Yob’s game before combining its world modeling with some of Eliza’s interface conventions and mixing the whole thing together with the world’s first true IF parser to create Adventure.
During the early seventies, Will Crowther had three passions that in combination led to the creation of Adventure. He was a computer programmer who in the course of his employment by BBN in Boston would create some of the foundational software that led to our modern Internet; he was an avid spelunker who with his wife Pat frequently traveled to Mammoth Cave in Kentucky to explore and map under the auspices of the Cave Research Foundation; and he enjoyed playing with friends a new tabletop game called Dungeons and Dragons, during which he took on the persona of a character named Willie the Thief for adventures in a Tolkienesque fantasy world (Hafner 206). Like many other hackers, Crowther considered the possibility of creating a computer game with some of the characteristics of Dungeons and Dragons for quite some time, but he was seemingly spurred into action only by an unfortunate event, namely the failure of his marriage. The newly divorced Crowther now began working on his ideas in earnest. He hoped to create an interactive textual simulation of caving, with some magical elements ala Dungeons and Dragons thrown in for spice, to give to his two young daughters whom he missed greatly and could see only occasionally. In his recent book-length study of IF, Nick Montfort theorizes that Sandy and Laura Crowther, aged five and seven, probably became the first players of a true IF game when they interacted with the first version of Adventure via an ASR33 Teletype connected to a DEC PDP-10 minicomputer in 1975 (Montfort 85). That original version unfortunately seems to have been lost, as much searching by myself and others has failed to turn up a copy. However, by all accounts it was quite austere by the standards of what would follow. Its geography was based largely upon that of the real Mammoth Cave that Crowther knew so well, and it was laden with caver jargon. Crowther apparently thought of his creation as little more than a trifle to amuse his daughters, and thus it was never polished into anything like a completed state. After Crowther and his daughters had had their fun with it, it was left to languish in various software repositories on the ARPANET, the predecessor to our modern Internet, to be toyed with the by the occasional curious hacker.
In 1976, one such hacker on the other side of the country, Stanford’s Don Woods, discovered Crowther’s game and was immediately smitten by its potential:
“Adventure made users feel like they were interacting more with the computer,” said Woods. “It seemed to be responding more to what you typed, rather than just making its own moves, like a silent opponent. I think this attracted a lot of players who might otherwise have been turned off the idea of playing ‘against’ a computer. This was playing ‘with’ a computer” (Hafner 207).
It probably did not hurt that Stanford’s computer science department was at the time in the midst of a full-blown Tolkien revival phase, to the point that the main printer there had been modified to handle three different types of Elvish fonts (Levy 140). Woods decided he would like to realize the potential he saw in Crowther’s work by squashing the bugs, expanding it, and generally making it into a complete, playable game. His first obstacle in doing so was the fact that he did not know Crowther and had no obvious way of contacting him, for, while Crowther had left his name on the program, it had apparently never occurred to him that anyone might want to actually get in touch with him about it. Woods’ eventual solution to the problem was to send an email to user ID “crowther” at every single domain on the then rather small ARPANET (Adams). Eventually he received a response from BBN in Boston giving him the permission he sought, and Woods set to work. The game he ended up creating, which was spreading like wildfire around the ARPANET by early 1977, looks like this:
YOU ARE STANDING AT THE END OF A ROAD BEFORE A SMALL BRICK BUILDING.
AROUND YOU IS A FOREST. A SMALL STREAM FLOWS OUT OF THE BUILDING AND
DOWN A GULLY.
YOU ARE INSIDE A BUILDING, A WELL HOUSE FOR A LARGE SPRING.
THERE ARE SOME KEYS ON THE GROUND HERE.
THERE IS A SHINY BRASS LAMP NEARBY.
THERE IS FOOD HERE.
THERE IS A BOTTLE OF WATER HERE.
YOU'RE AT END OF ROAD AGAIN.
YOU ARE IN A VALLEY IN THE FOREST BESIDE A STREAM TUMBLING ALONG A
AT YOUR FEET ALL THE WATER OF THE STREAM SPLASHES INTO A 2-INCH SLIT
IN THE ROCK. DOWNSTREAM THE STREAMBED IS BARE ROCK.
YOU ARE IN A 20-FOOT DEPRESSION FLOORED WITH BARE DIRT. SET INTO THE
DIRT IS A STRONG STEEL GRATE MOUNTED IN CONCRETE. A DRY STREAMBED
LEADS INTO THE DEPRESSION.
THE GRATE IS LOCKED.
THE GRATE IS NOW UNLOCKED.
IT WAS ALREADY UNLOCKED.
YOU ARE IN A SMALL CHAMBER BENEATH A 3X3 STEEL GRATE TO THE SURFACE.
A LOW CRAWL OVER COBBLES LEADS INWARD TO THE WEST.
THE GRATE IS OPEN.
Woods’ more Tolkienesque additions to Adventure, which include a troll, elves, and even a volcano standing in for Mount Doom, begin to appear as the player gets deeper into the game. The ultimate goal, and what semblance of a plot the game possesses, is to collect a number of treasures from the cave and return them safely to the surface. Needless to say, this is easier said than done.
Woods’ original release of the game was written in Fortran, which had no capacity for outputting lower-case letters. The parser was also extremely primitive, being capable of understanding only two words in simple verb-noun combinations. Nevertheless, all of the core elements of IF are present. The player is cast in the role of a character in a simulated world which is described in text, albeit ungrammatical upper-case text, and a true parser is used for input, albeit one of a simplistic and often infuriating design.
Attempting to play Adventure in its original form as a modern IF player is something of an exercise in masochism. The game is a veritable catalog of modern IF game design sins. As Carl Muckenhoupt writes in his review, “It has a verb-noun parser, minimal detail, two big annoying mazes, magic words, nonsense puzzles, and occasional death without warning.” Nevertheless, he concludes, “Download it anyway. You cannot consider yourself a true adventurer until you've played this game. (Muckenhaupt).” I am not sure I agree with Muckenhaupt. I think a short while spent mucking about with the game, perhaps with a walkthrough in hand to enable one to see the good bits quickly, is more than sufficient for any but the most dedicated IF historian. I think the simple fact that a game that was once fascinating now seems more annoying than anything else points to the growth that IF has undergone since 1977. In his classic essay “The Craft of Adventure,” Graham Nelson describes IF as “a crossword at war with a narrative.” Every IF game comes to terms with this uneasy balance in its own way, and every IF designer must decide how much emphasis to place on the gamelike qualities of her work, generally meaning puzzles, versus its literary qualities. There is no one perfect balance. This player, to choose one example, has certainly enjoyed games that ranged all along this particular continuum. Nevertheless, when a game fails to give at least a modicum of attention to literary sensibilities, and emphasizes puzzles that are difficult in an annoying or unfair way, it tends to win few friends in the modern IF community.
Neither Crowther nor Woods are truly at fault for their game’s failings. They were after all paving the way, with no corpus of game design theory to fall back on, and they were writing for an audience with very different expectations than the average modern IF player. Steven Levy compares playing Adventure to assembly language programming: “In a sense, Adventure was a metaphor for computer programming itself – the deep recesses you explored in the Adventure world were akin to the basic, most obscure levels of the machine that you’d be traveling in when you hacked in assembly code” (Levy 141). Levy means this as an endorsement of Adventure’s intellectual challenge, yet a new author who advertised, “Try my game! It’s just like assembly programming!” in modern IF circles would likely find few takers. Like most of the games that would follow it for the next several years, and unlike the best works of modern IF, Adventure is nothing but an elaborate puzzle box, with no sense or pretension of literary value.
Whatever its failings in modern eyes, Adventure caused something of a sensation in the computing world of 1977. Here was that rarest of all rarities, something truly new. Within months, one Jim Gillogly had ported the game from DEC Fortran to portable C code, making the game available to a host of new computer systems while, not incidentally for fans of proper English everywhere, conferring upon it the blessing of lower-case letters. Stories abound of computer science departments virtually shutting down for two weeks while everyone struggled together to solve Adventure’s many fiendish puzzles and “beat” the game. Graeme Cree’s review of the game for SPAG magazine conveys something of the glow of discovery that still surrounds the game today among those who were there:
Playing a game 70's style was very different from playing today. Since there were few personal computers, playing a game usually involved a trip to the local university computer room, generally after hours, with a bag lunch in tow (since the session would usually last quite a while). My own first experience with Adventure involved late-night trips to IBM with my programmer father. The long trek through dimly-lit windowless corridors to the terminal room was practically an adventure in itself, and since you couldn't just go and play whenever you wanted to, the game had plenty of opportunity to grow larger in the imagination in between sessions (Cree).
Hackers being hackers, as soon as many people completed Adventure they immediately set about improving it. By 1980, a bewildering number of iterations were to be found all over the computing world. Some attempted to improve the game qualitatively by grafting in a parser that could understand more than two words; some preferred to improve it quantitatively by adding more rooms and more puzzles to an already large game; many, of course, did both, and many of these “improvements” led to a game that was less playable than the original. Eventually, players started separating versions by the number of points that could be scored for solving puzzles, which provided a handy way to keep track of the game’s relative size. A player in the original could score a maximum of 350 points. Successive versions, sometimes building on one another and sometimes not, increased this number steadily over the years following the original’s release. The ultimate example of the game, in terms of size at least, arrived in 1992 when David Malmberg created a 1000 point version to demonstrate the capabilities of his Adventure Game Toolkit (AGT) shareware IF development system.
Adventure was one of the first games to appear on the personal computers that began to hit the marketplace at the end of the seventies, and at one point during the early eighties there were at least half a dozen competing commercial versions available for purchase. Among those selling Adventure at this time was a small company called Microsoft that got into the act when it produced a version of the game to bundle with the operating system it had written for IBM’s first Personal Computer. Even the game’s name was not static. In addition to Adventure, various versions referred to themselves as Colossal Cave, Humongous Cave, and Jewels of Darkness. Along the way, Adventure loaned its name to a whole genre of computer games – adventure games, naturally enough. In the early days, the term was synonymous with IF, but when graphical adventures hit the marketplace they borrowed and eventually co-opted the name. Ironically enough, when one sees discussion of “adventure games” today it is generally the point-and-click, graphical style of games ala Myst that are being referred to, rather than the textual genre of which the original Adventure game was the first example.
Even though he is noted as a proponent of strong intellectual property rights,
Bill Gates did not bother to compensate Crowther and Woods when his
company Microsoft “borrowed” their game for commercial release.
As for Crowther and Woods, their choice not to seek any copyright protection for their work, although understandable in an era when the very notion of copyright protection as applied to computer software was something of a foreign notion even to those working in the field, meant that they made virtually nothing from the game. Out of all of the companies who released commercial versions during the first decade of the personal computer, only one – the Software Toolworks, later renamed Mindscape – sought out the game’s original creators to award them a royalty for writing not only the most important work of IF ever (because it was the first), but one of the most important and widely distributed computer games of any type ever created (Adams). I work in the information technology field in a job that puts me into contact with older, “legacy” computer installations quite frequently. For years, I have made something of a hobby of typing “ADVENT” – the shortened abbreviation by which the game is known on most systems – at random command prompts when the opportunity arises. A surprising number of systems, even today, still have the hoary old classic buried in their recesses.
Before leaving Adventure, there is one final anecdote too fascinating to ignore. During the early eighties, a student in German literature at the University of California San Diego named Mary Ann Buckles was exposed to the game and the cult of players it had engendered, and was fascinated to enough to write her PHD dissertation on the subject. In it, she refers to Adventure as “the first major literary work of the new computer medium” (Buckles ix), and states that it “can be treated as literature because it is written in words, conveys stories, and can evoke powerful emotional involvement in the imaginary world it embodies” (Buckles 3). She concludes that “interactive fiction can develop into a serious artistic medium” (Buckles ix). For Buckles to be making these arguments in an academic thesis published in 1985, at a time when computer games were considered strictly children’s toys even among many who were creating them, is not only astonishingly insightful but equally brave. Sadly, most of her peers in academia were actively hostile to her work. In a recent interview, Buckles stated that, in her fellow scholars’ view, “You either had a soul, or you worked on computers” (Hebert). They “just didn’t like it – it wasn’t their thing. They’d never heard of the topic. They didn’t know what to do with it” (Hebert). Buckles was so disillusioned by the rejection of her peers that she retired from academia altogether to take up a career as a massage therapist. She deserves enormous credit for her foresight in seeing the potential of IF in particular and interactive narrative in general. In a supreme irony, her alma mater UCSD is now at the forefront of the study of so-called “new media.”
Sources for Further Investigation
Adventure. 1976. Will Crowther and Don Woods.
An MS-DOS executable created from the original Fortran source is available at
A more modern, user-friendly re-implementation by Graham Nelson is
Eliza. 1966. Joseph Wiezenbaum.
A faithful re-implementation in Java of Wiezenbaum’s original was written by Charles Hayden, and can be found on Dennis G. Jerz’s page devoted to the game at http://jerz.setonhill.edu/if/canon/eliza.htm.
Galatea. 2000. Emily Short.
Humongous Cave. 1992. David Malmberg.
Hunt the Wumpus. 1972. Gregory Yob.
A re-implementation of the original by Magnus Olsson is available at
Hunter, in Darkness. 1999. Andrew Plotkin.
Wumpus 2000. 2004. Muffy St. Bernard.
Zork I. 1980. David Lebling and Mark Blank; Infocom.
Released as freeware by Activision, and available at
Buckles, Mary Ann. “Interactive Fiction: The Computer Storygame Adventure.” University of California at San Diego PHD Thesis, 1985.
Cree, Graeme. Review of Adventure. SPAG Magazine Number 8 (February 5, 1996).
Hafner, Katie and Matthew Lyon. Where Wizards Stay Up Late: The Origins of the Internet. Simon and Schuster: New York, 1996.
Hebert, James. “Accidental Traveler in a Brave New World.” Sign on San Diego (January 22, 2006).
Jerz, Dennis G. “Eliza – Wiezenbaum (1966).”
Jerz, Dennis G. “Hunt the Wumpus – Gregory Yob (c.1972).”
Levy, Steven. Hackers: Heroes of the Computer Revolution. Penguin: New York, 1994.
Montfort, Nick. Twisty Little Passages: An Approach to Interactive Fiction. MIT Press: Cambridge, 2003.
Muckenhoupt, Carl. Review of Adventure. Baf’s Guide to the IF Archive.
Nelson, Graham. “The Craft of Adventure (2nd ed.).”
Wiezenbaum, Joseph. “Eliza – A Computer Program for the Study of Natural Language Communication Between Man and Machine.” Communications of the ACM Volume 9, Number 1 (January 1966).
Available online at http://i5.nyu.edu/~mm64/x52.9265/january1966.html.
Yob, Gregory. “Hunt the Wumpus.” The Best of Creative Computing, Volume 1. Ed. David Ahl. Creative Computing Press: Morris Plains NJ, 1976. Available online at http://www.atariarchives.org/bcc1/showpage.php?page=247.