This article is based on a couple of posts I wrote in response to Jim Aikin on rec.arts.int-fiction. This material has been edited together and in places rewritten for its current form.
The Function of Prose in IF
Authors and players of IF, wondering why interactive fiction is not more widely played, sometimes blame the prose quality of IF — arguing that it is not enough like a book, that it is too repetitive, that it doesn’t sound professional. It is certainly true that (because there are no barriers to releasing a work of IF) there is a great deal of poorly-written interactive fiction available on the IF archive. There is also well-written IF. Some of it is relatively book-like; some of it isn’t, but still presents text that — in an interactive context — is engaging, effective, and even beautiful.
So perhaps it is worth taking more of a look at what sorts of response IF authors receive from outside the community, and considering the essential qualities of good prose in IF (which may prove to be different from the essential qualities of good prose in a novel, an academic article, or a cookbook).
A major issue seems to be the work’s interactive accessibility, by which I mean how well the game communicates what the player can do, and how completely it responds to commands both serious and frivolous. I have found it both telling and amusing to read Victor Gijsbers’ indifferent review of Suveh Nux on IFDB, together with the strongly positive response the game has gotten from semi-outsiders who came to the work through the casual games site JayIsGames, where some commenters clearly follow IF but many just as clearly do
Of the two, it’s pretty obvious (at least to me) that the outsiders are the ones less concerned with questions of text aesthetics or story, and more concerned with clarity and operability. Clarity is rarely produced by technically incompetent prose, but competent or even inspired prose is not guaranteed to be easy to interact with.
By contrast, negative points made about IF in the wider world (and here I draw from threads on SlashDot about the yearly comp, and discussions on various gaming boards) often focus on the parser, not the writing, as the critical weakness of IF. It is the refusal to recognize nouns mentioned in the text, the insistence on very particular player behavior, the guess-the-verb nonsense, that turns off the most people, as I observe it. Remedying this might partly be about smartening up the parser, and I think Aaron Reed is doing good work in this direction. But some improvements can also be made by extensive polish, anticipating and responding to a larger than usual range of experimental behavior. People on JayIsGames specifically commented that Suveh Nux was more playable than average because of this richness of response; I’ve gotten similar feedback from non-regulars of IF to the effect that they liked Floatpoint, despite a general dislike of the genre, in part because it seemed better than average at understanding what they wanted to type. I don’t claim to have a complete lock on what people outside the community do and do not like about IF, of course, and objections vary from person to person as well.
Clearly, artificiality is an issue, since newcomers do raise the question “but why compass directions?” semi-frequently. This is a genre convention. It’s one I’ve defended because I think it genuinely contributes something in some games, enough to make it worth learning to deal with directions. To my mind this is no more ridiculous than asking musical-goers to learn and accept that a song sung at top volume can nonetheless represent the singer’s inner monologue and be completely unnoticed by the other characters standing around.
All the same, I hear the issue of playability cited much more often than the question of literary competence; and even criticism from authors of literary hypertext has tended to be more about why IF’s world models and structure are unproductive of rich characters and narrative than about problems in the readability of the output.
This is not at all an argument that prose style should be neglected. On the contrary, there’s much art to writing prose for IF, which involves constantly giving the player hints and directions without seeming to do so, building mood through specific physical detail, measuring the amount of exposition out into manageable pieces, and avoiding any commitments in prose that you’re not willing to honor in the world model.
I don’t think this skill can be acquired entirely through practice in static writing, though such practice does not hurt. But what about the rest of the skill? Where does it come from and how can we evaluate examples of it, if we are not evaluating IF prose simply by comparison with the styles of static authors?
I talked above about responsiveness, and I think that’s critical. There is one aspect of improving responsiveness that is about the writing: namely, that all text in interactive fiction is at least potentially a direction to the player, and the author’s willingness to direct the player is a critical part of making a game accessible. By this, I mean that text like
Something glints at the bottom of the lake.
Through the fragile surface of the glass ball, you make out the wavy
outlines of a key.
The princess is laid out on the antique sleigh-bed, deep in enchanted
sleep. As you look at her, she stirs in her dream and murmurs
something that sounds like your name.
directs the player to examine, to smash, to kiss. (Nick Montfort’s dissertation refers to this as the Suggester role of IF text.) Sometimes that direction depends on cliché (about how sleeping princesses should be treated, e.g.); sometimes it just consists of a willingness to foreground in description the functional aspects of an object, drawing the player’s attention gently toward those features of something that are most likely to be useful. We can still allow him to get a fuller sensory description with supplementary commands such as EXAMINE, READ, TASTE, SMELL, etc., in cases where those commands are not otherwise prompted by the text.
If we’re careful about directing player attention and always make it clear what items in a room are important and which items are not so important, I think the frustration of having to examine everything thoroughly decreases, so that items having a further description is no more problematic than things having a custom reaction to being tasted or smelled.
Similarly, we should be prepared — especially when the game’s direction is very heavy-handed — to cope with the player’s rebellion. Being directed, or given explicit options, is fun up to a point, and then it begins to feel like having one’s strings pulled; so players may react by trying just the opposite of what the game seems to be clueing. A game is more likely to get the player’s sympathy and respect in the long run if it has a response not only for the directed action but its antithesis.
This may all seem like an upsetting set of claims, since I am saying that quality in IF prose is slightly different from quality in static prose. I don’t think that’s the same as saying that IF prose cannot emulate the lyricism or the evocative power of writing in other media. It also doesn’t mean that IF writing should violate standard rules of narrative presentation, such as mentioning what is relevant and omitting what is not. The very important magical jewel of Fozbar absolutely should catch our attention if we happen to have it in the pommel of our sword.
Perhaps a given passage of writing in IF will be more work, requiring more thought and revision, than a similar passage of writing even in carefully-wrought static fiction. But I think that’s fair.
I may also seem to be placing the base functional concerns of the game over the more ethereal aesthetic claims of the prose. To that criticism, I can only point out that — however often we talk this way — the game and the text, or the game and the story are not really two separate creations sutured together. The prose aspects should make the game easier and more engaging to play, yes, but the interactive aspects should also enhance the story-telling and lend it strength. The compensation for the compactness necessary in a room description is that the very functionality of IF tends to make every object seem more important than it might on a page.
I am arguing for a kind of writing that avoids detail for detail’s sake. It’s certainly true that details set tone, reveal background, and reflect information about the viewpoint character that notices them — all good standard writing workshop advice, and IF makes good use of all those techniques, both through the visual and through other senses and memory. (Varicella’s descriptions rely heavily on memory; Aayela’s on senses other than sight; many of Robb Sherwin’s works on viewpoint character attitude; other examples abound.) But if we have no such purpose in mind, and the details have no bearing on game play, they are probably best omitted. This is a medium that rewards restraint.
If we truly have a great deal to say about a setting, then the best strategy is to reveal it bit by bit The immersive effect of rich detail works best, in this medium, if the player is allowed to discover some of the physical details himself, through EXAMINE, TOUCH, SMELL, etc.; having these richly layered creates far more of a You Are There experience than a room description heavy with detail and minimal follow-up descriptions.
This also need not prevent us from thinking about mood, atmosphere, memory, and the characterization of the narrator. Some of the most effective writers of mood create their effect not with a large number of common details (the flowers are red, the door is yellow, etc) but with a small number of very particular ones; and I think that that is especially true in IF. Words in interactive fiction individually carry more weight than they carry in static prose, if only because of the amount of attention we demand the player give to each one. We ask people to notice, to remember, to wonder about, in a way that static fiction authors do not. At the end of reading a short story, would you be able to sit down and write out a list of all the objects the author mentioned? It would be absurd — but I suspect you could come much closer to producing such a list for an IF game you’d just finished. In fact, for some games I can still visualize scenes and critical inventory objects years after playing, in some cases complete with tactile impressions and sounds and even smells. So there is, at least potentially, a lot more moodiness available in a given thing that you mention in IF, just because of how much the player gets involved with it.
I’ve often noticed how particular and detailed and physical are the descriptions in P. D. James’ mystery novels: she likes to spend quite a lot of time describing people’s clothes, possessions, and living spaces as a way of revealing their personalities. But, for all their apparent similarity to IF descriptions, I think I would find these to be overkill in an IF game. They’d need to be shortened and focused, because each sentence would do the work of three or four sentences in the static prose version. In this respect IF is closer to poetry than to conventional prose: it is worth taking more time to select fewer words, because each one will be inspected through a jeweler’s loupe.
There is one trait of commercially successful prose that I do think we could use in more IF, and that is confidence. The first page of a novel that sells presents itself with utter assurance. It knows what it is going to be about, it sets its tone, and it directs the reader’s attention to something interesting. It does not dither. It does not ramble. It knows what kind of book it is going to be, and it lets the reader know as well.
This is true equally of good best-sellers and of bad. The prose can be leaden, the characters rice-paper thin, but there will still be something authoritative and fearless about the opening. Sometimes the author is authoritatively and fearlessly aping the clichés of his genre, rather than asserting a style or vision of his own, but this can still impart the impression of professionalism, and it reassures
(say) the devoted romance novel reader that she’s about to read just the sort of book she wants to read.
This kind of contract with the player is just as important in IF, or even more so. We rely on the player’s trust and confidence in the author; without it, the player won’t experiment as much, won’t have as much faith that puzzles are fair and can be solved, won’t be as sure that the game is finishable.
Unfortunately, it’s hard to characterize this quality, other than the censor’s dodge: I know it when I see it. But I think that, together with greater game responsiveness, is the improvement most likely to catch the attention of a wider audience.