Zork III [Infocom >RESTART]

Cover of Zork III

IFDB page: Zork III
[This review contains lots of spoilers for Zork III, as well as minor spoilers for Zork I and Zork II. Also, I wrote an introduction to these Infocom >RESTART reviews, for those who want some context.]

Zork III opens in this location:

Endless Stair
You are at the bottom of a seemingly endless stair, winding its way upward beyond your vision. An eerie light, coming from all around you, casts strange shadows on the walls. To the south is a dark and winding trail.
Your old friend, the brass lantern, is at your feet.

Check out those adjectives! “Endless”, “eerie”, “strange”, “dark”. Immediately the game’s tone differs radically from its predecessors. Where parts 1 and 2 were light, playful, and adventurous, Zork III feels austere, somber, ominous. If not for the appearance of the brass lantern and the Elvish sword, it would hardly seem like it belonged in the same set as those other games.

>EXAMINE NEW TONE

Speaking of which, let’s talk about that sword. In Zork I, we find it hanging above the trophy case in a house. In Zork II we simply find it lying at our feet when the game begins. Zork III takes it to a much different emotional place, as just south of the opening room we find a junction, containing this:

Standing before you is a great rock. Imbedded within it is an Elvish sword.

Try as we might, we simply cannot pull our sword from this rock. (Though try often enough and we’ll get a jokey response from the game.) Up to this point, the sword has been a casual knockoff of Tolkien tropes, right down to the way it would glow in the presence of danger. It lived in a decidedly non-Tolkien world, existing side-by-side with robots and inflatable boats. Here, suddenly, this sword has taken on a mythic, Arthurian resonance. Now it’s not just another treasure to be collected, but something signifying destiny, echoing the Dungeon Master’s words in the opening text of the game — “Seek me when you feel yourself worthy!”. The SCORE command reinforces this theme: “Your potential is 0 of a possible 7, in 34 moves.”

The terms have changed, fundamentally. Where scores in the previous installments numbered in the hundreds, and incremented whenever a puzzle was solved or a treasure collected and stored, here the score is measured in a mere seven notches. Not only that, the game describes it as “potential” rather than “score”, and updates it without notice, sometimes after moves that are seemingly disconnected from any puzzle-solving activity.

Descriptions, too, take on a quality of solemn awe. There were incredible vistas in the previous two games, but the most remarkable ones tended to be sunlit — the volcano, the Aragain Falls. Compare that to the Flathead Ocean:

You are at the shore of an amazing underground sea, the topic of many a legend among adventurers. Few were known to have arrived at this spot, and fewer to return. There is a heavy surf and a breeze is blowing on-shore. The land rises steeply to the east and quicksand prevents movement to the south. A thick mist covers the ocean and extends over the hills to the east. A path heads north along the beach.

This is a marvelous image — an underground ocean! We’ve seen a reservoir, and a stream, and even a glacier, but this use of water feels like it comes from a different register, statelier and gloomier. We are in a much more serious & sad landscape now, nowhere near the unicorns and hot pepper sandwiches of previous games. Even when this game pays off the running “Hello Sailor” gag from the previous two games, it does so with an ancient mariner on a Viking-esque ship, sailing off through the mist — a grand and somewhat melancholy image. When the locations aren’t awe-some, they’re still solemn: a creepy crawl, a foggy room, a damp passage. There’s a distinct lack of cheer in this game.

What’s going on here? Primarily, I think it comes down to the fact that the Infocom Implementors had run low on pieces of the mainframe Zork to adapt for microcomputers. Zork I and Zork II covered the vast majority of that original mainframe game, leaving just a few set pieces for Zork III — pretty much just the mirror box, the royal puzzle, and parts of the endgame. As a result, they could create a game with its own sense of thematic unity, freed (mostly) from the hodgepodge aesthetic of mainframe Zork.

Map of Zork III

The theme that they chose reflected Infocom’s increasing seriousness about the potential of text adventures, or as they would come to be called, interactive fiction. In fact, I would argue that Zork III is the first game to cross the threshold from text adventure to interactive fiction. While Zork II had introduced an element of story with its wizard, Zork III brought a consistency of tone and style alongside an emphasis on actual character development for the PC — the journey to worthiness signaled first by the sword in its stone. At this time, Infocom was developing games like Starcross and especially Deadline, which would take the fictional element even further, so it’s little wonder that they gravitated towards seriousness and heft for Zork III.

>TAKE SWORD’S FIGURE

At the same time, moving to a new thematic register introduced new demands on the form, demands which its limited interface and draconian space constraints weren’t always prepared to handle. Take the hooded figure, for example. As we wander the thematically-on-point Land of Shadow, at some point or another we will encounter “a cloaked and hooded figure” who carries his own brightly glowing sword. When this happens, “From nowhere, the sword from the junction appears in your hand, wildly glowing!” This is one of those moments where the score increments without telling us — a puzzle has been initiated, not solved, but our potential has increased.

We’re meant to battle this figure, but not kill it. When it’s helpless before us, we are to remove its hood, prompting this scene:

You slowly remove the hood from your badly wounded opponent and recoil in horror at the sight of your own face, weary and wounded. A faint smile comes to the lips and then the face starts to change, very slowly, into that of an old, wizened person. The image fades and with it the body of your hooded opponent.
The cloak remains on the ground.
Your sword is no longer glowing.

Great, right? Straight out of Joseph Campbell, or at least The Empire Strikes Back. The idea is that we treat this opponent differently from the troll or the thief, showing compassion rather than murdering it, and in doing so we see ourselves in our foe, along with a hint of the ascension to come.

The problem is this: you can pick up very easily on the concept of the challenge and still not have the faintest idea how to achieve it. Dante and I attempted so many actions with the hooded figure, and were rebuffed at every turn. We tried talking to it. We tried giving bread to it. We tried “help figure”. We tried telling it to go away. We tried kissing it. We tried disarming it, stumbling in the process into weird parsing failures:

>take sword
You already have it.
The hooded figure attempts a thrust, but its weakened state prevents hitting you.

>take sword from figure
You already have it.
The hooded figure attempts a thrust, but its weakened state prevents hitting you.

>take figure's sword
The hooded figure isn't in the sword.
The hooded figure attempts a thrust, but its weakened state prevents hitting you.

>take sword's figure
You already have it.
The hooded figure attempts a thrust, but its weakened state prevents hitting you.

>[whatever man]
I don't know the word "[whatever".

Every one of these attempts was met with either confusion (as above) or some flavor of flat refusal — “The hooded figure does not respond to your words.” It’s the quintessential IF communication breakdown. We understood what we wanted to do. We understood that the game wanted us to do it, or something very like it. We just couldn’t make all the various ways we could think of to enact the concept connect with the one specific way that the game had prepared for the concept to be enacted. A more modern game — I’m tempted to say a better game, but I know the limitations of the early Z-machine — would have found ways to either let those alternate solutions stand as solutions, or respond to them in ways that hinted us toward the right answer.

>PUT RING ON SEAT

We ran into a similar issue with the time machine and the golden ring. Well, eventually we did — first we had to struggle through a bunch of false starts trying to figure out what the heck to even do with the time machine. See, we found our way to the Technology Museum well before we ever went into the Jewel Room, so right off the bat we got into the machine, set the dial to 000, and:

>press button
You experience a brief period of disorientation. The area around you seems to be solidifying! Rock formations close in on you and before you can react you are engulfed in stone!

** You have died **

An Apple ][ or something would have printed “END OF SESSION” after that and returned to a command line, but the Windows Frotz interpreter we were using just shut down! It was “[Hit any key to exit]” and goodnight, Gracie. Dante, especially, was stunned. “We perma-died!” he said. This was in stark contrast to the friendly death-evasions we were granted in the previous two games.

Then we perma-died a bunch more times, testing out the various possibilities of the time machine. Mind you, we had no idea what era to aim for, so we were just flailing around. From teleporting into solid rock, we eventually found ourselves… shot with ray guns? In these instant death scenes, the overall grim tone of the game turns suddenly comedic, as “a row of military people who, if appearances do not deceive, have the cumulative intelligence of a not yet ripe grapefruit” shoot us with a “waffle-shaped implement.” “Killed by a waffle!” said Dante.

Illustration of the Dungeon Master from Zork III

Even once we found our way to the Jewel Room, and therefore the sign hinting at what year to specify, we still struggled mightily with this puzzle. First of all, we found no clue at all that the machine can be pushed from place to place, but okay, let’s set that aside. It falls into the realm of fair, just barely. Much less so was the puzzle around transporting the ring back from the past. Once again, we couldn’t match exactly what the game had in mind, even though we had the idea.

Clearly we were supposed to somehow store the ring in the time machine. It’s not at all apparent why we could transport ourselves but not our possessions, nor why the time machine’s seat was immune from this property, but we quickly put together that the seat was a container, and therefore crucial to the answer. However, it turns out that the seat is two different kinds of container, and we only used one kind: putting things on it. We can put things on the seat and they travel from the present into the past. Voila!

Except, we can’t put things on the seat and have them travel from the past into the present. Why? It’s never explained, and in the end made the whole thing feel buggy, like maybe the present-to-past thing was an unintended accident. Not to mention, it makes for straight-up bugs like this:

Sitting on the seat is:
A golden ring (being worn)

This was after we’d already experienced the frustration of having some other logical approaches in the same section fail — sneaking past the guards with our invisibility potion, or using our magical-morphing key to unlock the museum door. Finally, in desperation, we looked at the hints and found we were supposed to LOOK UNDER SEAT, at which point the game cheerfully says, “You notice a small hollow area under the seat.” This is the very definition of a guess-the-verb puzzle, especially when the response to EXAMINE SEAT is “There is nothing on the seat,” which strongly cued us to believe that the seat is a surface for placing things onto. Dante said, “We had the right idea — it was the game’s fault we couldn’t solve this puzzle.” I quite agree.

On the other hand, once we do solve the puzzle, the plaque in the Jewel Room reads:

The plaque explains that this room was to be the home of the Crown Jewels of the Great Underground Empire. However, following the unexplained disappearance of a priceless ring during the final stages of construction, Lord Flathead decided to place the remaining jewels in a safer location. Interestingly enough, he distrusted museum security enough to place his prized possesion, an incredibly gaudy crown, within a locked safe in a volcano specifically hollowed out for that purpose.

Oh Zork. That was awesome. All is forgiven — I love you again.

>PUT OTHELLO BOARD ON TABLE

Other puzzles were much more satisfying. The Royal Puzzle is one I’ve solved numerous times in my life, but it’s so complicated that I’m always starting from scratch each time. That’s a characteristic of many puzzles in Zork III — where other episodes tended towards conceptual puzzles like the riddle, the prayer, or the cyclops, this game favored mechanical puzzles like the time machine, the mirror box, and of course the Royal Puzzle. While previous games (especially Zork II) spread many small puzzles across the game to create the layers of a larger goal, this game builds multiple layers into many individual puzzles.

The Royal Puzzle, for its part, certainly has numerous mini-goals to accomplish, all using the same basic mechanic. It pushes up against the limits of text games, resorting to crude ASCII maps of each location, and it certainly pushed up against the limits of our mapping capabilities. Dante and I ended up pressing an Othello board into service to track the borders, the movable pieces, and the goals. Solving it felt quite triumphant.

The opening screen of Zork III

As great as that was, though, for me in this playthrough the most thrilling puzzle was the Scenic Vista. That’s the one where you find a magic viewer that can take you into Zorks past, present, and future, and in the process visit some objects and places that will help you later on. Probably I loved this one the most this time because of the circumstances — playing through the games in succession with Dante was a bit like having that magic table, revisiting familiar landscapes with a new purpose in mind. Also, I find it hilarious that while Zork III resolves the “Hello Sailor” question, the broken timber is still a red herring.

And then there’s the earthquake. This is another paradigm shift from previous games — aspects of the dungeon change on a timer, not based on any action of ours. On the one hand, it’s another piece of the shift towards story and away from game — events in the world are not centered on the PC, and happen for their own reasons. On the other hand, as an interactive experience it amounted to a way that pieces of the game could close off, or open up, completely outside our control.

We came upon the broken aqueduct very late in our progress, and had pretty much forgotten about the earthquake. It took us a very long time to figure out what to do… and then realized we had to replay. So yay, another way Infocom found to force us to restart. No light limits this time, but yet another time limit snapped the game into an unwinnable state which we didn’t find until we’d played large swaths of it, including the “who would ever want to replay that?” Royal Puzzle. Once again, I ran through it by myself, Dante waiting for me to finish so the fun part could resume. Though on the bright side, our travails at the Aqueduct View and subsequent Aqueduct prompted him to observe, “Whenever there’s a location in Zork that has ‘View’ in its name, you can bet that you’ll be visiting the place that it is a view of later.” That was enjoyable.

>LEAN ON STAFF

Finally, we made our way past the guardians, employing another complicated and multi-layered mechanism, and reached the endgame at last. We had figured out at this point that we were supposed to be emulating the Dungeon Master himself, and were excited to see what final challenge the game had in store for us. Then we found it and… were thoroughly confused. We took many, many runs at this final puzzle, but couldn’t make a lick of sense out of what was supposed to be going on — shades of the Bank of Zork. After numerous iterations of this, we said:

>lean on staff
Are you so very tired, then?
>YES

Finally, we looked at the hints and… were STILL confused. Oh, we knew how to solve the puzzle. We just didn’t know why that was the solution. We read several Internet accounts of how to solve this puzzle, and found ourselves no further enlightened. Why cell 4? What… just what?!?

The puzzle just seems so arbitrary, so at odds with the many clever constructions that preceded it. I mean, it’s possible I’m just too thickheaded to get the brilliance at work here, but whatever the case, it was an anticlimactic ending to the trilogy. The final text about the treasure and the power and the ascension and all was cool, but it would have felt much more exciting had it been the culmination to a challenge befitting the trilogy, and for us it just wasn’t.

Nevertheless, it was the end of a long and mostly pleasant road. It was a joy to replay this trilogy with Dante on board, and we set our sights next on… the beyond.

Interactive Fiction And Reader-Response Criticism (academic paper) [Misc]

[As I mentioned in the previous post, I wrote a paper about IF for a graduate class in literary theory, circa 1994. I think it’s the first thing I ever wrote about interactive fiction — before I knew anything about the indie IF scene or the int-fiction newsgroups at all. Below is that paper, HTML-ized and with various links added, but otherwise identical to what I turned in.]

>examine the map
The map shows a network of boxes connected by lines and arrows, with many erasures and scrawled additions. Something about the pattern is maddeningly familiar; but you still can’t put your finger on where you’ve seen it before.

>examine the book
The open book is so wide, it’s impossible to touch both edges with your arms outstretched. Its thousands of vellum leaves form a two-foot heap on either side of the spine; the rich binding probably required the cooperation of twenty calves.

The magpie croaks, “Then stand back! ‘Cause it go BOOM. Awk!”

>read the book
It’s hard to divine the purpose of the calligraphy. Every page begins with a descriptive heading (“Wabewalker’s encounter with a Magpie,” for instance) followed by a list of imperatives (prayers? formulae?), each preceded by an arrow-shaped glyph.

The writing ends abruptly on the page you found open, under the heading “Wabewalker puzzles yet again over the Book of Hours.” The last few incantations read:

>EXAMINE THE MAP
>EXAMINE THE BOOK
>READ THE BOOK
Trinity with Paul O’Brian

An electronic fiction is not only a fictional universe but a universe of possible fictions.

–Jay David Bolter (Tuman 34)

The long passage above is from a fictional text about an adventurous individual who discovers the secrets of a magical land in order to avert nuclear annihilation on Earth. It has all the qualities of a traditional1 fantasy narrative — fantastic setting, magical transportation, eerie symbolism — but it is not a traditional fantasy narrative. It has no pages, no hard covers, in fact it isn’t a book at all. It is a piece of computer software, an adventure game entitled Trinity, one which boasts no graphics or fancy features — just text and a “parser” programmed to understand simple sentences typed by the player. Its makers, a company called Infocom, call it “interactive fiction,” and the sentences following the “>” sign were typed by me, accepted into the computer, and processed for a response. The scene I chose was a rather metafictional one, where the player/reader comes across a book within the context of the game which records every move that he or she types at the prompt (“>”).

Interactive fiction (hereafter shortened to IF) is an extremely new form, the most primitive incarnation of which surfaced only about 20 years ago, but already it challenges deeply entrenched notions of narrative, authorship, and reading theory. It demands (and helps to enact) new theories of subjectivity, and presents us with a narrative form which is familiar in some ways, but which in other ways is utterly alien. Its novelty and its radical differences from print narrative make it difficult to theorize, but I hope to make a few definite steps in that direction here. As I see it, literary criticism of traditional texts, especially reader-response criticism, can help us in an endeavor to understand this new form. Theorists like Wolfgang Iser, Norman Holland, and Stanley Fish have created theories aimed at traditional texts which seem custom-made for interactive fiction. In this paper, I will attempt to theorize the computer text through the lens of their theories, and to explore the avenues opened by interactive fiction which offer both a better understanding of this tremendously potent literary form and a fresh perspective on literary theory itself. Before this is possible, however, we need a clearer understanding of the subject at hand.

WHAT IS IF?

Previous critical approaches to IF have had to concern themselves so much with explaining how it came to exist that by the time they got to theorizing, they had largely run out of steam. I hope to avoid that trap by providing an extremely brief history of the evolution of IF, and referring the curious reader to Anthony Niesz and Norman Holland, who provide the best concise narrative of IF’s development.

IF developed in the early 1970s on large university mainframe computers, at which point it was little different from the Choose-Your-Own-Adventure novels of print fiction (“If you want to get in the helicopter, turn to page 12. If you want to go home, turn to page 16”). David Lebling and Marc Blank, working on the mainframe at the Massachusetts Institute of Technology, co-authored Zork, the first game whose parser could understand complete sentences (“Get all but the dagger and the rope”) rather than simple two word commands (“look frog”). The game, a sword-and-sorcery treasure hunt in the Dungeons and Dragons pattern, spread rapidly to other mainframe systems and was such a tremendous success that when personal computers began to proliferate, Blank and Lebling transposed Zork to the PC format and founded a company, Infocom, from which to sell it. Again, it was successful, and Infocom followed it with four sequels (a fifth is forthcoming) and more than 25 other works of IF in the literary traditions of mystery, science-fiction, adventure, romance, and others. Many other companies created programs which followed this trend, but Infocom remained the publisher not only of the highest quality software, but that most dedicated to textual IF rather than graphics or sound-augmented games; for the purpose of clarity and simplicity, this paper will deal only with Infocom programs.2

As indicated by the passage quoted from Trinity, the text of interactive fiction comes in rather short blocks, interrupted by prompts. At a prompt, the program will wait for the player to input her next move. The narrative will not continue until the player responds. Input can consist of anything from the briefest commands (“yell”) to fairly complex sentences (for example, an instruction manual lists “TAKE THE BOOK THEN READ ABOUT THE JESTER IN THE BOOK” as a possible command). The program addresses the player in the second person singular (“you still can’t put your finger on where you’ve seen it before”), a mode which raises interesting questions about the reader’s subjectivity.

IF also calls into question the basic notion of narrative linearity. An Infocom advertising pamphlet reads:

you can actually shape the story’s course of events through your choice of actions. And you have hundreds of alternatives at every step. In fact, an Infocom interactive story is roughly the length of a short novel in content, but there’s so much you can see and do, your adventure can last for weeks, even months. (Incomplete Works)

If the shape of a traditional narrative is a line, an interactive narrative is better conceived of as a web.3 Each prompt, each location in the story, is a nexus from which hundreds of alternate pathways radiate, each leading to new loci. This is not to say that IF destroys linearity, but rather that it removes the responsibility for creating linearity from the author and hands it over directly to the reader. Readers of an IF create their own line through the text, a line which can be different at every reading.

IF AND TRADITIONAL TEXTS

Though it is highly divergent, IF is not completely divorced from traditional texts. For one thing, most IF themes, settings, and structures are drawn directly from literary traditions established by traditional print authors. This extraction can range from simple genre duplication, as in the generalized science-fiction feel of Starcross, to the direct appropriation of character and setting, as in Sherlock: Riddle of the Crown Jewels, a mystery based in Arthur Conan Doyle’s London.

Moreover, the programs often incorporate traditional texts into their packaging, texts which frequently contain clues to the puzzles contained within the game. Trinity, for example, includes a comic book dealing with the themes of the text. A Mind Forever Voyaging, which Infocom calls its “entry into the realm of serious science fiction such as 1984” (Unfold), contains an initial chapter in print which sets the scene for the computer narrative and introduces the player to his “role” within the game.

The line between interactive fiction and traditional print texts becomes further blurred when the player uses the SCRIPT function. SCRIPT commands the computer to print out a transcript of the game as it is being played, so that players can have a hard copy of the text they produce in collaboration with the game. Because of its radically different configuration, IF disrupts many of the characteristics often taken for granted in print fiction. For example, as Niesz and Holland point out, “Because the fiction is inseparable from the system that enables someone to read it, one cannot, as it were, hold the whole novel in one’s hand… One cannot look back at what went before” (120). One way to address these differences is through the SCRIPT function. Indeed, a game like Infidel, which includes a complex system of “hieroglyphics” for the reader to decipher, almost demands some type of transcription if the player is to be able to put the pieces of the code together. Furthermore, most players of IF make maps on paper in order to keep track of the complex web of locations the story. Though IF is essentially a computer-bound form, its readers often find themselves taking pen (or printer) to paper in order to resist the evanescence of screen scrolling.

However, IF’s removal from traditional texts is also one of its greatest advantages. In fact, I would argue that without computer technology, creating interactive fiction would be extremely difficult if not impossible. Choose-Your-Own-Adventure novels not only demand an arduous amount of page-flipping, the options they provide are only of the multiple-choice variety, as opposed to the much greater vocabulary-oriented option locus of IF. As Stuart Moulthrop and Nancy Kaplan argue, “the more intricate page-turning a text demands, the more conscious its reader is likely to become of the native sequence that he is being made to violate” (12). The computer’s advanced ability to handle high-speed calculations allows the text to become more truly interactive without any of the alienation produced by the print text which violates its own conventions. This makes the computer a necessary tool for IF; it is the only medium which allows the type of interactivity at issue here.

Finally, it is important to remember that although IF differs radically and in many ways from traditional texts, it is still a text. Its sentences and paragraphs may be interspersed with and guided by player input, but those sentences (as well as the sentences typed by the player) are still units of language susceptible to theories designed to analyze all such units. Literary theory is not obviated by the interactive text; rather it is pushed into new arenas.

THE NARRATIVE CURTAIN

Neil Randall’s analysis of IF includes a lucid explanation of one of its most noteworthy features: “interactive fiction disrupts the concept of the novel’s ‘last page’… an interactive fiction work’s only indication of ending… is the story’s announcement that the quest has been solved” (189). This effect is part of what I call the narrative curtain. IF allows its reader a certain initial area to explore, but draws a metaphorical curtain around its boundaries, a curtain which promises further expansion at the same time as it blocks the way. The reader’s task is to find the tools necessary to draw back the curtain and continue her progress through the narrative.

This curtain also operates at the level of player input. A player encountering an IF text for the first time has no idea of the limits of the program’s vocabulary, and must discover these limits by trial and error. Even after lengthy interaction, the player is always unable to construct a definitive list of words the parser understands — there always remains the possibility that the program understands words that the player hasn’t attempted. Thus, the curtain remains, promising expansion but blocking the reader’s vision of that expansion. This curtain is again made possible by the computer’s ability to hide parts of the text from the reader, a capability which print simply does not have, unless the narrative is serialized and forcibly withheld from its reader. It is this curtain which allows Niesz and Holland to conclude that “writing and reading as processes replace writing and reading as products” in IF (127).

“In general, the structure [of IF texts] is the Quest” (115), assert Niesz and Holland, and I would argue that this structure has been dictated by the power of the narrative curtain. Without a quest for the player to complete, the narrative curtain becomes more frustrating than satisfying. If the curtain is a puzzle to be solved, then passing through it is an achievement, but if it is simply an arbitrary barrier, the desire to cross it decreases along with the reader’s interest in the story. However, I would argue that without the narrative curtain, a player’s desire to make a line of IF’s web suffers serious attrition. If the text simply sprawls out before the reader with no particular narrative thrust, IF becomes less powerful and engaging than traditional fiction. However, the curtain still remains on the level of input, and if the text presents a world for the player to explore, the reader’s desire to discover the intricacies of that world erects yet another form of the narrative curtain — the allure of textual frontier, of undiscovered country. I would argue that the barrier of the narrative curtain is IF’s analog to the element of conflict in traditional narratives — a text can exist without it, but that existence is a dull one indeed.

A further effect of the narrative curtain is its impact on narrative time. The reason why an IF text can take “weeks, even months” to complete is because it takes time to discover how to pass through the various curtains it constructs for its reader, and the program won’t allow its reader to continue until she has discovered the secret of its puzzle. In other words, unlike readers of a traditional narrative, IF players can’t turn the page until they “deserve” to, that worthiness being earned by puzzle-solving. Computer fictions can adjust narrative time not only through forcing the reader to wait for blocks of text, but by challenging him to pass tests before the story can continue.

This brings up another salient point about IF — the frustration it causes when its reader finds herself unable to solve the puzzle. Nearly every player of Infocom games has experienced the sensation of being “stumped.” In the practical arena this has translated into hint book sales for Infocom, Inc. But what of IF theory? I would suggest that when the narrative curtain becomes impenetrable, IF is in its greatest danger of becoming unfriendly to the reader. The player is likely to simply stop playing if the obstacles become too difficult, allowing the curtain to become permanent and leaving the reader with a frustrating sense of lack of closure. However, there are important caveats to this point. Randall articulates the first:

Even when the reader cannot formulate a solution to advance the interactive text, she can usually back-track to a previous screen or side-step to another… In backtracking, the reader re-reads portions of the text, often many times over, in an effort to find a clue that will allow the barrier to be breached. By side-stepping, the reader hopes to return to the barrier with a new sense of how to surmount it. In either case, what is continuous is not the plot but rather the development of the reader’s knowledge of the world in which her character is travelling. (190)

The other point is that even when the reader leaves the program, the narrative curtain still possesses power. Players often return to a puzzle which is stumping them after weeks or months of inactivity, whenever a new approach occurs to them. When these are the circumstances and the curtain finally is pierced, the satisfaction of achievement is proportionally heightened.

The notion of the narrative curtain brings out one of the most interesting features of IF, its casting of the reader as bricoleur. The closest analog to this casting in traditional texts is in the mystery novel, where the reader often watches actively for clues in order to beat the textual detective to the solution of the crime, gaining thereby a certain sense of satisfaction and even superiority. However, in a mystery novel the answer will be revealed as long as the reader keeps reading. IF, on the other hand, forces the player to collect and combine pieces of the text and recontextualize them in order to pass through the narrative curtain, even if that bricolage is as simple as finding a key for a locked door. Until the player makes the connections the program wants, the Quest simply will not progress, though the line of the story may continue.

THE INTERACTIVE SUBJECT

The question of reader subjectivity is already incredibly complicated, and the advent of IF only adds more questions rather than answering any of those already extant. Thus, this section can only scratch the surface of this intriguing question, but I hope to address some of the main points.

Georges Poulet addresses questions of subjectivity in his essay “Criticism and The Experience of Interiority,” and brings up a point which is useful to the analysis of IF: through the medium of language, both author and reader are forced by the limited nature of their communication to exclude their actual lives and take on the “roles” which the language demands. Of course, the reader’s adjustment is greater because the author is choosing the words while the reader is making sense of them. IF texts complicate the assumption of such “roles.” Since the player is the direct agent of action in the story, one could argue that he is more himself than when reading a traditional text. However, this argument has a complicated flip side.

First of all, the reader’s subjectivity is constrained by the inevitable limits of the parser. Thus when I engage with Zork I am forced into the role of someone who cannot use certain words (the vast majority of words, actually). Moreover, the Quest structure demands that its readers assume the role of someone who cares about completing the task at hand. Thus, Zork asks me to become someone who would kill trolls and remove treasures from an underground labyrinth. Further, IF settings are often quite removed from reality, demanding another role adjustment, belief in hyperspace or magical mushrooms, for example. Finally, the text sometimes forces the player into a more specific persona. In Infidel, for example, the reader’s character is a ruthless, greedy archaeologist who is deserted at the narrative’s outset by his underpaid and overworked crew. Plundered Hearts, an interactive romance set in the 18th century, casts the player as a young woman who is captured by pirates. This role-casting can become highly specific, as in Hitchhiker’s Guide To The Galaxy, where the player is cast as Arthur Dent, the protagonist of the Douglas Adams novel from which the game is adapted. Other games however, allow interactive adjustment of the player’s role. Beyond Zork, for example, allows the player to create a “character”, which she names and for which she chooses the degrees of six different attributes (strength, compassion, etc.) from a fund of percentage points. Bureaucracy asks the player to fill out a form listing certain vital statistics (name, address, boy/girlfriend, age, etc.) and then further confounds that player’s subjectivity by forcing him into a specific situation with little or no bearing on reality and using the statistics provided to torment the player’s persona (The boy/girlfriend runs off, etc.)4

I would argue that no matter what the level, the player of IF texts is forced to some degree to assume the type of role which Poulet asserts for traditional texts. This role is further enforced by the program’s second person singular mode of address. When a typical sentence of IF is “You can’t go that way,” it’s clear that the text enacts an enforced recontextualization on the reading subject. This is where postmodern theories of subjectivity present themselves as useful tools for examination of the interactive subject. Postmodern theory generally accepts the subject as already fragmented, a fragmentation which is easy to see in the interactive subject. I would suggest that readers of IF texts, especially while they are in the process of reading, are excellent examples of “cyborgs,” as Donna Haraway uses the term. In “A Manifesto For Cyborgs” Haraway argues that the distinction between human and machine has already become so blurred that we experience day to day life as cyborgs, thinking of mechanical devices as extensions of our own subjectivity. For the IF player, reading a computer screen and typing English commands on a keyboard in order to direct a mystery narrative in the persona of Dr. Watson, this assertion isn’t difficult to accept.

INTENDED, IMPLIED, AND IDEAL READERS

The nascent stages of the formation of reader-response criticism were characterized by theorists like Walker Gibson and Gerald Prince, who constructed theories of readership which attempted to complicate the idea of reader reception by suggesting that texts ask readers to adjust their own perceptions of reality in order to conform to textual paradigms. We have already seen that IF forces a complicated and fragmented subjectivity on its reader, and examination of Gibson’s and Prince’s theories can help us to better understand this process.

Gibson, in “Authors, Speakers, Readers, and Mock Readers,” postulates a “mock reader,” arguing that “every time we open the pages of another piece of writing, we are embarked on a new adventure in which we become a new person — a person as controlled and definable and as remote from the chaotic self of daily life as the lover in the sonnet” (1). IF seems to literalize this process, while complicating it with the question of player agency. Although the player does assume a certain role, she also directly controls the actions of that persona. Thus, like traditional texts, IF constructs a niche into which the player must force herself, but within that niche she is allowed a freedom of movement unheard of in traditional texts.

Gerald Prince’s article “Introduction to the Study of the Narratee” introduces (appropriately enough) his concept of the narratee, “the narratee being someone whom the narrator addresses” (7). Prince goes on to use this concept to subdivide the types of readers. He distinguishes the real reader, the individual with the book in his hands, from the virtual reader, the intended audience of the author. He also posits an ideal reader, who would theoretically have a perfect understanding of the text in all its nuances. This ideal reader is, of course, a fiction. The narratee is separate from of these, though it incorporates elements of each. Like Gibson’s “mock reader” the narratee is someone who the real reader becomes in the process of reading. Since the narratee is the person to whom the narrative is directed, the text itself dictates the narratee.

These classifications are intriguing, but IF demands not only new categories but new scrutiny of the old ones. I would suggest three basic categories which have the mnemonic advantage of all starting with the same letter: Intended, Implied, and Ideal readers. The Ideal reader of an IF text is just as fictional as the one Prince stipulates. An Ideal reader would know the entire vocabulary of the parser, understand immediately how to pass through each narrative curtain, and appreciate every aspect and nuance of the interactive text. The Intended reader, on the other hand, while still a fictional construct, is more firmly within the realm of possibility. Intended readers are who the author of the IF text had in mind as average consumers. They would be able to appreciate the text and solve the puzzles, and would hit on some of the IF’s features, but would still leave some narrative curtains standing, and may miss some textual nuances. Implied readers are analogous to mock readers and narratees; they are the reader-personas to whom the narrative is directed. The text makes certain assumptions about them, such as their fluency in English and their basic grasp of textual conventions, but none about their ability to pass through narrative curtains or to grasp textual nuances. Each of these readers is a fictional construct, and together they form a continuum, from Ideal to Implied, into which the real reader necessarily must fall.

The advantage of this system of classification is not so much in its accuracy, but in its utility for further complicating questions of authorship and intentionality. For example, one problem with the Ideal reader is that, although she would grasp textual nuances perfectly, her perfect ability to instantly pierce narrative curtains would detract from her overall interactive reading experience. I would argue that part of the pleasure of the text lies in the challenge of bricolage which it issues the reader, and if solving the puzzles is no challenge, then that pleasure is drastically decreased. The Implied reader, on the other hand, while he may be familiar with basic conventions of IF, may also be immediately stumped or may disagree violently with the ideological content of the text. Thus, the “narratee” of IF may in fact be alienated by the text to an even greater degree than he would be by a traditional fiction.

Furthermore, the idea of intentionality is a highly dubious one, and even more so in IF. If we accept that traditional fictions are created by readers in conjunction with texts, as reader-response critics assert, then the author drops out of the picture. IF encourages this process through the tendency of computer software publishers to elide the names of programmers. More importantly, however, the notion of intentionality receives a staggering blow from the basic form of the text: one which refuses to acknowledge itself as complete until the reader fills in its gaps.

FILLING IN ISER’S GAPS

The notion of the narrative gap was first articulated by Wolfgang Iser. In “The Reading Process: A Phenomenological Approach,” Iser asserts that “a literary text must therefore be conceived in such a way that it will engage the reader’s imagination in the task of working things out for himself, for reading is only a pleasure when it is active and creative” (51). If this sounds remarkably like IF theory, the reason is that Iser conceived of reading as an active, creative process, whose agent is continually filling in the blanks for the text. Iser finds these blanks, or “gaps”, in areas of the text which somehow thwart our preconceived expectations: “whenever the flow is interrupted and we are led off in unexpected directions, the opportunity is given to us to bring into play our own faculty for establishing connections — for filling in the gaps left by the text itself” (55). I would argue that this description still applies to IF — the blocks of text in an IF program are just as capable of producing such moments of unexpectedness as any in print fiction.

However, IF obviously adds another level to the notion of gaps in a text. As Randall points out, “interactive fiction displays two types of narrative gap: the traditional type filled unconsciously by the reader, and a manifest type shown on the computer’s screen” (189). Every time the reader sees a prompt, she is expected to input some command, without which the story is simply halted. This prompt, this halting, is clearly a new type of narrative gap, one which the reader fills in more literally than Iser ever imagined.5 Iser assumes that readers of traditional texts always imaginatively fill in the gaps left by the text at its unexpected moments. IF, however, goes one step further, by forcing an active reading. It is this forceful narrative gap which makes IF so friendly to analysis in reader-response models. Far more than in a traditional text, the reader must respond if the text is to continue to reveal itself. Moreover, that response must consist of elements which the text can parse, or else the reader receives a message like “I don’t know the word ‘deconstruct'” (for example) and is returned to the prompt.

Iser uses the metaphor of constellations to clarify his theory: “two people gazing at the night sky may both be looking at the same collection of stars, but one will see the image of a plough, and the other will make out a dipper. The ‘stars’ in a literary text are fixed; the lines that join them are variable” (57). Iser’s stars are the printed units of language, and his lines are the conscious or unconscious connections drawn by the reader. I would contend that IF adds even more resonance to this metaphor. The reader of IF not only draws the connections Iser refers to, but activates explicit connections between textual blocks at every prompt, creating her own unique “constellation” of narrative, one whose shape clearly manifests itself in the line of the story under her guidance.

IDENTITY THEMES IN IF

If IF indeed helps to extricate and document the ways in which readers fill in narrative gaps, what can we learn from the differences in the constellations they create? One theory which presents itself as useful for answering this question is the one presented by Norman Holland in “UNITY IDENTITY TEXT SELF”. In that essay, Holland argues that unity and text have an analogous relationship to identity and self, the former term in each pair being an abstraction from patterns evidenced by the latter term. He argues further that because we are each possessed of and immersed in an identity, we extrapolate textual unity according to certain themes inherent in that identity (and the same applies to our extrapolations of the identities of others). For Holland, reading is a process in which, the pieces of self interlock with the pieces of text to form a reading, an interpretation of text which is inevitably stamped with the reader’s “identity theme.” Under this paradigm, interaction with an Infocom game operates like zipping up a zipper — the self and text interpenetrate to form a seamless narrative line, the direction of which must indicate elements of the reader’s identity theme.

However, there is a definite danger of oversimplification in this model. For one thing, no two readings of an interactive text, even by the same reader, are exactly the same. This fact calls into question the notion that identity themes can be traced even when the reading process is made more visible. In fact, it seems entirely plausible to me that the varying narrative lines of the interactive process are part of another binarism analogous to those Holland discusses. Rather than a pattern which shows itself in one reading, the visible identity theme evoked by IF is more accurately (I would argue) an extrapolation of tendencies within a grouping of several narrative lines from the same reader. Thus, as identity is to self and unity is to text, so the visible identity theme is to the group of narrative lines.

Moreover, Holland’s model allows no room for change. I would argue that the experience of interactive fiction has a direct effect on the reader, and that the crossing of a narrative curtain drastically affects the next reading of that same situation. Once I’ve discovered, for example, the secret of the Echo Room in Zork, I am highly unlikely to spend as much time there as when I was trying to puzzle it out. Conversely, the crossing of narrative curtains often allows the text to expand, showing the reader new locations or new features. It seems plausible to me that the text and player of IF can be read as providing growth stimuli for each other. Thus interactive fiction is a mutual growing experience, not only for the reader, but for the text as well, and the result of such a reading is not a text imprinted with the reader’s identity theme or a reader stamped with the themes of the text, but a unique collaboration of reader and text, one which partakes from both in order to create an unprecedented result.

READING AND WRITING: OBSOLETE CATEGORIES?

As we progress further and further into theorizing IF, it becomes increasingly clear that notions of authorship and readership are becoming more and more difficult to define. It is at this point when the theories most useful to us are those of the paradigmatic theorist of reader-response criticism, Stanley Fish. “Literature in the Reader: Affective Stylistics” was the article which announced Fish’s perspective on the reading process, a perspective which has come to be identified almost programmatically with reader-response criticism. In it, Fish posited meaning as an event, a “making sense” enacted by the reader. By closely reading sentences, and analyzing their capacities for forming meaning in a temporal dimension through the succession of words, Fish showed that meaning is not extracted from a text as if every word in it were received simultaneously. Fish analyzed the dynamics of expectation to demonstrate meaning as a sequence of mental events. His article attempted to change the typical critical question “What does this sentence mean?” to “What does this sentence do?”

I would argue that in the creation of an IF narrative, both the author and the reader are continually asking themselves this crucial question. Authors of interactive fiction write with the consciousness of their medium. They are aware of the narrative’s enforced gaps, and of the narrative curtains they are constructing, and this awareness inevitably affects their writing. For example, text in IF often contains hints about how to pass through certain narrative curtains. The magpie in the passage from Trinity I quoted at the beginning of this paper is squawking part of a formula by which the player may pass through a crucial narrative curtain. Without the magpie’s help, the bricoleur/player would have no idea how to cross this juncture in the narrative. Those sentences were clearly constructed with a mind towards what function they would serve, what they would do. Players of IF texts confront the same question at every prompt. The player typing commands into the computer is forced to continually ask herself “What will this sentence do?” so that she may better communicate with the program. The experience of writing for a parser provides constant reminders of authorship, audience, and action of units of language. It forces upon us the question which Fish only asserts.

The answers to this question seem to show up more clearly in interactive fiction, but this appearance itself is highly dubious. Though authors discover the effect of their sentences on game testers, they obviously cannot learn this effect for every player at every juncture. In fact, since meaning is created through the interaction of player and text rather than player and author, the author’s thoughts become irrelevant. It is the text which ultimately asks the crucial questions. Conversely, though the player always sees the results of his commands, they are not always clear or easily interpreted. While the text usually guides or blocks input, it occasionally does both, and sometimes does neither, producing a response like “Try to phrase that another way.” Though both player and text are asking “What will this sentence do?” the answer is neither easy nor clear.

This is where Fish’s notion of interpretive strategies comes into play. In “Interpreting the Variorum,” Fish asserts that “Texts are written by readers, not read, since, the argument now states, the formal features of the text, the authorial intentions they are normally taken to represent, and the reader’s interpretive strategies are mutually interdependent” (Tompkins xxii). Fish’s argument for traditional texts must be modified a little for IF, because the form muddies the distinction between author and reader. Not only is the reader utilizing interpretive strategies, the text is as well, and furthermore those strategies must coincide if the narrative is to progress. For a narrative curtain to be drawn, the reader needs to have a strategy which will allow her the correct pieces for her bricolage, and the text must have a strategy with which to interpret the player’s input. In this paradigm, the player and computer together fit rather well into Fish’s concept of the “interpretive community,” “made up of those who share interpretive strategies not for reading (in the conventional sense) but for writing texts, for constituting their properties and assigning their intentions” (“Variorum” 182).

However, to truly accept this concept, it is crucial that we absorb one final ideological prop, the notion of self as text as advanced by Walter Benn Michaels in “The Interpreter’s Self: Peirce on the Cartesian ‘Subject'”. In Michaels’ interpretation of the pragmatic philosophy of Charles Sanders Peirce, “our minds are accessible to us in exactly the same way that everything else is. The self, like the world, is a text” (199). Once we accept this dictum, several previously unsettling aspects of IF finally fall into place. First of all, if we can read the self as a text, we can just as easily read the text as a kind of self, especially the text of IF, which actively seeks input. With this understanding, interactive textuality is thrust into a new light. In the interaction of the computer program’s self/text with the player’s self/text, there is no reader and writer. They read and write each other. In the radical realm of interactive fiction, the theories of reader-response critics become facts, and in fact are surpassed. Not only does the reader actively fill in textual gaps, but the text fills gaps in the reader, plants seeds for ideas which haven’t come yet. Human and computer, when both are understood as both self and text, become a true interpretive community, each community producing its own unique fictions.

“Reader” and “text” become provisional terms. The text is a reader just as the reader is a text. Even “player” and “program” are problematic. The IF is, after all, attempting to program its human reader into certain responses, and is doing so with a significant element of playfulness. Interactive fiction, by enacting reader-reception theory, has stripped the most basic terms of literature of their apparent meaning, and cast us into a new location. It is one which is alien and unfamiliar, but it is also a nexus from which radiate thousands of branches of possibility.6

Endnotes

1 Because of its utility and clarity, I here reprint and appropriate for my own use Neil Randall’s definition of a “traditional” text:

Throughout this article I will use the term “traditional” to refer to literature printed on paper, usually in book (rather than magazine) format. I considered such terminology as “book fiction” or “Gutenberg literature” to avoid collocating such diverse novels as Mitchell’s Gone With The Wind and Barth’s Sabbatical, which have in common only the fact that they are written by Americans and printed on paper, but such terminology proved more cumbersome than it was worth. “Traditional”, therefore, says nothing of the degree to which a work is or is not avant-garde; it is simply a reference to medium. (190)

[Back to reference]

2 The sources of the programs I refer to are two anthologies of Infocom programs, The Lost Treasures Of Infocom volumes I and II. Since these packages contain a total of 31 texts, I find it easier to simply cite the collection than each individual text.
[Back to reference]

3 My analysis of the shape of IF owes much to theories developed for hypertext, a type of computer construction related, but not identical to, IF, the primary difference being that hypertext’s method of branching involves “clicking” with a mouse on a significant word or phrase rather than typing words for a parser. Discussions of hypertext which are useful for analysis of IF are included in my list of references.
[Back to reference]

4 Other games vary the persona of the reader even within the context of the program. Border Zone, for example, is divided into “chapters,” each of which casts the reader in a new character.
[Back to reference]

5 However, even this literalization is problematic due to the fact of the parser. I would argue that Fish’s notion of societally mediated boundaries to interpretation comes in handy here. Just as the reader’s imagination is not free to fill in the gaps in just any way, so the player’s input is restricted by certain ambiguous boundaries.
[Back to reference]

6 This paper only attempts to examine IF as it looks through the lens of reader-response criticism. A more complete discussion of IF would include sections on the idea of the varying levels of code inherent in reading an IF text, some system for evaluating quality, a consideration of the political implications of simulation, questions that still remain open, and of course the obligatory “speculation on the future” section. I had originally planned on these, but the project simply became too large. When I revise this paper to turn it into an article, those sections will be included. [LOL Yeah, right. That never happened. –PO 2000]
[Back to reference]

References

Bolter, Jay David. “Literature in the Electronic Writing Space.” In Myron C. Tuman (ed.), Literacy Online: The Promise (and Peril) of Writing With Computers. Pittsburgh: University of Pittsburgh Press, 1992: 19-41

Campbell, P. Michael. “Interactive Fiction and Narrative Theory: Towards An Anti-Theory.” New England Review and Bread Loaf Quarterly 10.1 (1987): 76-84

Fish, Stanley. “Interpreting the Variorum.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 164-184

Fish, Stanley. “Literature In The Reader: Affective Stylistics.” In Jane Tompkins (ed.), Reader-Response Criticism:From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 70-100

Gibson, Walker. “Authors, Speakers, Readers, and Mock Readers.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 1-6

Haraway, Donna. “A Manifesto For Cyborgs: Science, Technology, and Socialist Feminism in the 1980’s.” Socialist Review 80 (1985): 65-107

Holland, Norman N. “UNITY IDENTITY SELF TEXT.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 118-133

Infocom. The Incomplete Works of Infocom, Inc. (advertising pamphlet) Cambridge: Infocom, 1985

Infocom. The Lost Treasures Of Infocom. Cambridge: Infocom, 1992

Infocom. The Lost Treasures Of Infocom II. Cambridge: Infocom, 1992

Infocom. You Are About to See The Fantastic Worlds Of Infocom Unfold Before Your Very Eyes. (advertising pamphlet) Cambridge: Infocom, 1984

Iser, Wolfgang. “The Reading Process: A Phenomenological Approach.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 50-69

Kelly, Robert T. “A Maze of Twisty Little Passages, All Alike: Aesthetics and Teleology in Interactive Computer Fictional Environments.” Science Fiction Studies 20 (1993): 52-68

Landow, George P. Hypertext: The Convergence of Contemporary Critical Theory and Technology. Baltimore: Johns Hopkins University Press, 1992

Michaels, Walter Benn. “The Interpreter’s Self: Peirce on the Cartesian ‘Subject’.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 185-200

Moulthrop, Stuart. “Reading From The Map: Metonymy and Metaphor In The Fiction of ‘Forking Paths’.” In Paul Delany and George P. Landow (eds.), Hypermedia and Literary Studies. Cambridge: MIT Press, 1991: 119-132

Moulthrop, Stuart and Nancy Kaplan. “Something to Imagine: Literature, Composition, and Interactive Fiction.” Computers And Composition 9.1 (1991): 7-23

Niesz, Anthony J. and Norman N. Holland. “Interactive Fiction.” Critical Inquiry 11.1 (1984): 110-29

Poulet, Georges. “Criticism and The Experience of Interiority.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 41-49

Prince, Gerald. “Introduction To The Study Of The Narratee.” In Jane Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: 7-25

Randall, Neil. “Determining Literariness In Interactive Fiction.” Computers and the Humanities 22 (1988): 183-91

Tompkins, Jane. “An Introduction To Reader-Response Criticism.” In Jane
Tompkins (ed.), Reader-Response Criticism: From Formalism To Post-Structuralism. Baltimore: Johns Hopkins University Press, 1980: ix-xxvi

Ziegfeld, Richard. “Interactive Fiction: A New Literary Genre?”. New Literary History 20.2 (1989): 341-372

Twisty Little Passages by Nick Montfort (book review) [Misc]

[This is a review I wrote in 2004, of the first book-length academic study of interactive fiction: Nick Montfort’s Twisty Little Passages. I also posted a version of this review on the Amazon page for that book.]

Just over ten years ago, I was holed up in the University of Colorado at Boulder’s Norlin library, researching interactive fiction. I was a grad student in English, and had a final paper due in my Literary Theory class. Activision had recently released the Lost Treasures of Infocom bundle, reawakening my childhood love of IF, and I felt inspired to write a paper that connected reader-response theory to the actual reader-responsiveness of text adventures. I wanted to cite and to engage with previous academic work on IF, but unfortunately, though unsurprisingly, it had received very little serious critical attention. Sure, I found a few articles here and there, but what I really needed was something substantial, something that offered a critical vocabulary for talking about interactive fiction, that placed it in a literary context, and that presented a basic history of the form.

What I needed was Nick Montfort’s Twisty Little Passages. How strange and funny that ten years later, the paper I wrote for that class finds itself cited in the first book-length academic treatment of interactive fiction. True, the citation only occurs in a passing (and correct) dismissal of reader-response theory as anything but a very limited way into talking about IF, but it makes me feel like part of history nonetheless. Montfort’s book is just what IF needs to establish its rightful place in the scholarly discourse surrounding electronic literature, and indeed literature, full stop. It never fails to be informative, and frequently succeeds at being sharply insightful about the literary elements of IF.

However, Twisty Little Passages is quite suitable for readers outside the ivory tower as well. Though the book is clearly aimed at an academic audience, Montfort’s prose is blessedly jargon-free, clear, and effective, with generous doses of humor thrown in for good measure. Even in its most theoretical moments, the book manages to balance impressive rigor with unfailing clarity, a feat all too rare in literary theory. Consequently, it’s an entertaining read for general audiences and English professors alike. If you’re an IF aficionado like me, you’ll find Twisty Little Passages enlightening and fun, and if there’s anyone in your life who genuinely wants to know what interactive fiction is and why they should care, hand them this book.

Just the bibliography alone is a noteworthy achievement; Montfort has synthesized the already extant body of formal IF scholarship and mainstream coverage with much of the important amateur IF theory produced by people like Graham Nelson and Emily Short. Also included are a range of other contributions from the IF community and pieces covering the book’s other concerns, including riddles and computer science. In addition, there is a formidable collection of IF works cited, a list comprising much of the most influential interactive fiction of the past thirty years.

Something else that the bibliography makes clear is the value of Montfort’s personal connections. It’s peppered with references to emails and private conversations with some of the leading lights of IF history: Robert Pinsky, Graham Nelson, Steve Meretzky, and others. Montfort’s ability to gather such firsthand information highlights one of the most important things about Twisty Little Passages: not only is it the first book-length treatment of interactive fiction, it is the first formal treatment I’ve seen that approaches IF from the inside out, rather than from the position of a quizzical spectator. Montfort’s extensive experience in both the academic and IF communities lends him a brand of authority that previous commentators on IF lacked.

Of course, authority only gets you so far — it’s what you do with that authority that counts. That’s what makes the first two chapters of Twisty Little Passages such a particular pleasure: Montfort not only knows what he’s talking about when it comes to IF, he’s got quite a bit of original insight to offer about its literary and theoretical contexts. As with many works of literary criticism that seek to approach an underscrutinized area, the project of this book’s first chapter is not only to expose the topic’s theoretical underpinnings but to define and delimit a specific vocabulary for use in discussing it. Montfort does an excellent job of providing a clear definition of IF (and indeed of making the case for the term “interactive fiction”) and of defining a set of terms to identify the subcomponents of the IF experience.

For example, according to the book, a session is “what happens during the execution of an IF program. [It] begins when an IF program starts running [and] ends when the program terminates”, while a traversal is “a course extending from a prologue to a final reply, and from an initial situation to a final situation” — thus a traversal can encompass many sessions (as frequently happens in the case of long, complex games), or a session can encompass many traversals (as happens with short games with high replay value.) Of course, at bottom the choice of terms is more or less arbitrary, but it is crucial that we be able to name the various parts of the IF experience — they are our stepping stones to more sophisticated discussions. Twisty Little Passages lays this groundwork admirably.

On the whole, this book seems more interested in surveying the territory of IF than in making unified arguments about it, but the exception to this is chapter two, where Montfort makes the case that the most important literary progenitor of IF is the riddle, and takes a counterpoint to the most famous analogy and contextualization of IF from the last decade:

The riddle, like an IF work, must express itself clearly enough to be solved, obliquely enough to be challenging, and beautifully enough to be compelling. These are all different aspects of the same goal; they are not in competition. An excellent interactive fiction work is no more “a crossword at war with a narrative” (Nelson 1995a) than a poem is sound at war with sense.

This is a brilliant and entirely convincing comparison. Montfort gives us a brief history of the riddle, and draws the necessary parallels to demonstrate IF’s similarities with it, leaving us with a new paradigm within which to view interactive fiction. Best of all, this angle of approach allows IF to be both story and game, both art and amusement, without detracting from the value of either.

Chapters three through seven, indeed the bulk of the book, are devoted to delineating the history of IF, from its mainframe beginnings to the current amateur renaissance. It’s an entertaining journey, and Montfort’s encapsulation of IF history is concise, approachable, and extremely informative. I found it a little frustrating, though, because it must of necessity skim over the ground rather quickly, especially as it moves into the Infocom era and beyond. Consequently, there are many moments of intriguing literary analysis of IF games, but they end almost as soon as they begin — practically every page contains material that could make a full article in itself. By the time I reached the end of the book, a sort of epilogue that takes inventory of the various ways in which the tropes of interactive fiction have made their way into our culture, I was already wishing for a sequel, one that assayed a more in-depth discussion of games like Mindwheel and Photopia instead of the tantalizing tidbits we get here.

Montfort has already done some of this sort of work, such as an article written with Stuart Moulthrop for an Australian digital arts conference, analyzing the role of Princess Charlotte in Adam Cadre’s Varicella. However, not much work of that sort could appear without Twisty Little Passages preceding it, just as in-depth conversations can rarely occur prior to introductions. This book creates a foundation for the inclusion of IF in literary discussions, and for further examination of specific IF works. Perhaps if we look back on IF criticism in another ten years, we’ll see that introduction as the most important service Twisty Little Passages performs.

What’s next for >INVENTORY

With the publication of my Identity review, I’ve now completed the journey of reprinting every IF Competition review I wrote between 1996 and 2004. There are over 300 of them!

As I said in this blog’s first post, I intend it to house not just my comp reviews, but everything else I’ve written about IF. So now that the comp reviews are all here, I’m going to cast around for everything else of mine I can find on the topic (newsgroup posts and SPAG editorials excepted… for the most part) and get it in here.

That means a few different things:

  • Essays I’ve written about IF
  • Interviews I’ve given
  • Solicited reviews I’ve written
  • Reviews I wrote for the onetime paying home of longer works, Mark Musante’s IF-Review
  • Reviews of commercial adventure games
  • Miscellaneous stuff — correspondence, haikus, IF material from my other blog, etc.

I’m also planning to spruce up the site a bit with easier indexing of its material, and better links from IFDB. In addition, along the way I’ve found it in me to write some new stuff too! The Infocom >RESTART reviews are the first piece of that — so far I’ve written up my experience of playing Zork I and Zork II with my teenage son, and there’s more of that to come. (You can probably guess which one is coming next.)

Beyond that, I’ve found myself curious about what lay beyond the horizon after 2004 in comp-world, so I’ve decided to play and review some games from 2005. As you might expect, this will happen in fits and starts — my life has changed considerably since that era — but so far it’s been a lot of fun. So once I’ve collected and deposited all the stuff I listed out above, there will be new comp reviews! Uh, of very old games. But still new to me!

So although the 1996-2004 comp reviews are done at last (after 16 months!), there is a lot more inventory to unpack! I hope someone finds it useful — it’s been tons of fun to revisit it all.

Identity by Dave Bernazzani [Comp04]

IFDB page: Identity
Final placement: 15th place (of 36) in the 2004 Interactive Fiction Competition

Man, what did I say about the doomed starships with the cryogenic pods? Just two reviews ago I was talking about how they’re the hot new comp game setting, and now here comes another one. Crazy zeitgeist. No doubt each of these authors is bummed to have somehow locked into the concept that’s freakishly dominating the comp this year, because the games are inevitably going to be compared to each other, and when there’s a comparison, somebody suffers. In this case, the game doing the most suffering will certainly be Getting Back To Sleep, but between the two remaining contenders, Identity and Splashdown, the contest is much closer.

They’re both implemented solidly enough to be quite playable, but then again they’re both plagued with numerous typos, formatting errors, and minor bugs. Neither one manages to do anything particularly original with the “crash survivor” plot, though both provide a series of relatively enjoyable (if rote) puzzles. In the end, though, I’m afraid I have to give the nod to Splashdown, because that game just seems to be a little more interested in telling a coherent story than Identity is. Splashdown‘s comparative richness of story and setting arises from fillips like its PDF feelie, but also from the choice of story elements like a more interesting origin for its craft’s demise.

One of the weakest points of Identity is actually one of its main points: the PC’s amnesia. Even setting aside the fact that the amnesiac IF PC is an exhausted cliché, there is no reason that I can see for this PC to be thus stricken. From the first 30 seconds of the game, we can piece together that the PC is a guy who was on a starship in cryogenic sleep, and that the starship has now crashed. After playing through the whole thing and encountering several scenes where the game tries to fill in more memories, what we know about the PC by the end is… that he’s a guy whose starship crashed.

In contrast to a game like Square Circle, where the revelation of the PC’s identity puts something at stake due to the memories and knowledge about him possessed by various characters in the game’s milieu, Identity‘s PC wanders around on a planet full of strangers who seem actively and unnaturally disinterested in who he is. His true name matters to no one, even himself, and thus concealment of it buys the story nothing. The game would have worked exactly the same way if the PC’s memories had been complete and intact at the beginning of the story, and I’m not sure why the game gives him amnesia at all, except to conform to some misguided notion that all good PC’s don’t know who they are.

Either that, or perhaps the original plan for the game included some subplot in which the character’s identity mattered, a subplot that may have been cut to meet the comp deadline. In this latter case, though, the amnesia should have been excised. It wouldn’t have been too hard — just the removal of a few extra bits of prose and a retitling.

I’m more inclined to suspect that the amnesia was introduced in order to conform to a sort of 80’s old-school blueprint, because the game itself feels like a direct descendant of some of the sub-Infocom work from that decade. Everything feels very mechanical, including the people and animals. For instance, there’s a yak in the game, but the writing does very little to evoke anything yak-like about it, and instead it behaves somewhat like a horse, somewhat like a cat, but mostly like a yak-shaped car whose ignition key must be obtained (naturally) by solving a puzzle.

This puzzle, like many of the puzzles in the game, involves observing what few things are implemented and figuring out how they might interact with each other in game-logic. Not natural logic, of course, or else the solution to the yak puzzle would have worked equally well in another puzzle with a virtually identical objective. This approach isn’t my favorite — I prefer the realism that’s come in with the best IF of the last decade. Still, it’s enjoyable enough for what it is, and one area in which this game deserves praise is in its handling of unexpected verbs.

For instance, as an antidote to amnesia I tried REMEMBER, and was quite pleased to see that the game handles it (albeit with a default response.) Elsewhere, I found myself in a chair with some safety straps I could fasten, and smiled when SECURE STRAPS worked as expected. This game has clearly been tested, and that counts for a lot with me. Unfortunately, the testing didn’t quite weed out all the bugs, both in coding and in prose mechanics, so another round is required. Identity isn’t an unpleasant way to spend an hour or so, but for me it felt mostly like a missed opportunity.

Rating: 7.8

Zero One by Edward Plant [Comp04]

IFDB page: Zero One
Final placement: 31st place (of 36) in the 2004 Interactive Fiction Competition

Because I am a whiny malcontent who is never satisfied, I’m beginning this review with another complaint about Alan‘s transcripting capability, or lack thereof. Last year, I moaned about the fact that Alan doesn’t offer a SCRIPT command, and therefore I was having to periodically copy and paste from the Glk scrollback window into a text editor in order to have a transcript for reference while I wrote my reviews. In response, a few friendly people informed me that if I start the Alan interpreter from the command line with a “-l” switch, it will indeed log a transcript.

This, though a little annoying, is happy news. I tested the method this year, and it works! Sort of. For no reason that I can ascertain, the transcripts are saved in an extremely goofy format, with one line per turn, a line that begins with output generated by the game, then lists the name of the current room, and ends with whatever I type at the prompt. Thus, the slightly more readable parts of the transcript look like this:

There is nothing special about the door.Cell> x walls
‘walls’? I don’t know that word.Cell> n
The door’s closed.Cell> open door
It’s locked.Cell> unlock door
You can’t unlock that!Cell> knock on door
You knock on the door.Cell> z

The less readable parts, which are predominant, occur when the game has anything substantial to say — they stretch off into the distance or wrap (depending on the text editor) to form a busy jumble of unformatted verbiage. Also, on a more minor point, I don’t get to choose the filename for the transcript, and Alan uses an inexplicably super-funky naming convention that gave my log files titles like “011100149966.log.” This transcripting capability is better than nothing, but the quality is still unacceptable. Come on, Alan. Transcripting is kind of a basic IF function, going way back to the 80s. Help a critic out.

Now with that screed out of the way, on to the game. I’m afraid that I don’t have many good things to say about it either. Zero One (or 01, as it likes to nickname itself) is an extremely silly game, cliche-hampered, lacking any sort of logical story, bug-ridden, and incomplete. If you were setting out to write a totally hackneyed IF game, what would be the starting location and situation of the PC? If you said “stricken with amnesia and locked in a cell,” you are today’s winner! That’s exactly the story with the PC of 01, but unlike, say, Square Circle, which builds an honest-to-gosh story around this situation, this game is totally uninterested in revealing the PC’s actual identity or the circumstances the led up to his incarceration.

Oh, it makes a couple of halfhearted gestures at explanation, but these are totally insufficient to actually build any real understanding, and besides, they’re totally overwhelmed by the weight of random events and situations. A good example is the kitchen drawer, in which you’ll find a dead fish along with the cutlery. Why? Aw, who cares? What bits of information do exist are burdened by a juvenile fascination with weapons and gore, like the pool of blood and splattered head that awaits the PC just outside his cell, or like this, after you find a handgun (complete with make and model info) and magazine of ammo:

> put magazine in beretta
Lock and Load!

Bro-THER. Throw in a little queer-baiting, and you’ve got a game that just screams “12-year-old male.”

The game is good for a few unintentional laughs, though, due both to its harebrained shadow of a plot and to its buggy implementation. A great example is the doors to the prison, which are secured by a padlock that, to the game’s own surprise, can be unlocked with the first key you find (the bracketed comment is from me):

> unlock green door
There is a padlock on the door and you don't have a key. [Actually, I
do.]

> unlock green door with key
The green door is now unlocked.

Sadly, this change just makes the game channel one of the maze rooms from Zork — going through the door will just loop you back into the current room. Oh, and I also managed to crash the interpreter entirely, though I’m not sure whether this was 01‘s fault or Alan’s. I write comments at the prompt as I go through the game, and after a particularly long line, the interpreter itself just up and shut down, much to my surprise. Luckily, I’d just saved my game, so I didn’t lose much.

Actually, I wouldn’t have lost much if I had just stopped right there and never opened the game again. The ending text insists that “ZERO ONE is not yet finished… Expect a return!”, but given the quality level of this game, that seems more like a threat than a promise.

Rating: 2.8

Splashdown by Paul J. Furio [Comp04]

IFDB page: Splashdown
Final placement: 8th place (of 36) in the 2004 Interactive Fiction Competition

Apparently, malfunctioning slower-than-light starships with passengers in cryogenic sleep have replaced isolated scientific complexes as the go-to comp game setting this year. Of course, in fairness, we did have a scientific complex in All Things Devours, though it wasn’t underground or in Antarctica or anything. Oh, and I guess my own entry could be considered an “isolated scientific complex” game, sort of. Still, the malfunctioning starships are making a strong showing this year. Happily, Splashdown is leagues ahead of its competitor, Getting Back To Sleep, though they both feel rather too much like Planetfall knockoffs.

One refreshing difference is that at least Splashdown acknowledges its debt to Steve Meretzky, not to mention the fact that its implementation is (pardon the pun) light-years ahead of GBTS‘s creaky homebrew. The game even provides a nifty PDF feelie that rivals Infocom in quality. Nevertheless, there are times when Splashdown feels just too derivative, especially when it introduces a cute little robot companion who follows the PC around, spouting random funny dialogue and helping out with the occasional puzzle. Sound familiar?

Besides that, there’s the fact that the crux of the story really doesn’t make much sense. Apparently, the ship is heading off to colonize a distant planet, but something goes wrong, so its computer picks a random colonist to reanimate, in hopes that this person can address the problem. This random colonist is the PC, natch, and if it were me, the colonists would be doomed, because I certainly didn’t win the game my first time through. Isn’t this an unbelievably dumb disaster plan? Why in the world wouldn’t there be a designated person to reanimate in situations like this? Putting the fate of 500 (or maybe 300, depending on whether you believe the game or the hint files) people in the hands of some randomly selected dude isn’t a strategy I can see even the most dunderheaded government or corporation assaying.

That complaint aside, Splashdown presents an entertaining story and a believable setting. I particularly enjoyed figuring out the reason why the ship malfunctioned, a comic situation worthy of Meretzky. There are a nice variety of puzzles, and they’re blended pretty seamlessly into the story, which I greatly appreciate. Somehow, though, many of these puzzles felt rather counterintuitive to me. Looking back at my transcripts, I see a few different root causes for this problem. One issue is that the game’s description of certain objects doesn’t really jibe with my understanding of how those objects ought to work in real life. In particular, I don’t expect a spigot to do anything useful unless I turn a faucet or turn on a pump or something, and if I do so, I expect that spigot to start spouting whether or not it has anything attached to it. These assumptions played me false as I was flailing around Splashdown‘s ship, trying to figure out anything to do that would make any sense at all.

Another issue is that I had the sense that I was missing just a little bit of documentation. In one or two of the final puzzles, I only knew what I needed to do because the hints told me so, not because of anything in the game that gave me the clue. This may be down to a case of me being slow on the uptake, or it may be that the game makes a few too many assumptions about how familiar players are likely to be with its setting. Finally, in some situations, too few verb forms are implemented. Particularly on one of the initial puzzles, I grasped the concept of what needed to be done, and tried a few different ways of expressing it, only to be rebuffed each time. Consequently, when I saw the solution in the hints, I felt annoyed rather than relieved.

Actually, the lack of synonyms and alternate verbs plagued me outside of the puzzles as well. For instance, there are cryotubes in the game that can’t be called “tubes.” There is no good reason not to provide those sorts of synonyms. In addition, one section of the game requires a lot of talking to the computer, using syntax along the lines of COMPUTER, DISPLAY HELP SCREEN. You can’t call the computer COMP or anything like that, and you can’t just say, for example, DISPLAY HELP, or better yet, HELP. Given the number of times I had to type out commands like this, I was mighty annoyed at the lack of abbreviations after a while.

On the other hand, the implementation is almost comically rich in a couple of areas, particularly the cryotubes themselves. There are 125 of these implemented, each with its own personalized nameplate. I was so gobsmacked at this that I had to examine each one, and was rewarded with occasional jokes and geeky insider references. And so the ship’s systems gradually failed as I went around autistically reading nameplates, but I loved it. Despite the occasional moment of breathtaking implementation, though, Splashdown feels like it’s not quite out of beta yet. There are a considerable number of typos, and sloppy formatting is rampant, especially when it comes to the robot companion’s random dialogue. In addition, I encountered a few minor bugs and glitches here and there. I hope very much that the author takes reviews and feedback to heart and releases a post-comp edition of this game. With some polish, I think it could be a really fun Infocom-style ride.

Rating: 8.0

Ruined Robots by Nicholas, Natasha, and Gregory Dudek [Comp04]

IFDB page: Ruined Robots
Final placement: 34th place (of 36) in the 2004 Interactive Fiction Competition

Okay, I’m not sure how to review this game without sounding like an ogre. Per the credits and a quick perusal of the web page, it looks like Ruined Robots was written by two kids and one adult, so I hardly want to come off like a big bully, but there’s just no getting around the fact that this game is terrible. Being charitable, it feels like a child’s drawing that’s been taken off the refrigerator and entered into a competition filled with talented painters and sketch artists. In one context, it’s something to be proud of, but in the other, it can’t be anything but bad. I mean, there are a couple of cool features — some of the ambient sounds are nice, and the little “common commands” toolbar could be useful to players with a different style from mine. But the crux of the game — the writing and the coding — is just really poor.

Detailed analysis of everything wrong with Ruined Robots would be belaboring the point, and though point-belaboring is one of my hobbies, I’ll try to steer clear of it this time, and instead just mention a few ways in which this game could be improved:

  • Get rid of the hunger timer. It is pointless. (By the way, what is it with TADS and hunger puzzles? Are they the default in the TADS library or something?)
  • Figure out what’s causing the freaky line of double-strikethrough letter Ys next to all the room titles, and fix that.
  • It’s = it is. Its = belonging to it.
  • Proofread in general. Pretend you’re turning it in for school and will get marked down for every English error or something.
  • Get your game beta-tested. If your testers can’t figure out the puzzles, decipher the prose, or finish the game without your help, fix whatever’s confusing them.
  • Try to have the puzzle-solving actions make sense. You can’t expect players to perform some random action you’ve given them no reason to try.
  • Speaking of randomness, cut it out with the million-and-one random objects that break mimesis and serve no purpose. If there’s a snowman on the lawn in the middle of spring, for instance, there’d better be a good and interesting reason for that. I don’t want to hear, “The snowman isn’t important.”

Finally, if you’re going to provide a walkthrough, please make sure it actually works to solve your game with. Otherwise, you’ll have judges who decide that your game is totally unplayable and deserves a 1.

Rating: 1.0

Order by John Evans [Comp04]

IFDB page: Order
Final placement: 24th place (of 36) in the 2004 Interactive Fiction Competition

At this point, I feel like I could really save myself a lot of time and energy by just writing up a template for reviews of John Evans games. The basic gist of the template would be “This game has some cool ideas and a lot of potential, but it’s not really finished and it seems untested, so it sucks.” Then I could just fill in the blanks with some specifics about the game, and be done. Evans has submitted games to the last five IF competitions, and they’ve all fit this mold, so why shouldn’t I just keep writing the same review over and over again? I kinda have to.

I mean, I guess they’re improving. The first couple (Castle Amnos and Elements) were way too big for the competition, and that problem got corrected. Unfortunately, it got corrected by switching to total unfinishability due to scores of heinous bugs. I gave a rating of 1.0 to Evans’ last couple of games because they were just totally incomplete. Order isn’t as bad as that. It’s the right size for the comp, and it is finishable. But it is still not finished. See, if you’ve got a command called BUGS that lists out the bugs in the game, your game’s not finished. Actually, one of the funnier parts of Order is that even the BUGS list is buggy, as some of the things listed actually do work (though plenty don’t.) Also, if a crucial puzzle in your game rests on a piece of scenery that you don’t mention anywhere, your game’s not finished. A tester would catch that. That’s, y’know, what testers are for. Use them.

Actually, beta-testing seems particularly critical for a game like Order, because the game revolves around coming up with actions quite spontaneously, and the better sample you have of what actions people are going to come up with, the better you’ll be able to implement them. (Here’s where I start filling in the “cool ideas” part of the template.) I love the basic concept of this game — you’re some kind of magical spirit, and you’ve been set a series of tasks by your summoner. So far, still pretty bland, and strongly reminiscent of J.D. Berry’s The Djinni Chronicles. But the nifty twist that Order puts on things is that you have a CREATE verb at your disposal, and you must use it to solve every puzzle.

So, for instance, you find yourself faced with a locked door, and must CREATE a key. Done right, this could be an amazingly powerful device, giving the game a feeling of almost limitless possibility. And indeed, there are times when Order feels like that. Of course, there are way more times that it’s disappointing, and not just because I was trying crazy commands like CREATE PHASER and CREATE WETSUIT. Lots of much more reasonable attempts aren’t implemented, and I can’t help but think that some beta-testing would have greatly improved the game’s range of options. Still, there are multiple solutions to each puzzle available from creating various objects, and that aspect of the game is really fun. Too bad the other parts are such a drag.

Oh, there is one more good part — the hint system. These hints are nicely implemented, InvisiClues style, and it’s a good thing too, because nobody is finishing this game without the hints. As I alluded earlier, there’s a critical puzzle in the game’s midsection that is simply not solvable without hints, because its main components aren’t mentioned anywhere. Remember Bio, from Comp03, the game that starts out with gas seeping into your room and you have to get a gas mask out of an armoire, except the only place that says anything about an armoire is the walkthrough? Pretty tough puzzle, right? Well, Order takes inspiration from it with a puzzle where you must perform an action on an orb encased in a steeple. Problem is, nowhere in any room or object description are the orb and steeple mentioned. There is a dome mentioned, but like many of the scenery objects in room descriptions, it’s not implemented. Actually, even for some of the objects that are implemented, the descriptions aren’t exactly superb:

>i
You are carrying a set of white robes (being worn).

>x robes
A set of white robes.

Ooo, spellbinding. So okay, I’ve almost got my template ready. The only thing left is to figure out how to wind it up. I’m leaning towards a plea for the future: please help your cool ideas reach their fullest potential by finishing and testing your games! Please! I may have to edit that part out though, as its refusal becomes more and more certain.

Rating: 4.9

Redeye by John Pitchers [Comp04]

IFDB page: Redeye
Final placement: 28th place (of 36) in the 2004 Interactive Fiction Competition

Okay, before I launch in, let me just say this: Redeye seems like a well-intentioned game. It’s seriously burdened with problems, but it makes a reasonable attempt at story, writing, puzzles, and so forth. Moreover, it does one thing that I thought was really cool, which is to take advantage of my adventure game sensibilities for a crucial scene. The PC witnesses a shooting, and on examining the victim’s body, finds a gun. Because of my ingrained “pick everything up” IF habits, I added the gun to my inventory, and then went outside… to be greeted by the police, shouting at me to drop the gun. Oh yeah. Guess it’s not a good idea to pick up a gun right after a murder. Looks kinda suspicious. Of course, as it turns out, making that mistake is the only way the story can continue, which is not so good. Still, things like this help me believe that this game can be improved, and that authors can learn from its mistakes for the sake of future games.

In that spirit, then, allow me a few suggestions for how to make IF better than Redeye. First of all, put some thought into presentation. The first thing this game did was assault me with an “angry fruit salad” melange, blocks of text in no less than eight different colors. I found all these colors pointless and distracting — they add nothing to the game. Worse than this, the game set the background to black but somehow failed to set the color for the foreground text, and consequently I couldn’t see what I was typing, since it was black-on-black. Even the most minimal testing would have found this problem, which leads me to believe that the colors were changed at the last minute.

Presentation is just the tip of the iceberg, though. Another pointer is to proofread your grammar and spelling, and have some people who are good with words look at it, too. In particular, learn how apostrophes work. “The suns rays” doesn’t mean “the rays of the sun”, it refers to multiple suns and multiple rays. “Tell her your OK” doesn’t mean “tell her that you are okay”, it means “tell your okay to her.” “Your” means “belonging to you.” “You’re” is short for “you are.” Learn grammar or your writing will suck. While you’re at it, learn the difference between sentence fragments and complete sentences. Avoid the former, and cleave to the latter. Next: pick a voice and stick with it. Second-person voice is “you”, and first-person voice is “I.” Do not mix them, like so:

You have absolutely no idea where you are or how you got here…

Hmmm. Where is Arthur? I hope nothing has happened to him…

You open your eyes. Crikey. Where in hell am I? It looks like the
middle of the bloody desert!!

Crazy seesawing between voices like that makes no sense and is very annoying to read.

Well, that gets us through the intro. Now some tips for the game. First of all: please, please, PLEASE no hunger puzzles. They are just lame. They are especially, excruciatingly, super-duper lame when there is NO FOOD IN THE GAME! Secondly, please allow for reasonable actions. Plots on rails (and this one most certainly is one of those) are fine, but you have to find a good reason to turn down the player’s actions, not just fail to implement them. If the PC witnesses a murder, many players’ first instinct will be to CALL THE POLICE. “I don’t know the word ‘call'” is not a sufficient response to this.

To avoid problems like this, have your game tested by a number of people and enhance your code in recognition of the actions they attempt. I’d have thought that one was bloody obvious, given that the comp organizer sends out emails with specific instructions to do so, but perhaps not. Also, please implement descriptions for nouns mentioned in room and object descriptions. If I’m told I can see a highway, a store, and a hotel, I want to be able to examine them. I do not want to be told that the game doesn’t know those words. (Yes, that’s a lot of work. Writing good games is a lot of work.)

In addition, provide all the synonyms you use in your descriptions. For instance, if there’s a group of bikers alternately described as a “horde” and a “crowd”, X HORDE, X CROWD, and X BIKERS should all result in the same response. Implementing only one of them is… bad. Also, this thing:

>x urinal
Which urinal do you mean, the toilet, or the toilet?

Please avoid that. On another note, recognize when you are cueing the player and respond accordingly. For instance:

>x motorcycles
The motorcycles are predominantly Harley Davidsons. Low, sleek and
highly modified. I wouldn't touch them if I were you.

>touch motorcycles
Touching the harleys doesn't seem to have any effect.

That is anticlimactic and disappointing. Again, at least one tester probably would have caught this. Okay, there are certainly plenty more changes that would significantly improve the game, but that’s enough for now. Redeye is quite a poor game, but it’s quite a good cautionary example. If you play it, you’ll probably come up with your own list of what not to do in creating IF.

Rating: 4.0