Adam Cadre’s Radio K podcast, continuing its coverage of IF games from the early to mid 2000s, now covers Slouching Towards Bedlam, Gourmet, and The Dreamhold.
Part of Adam’s discussion of Slouching riffs on what he says was my view at the time, that this was one of the first IF games to seriously address moral choice. He objects to that view because the player is very likely to explore all the possible endings, and therefore it’s unlikely that they’ll feel much weight in their decision, and also because he doesn’t consider the final question morally all that interesting.
It’s possible I did characterize Slouching that way at some point in a newsgroup discussion – I don’t now recall precisely. It’s not really quite what I would say now, though — and actually it’s not exactly what I say in my contemporary review.
I bring this up not to nitpick Adam’s generally excellent podcast, but because thinking about Slouching Towards Bedlam from the perspective of the current IF scene sent me off on something of a tangent of further thoughts.
So with apologies to Adam, here’s what I would say about it now: Slouching is an early, and still relatively rare, example of parser IF that makes diegetic outcomes depend on a complex set of world model machinations on the part of the player. First you have to figure out how the world model affects the story, and then you have to use that information to bring about the ending you want. You’re likely to luck into several endings you don’t want at the same time.
This is not a desperately hard puzzle, but it does work rather differently from most of the branching narrative endings that existed in IF up to that point, which often (as in Metamorphoses, or Masquerade, or Plundered Hearts) rely on a single endgame choice. Slouching‘s structure is instead is more like what we see in much more recent work like Midnight. Swordfight. or Map. I tried to do something a bit similar with Floatpoint, but that was definitely directly inspired by Slouching. Or, for that matter, you can find this in visual novels – Black Closet is very much the sort of game that requires significant learning before you can steer the story in the direction you want.
I would say that the amount and type of control over the story we have in Map is more satisfying than in Slouching – there are harder choices to make, and they involved characterized humans rather than abstract Lovecraftian evils. But Slouching was one of the first times I saw a parser IF game do this in a meaningful way.
In general, despite how much people talk about how parser IF has more agency than CYOA, in practice, most older parser IF has offered the player virtually no agency that affected plot beats; and diegetic agency is still much rarer in parser IF than in most choice based games, gamebooks, or dating sim-style visual novels. When I wrote Metamorphoses in 2000, having multiple endings was not novel exactly, but still enough of a surprise feature that it got quite a bit of discussion. Counterfeit Monkey also has multiple endings, but by the time it rolled around, multiple endings in parser IF had become far too common to be considered worthy of explicit note.
But that’s an all-linear-until-the-end structure. Parser IF that does even a modest amount of branch and bottleneck is still vanishingly rare. Yeah, you might get a choice of which rooms to traverse first or which puzzles to solve in which order, and maybe if you’re lucky there will be variant outcomes depending on how thoroughly you solved things – but a significant story variation in the midgame? Highly unusual. The reason that Sam Ashwell-style charts and narrative typologies have only made their way into mainstream IF community discussion in the last few years isn’t that parser IF plots are too complicated to chart that way. It’s because they’re usually too simple for anyone to bother.
Likewise, Adam questions the importance of a moral choice if the player is able to SAVE or UNDO and immediately experience some alternatives.
It’s true that some IF blocks off trivial replay partly in order to make those decisions more powerful: most of your choices in Fallen London you can’t undo, though occasionally there were ways to earn or buy a reset. And in Choice of Games pieces, you can’t back up, so if you want to look at another ending, you have to play through the whole thing again. Matter of the Great Red Dragon instructs you, more or less on your honor, to play only once. So there are various ways that people try to get around this.
It’s also true that “you can reset this moral choice and replay!” would have struck me as a pretty significant criticism back in 2003. I might then have said Slouching does a better job than most because, again, to execute a change in outcome, you have to replay a lot of the game to get there, rather than just restoring a state near the endgame. But of course I did play to all five endings at the time, and I would consider playing to all the endings to be the appropriate way to play Slouching.
Over the last dozen years, though, I’ve become less committed to the idea that a moral choice in a game can work only if the player is forced to commit to the outcome of that choice. In Map, it wasn’t having to stick with a moral decision that made it work for me, but the complexity and ambiguity of the choice and the number of factors I had to weigh in assembling an outcome. In Floatpoint, I tried to set things up so that the very act of trying to bring about a particular ending would mechanically force you to encounter more of the reasons why that choice might be a complicated one. (I don’t know how well that worked for people, but it was a consideration at the time.)
Or again, at the very end of Black Closet, I ran into a key choice and it was pretty obvious I would have to replay for hours if I wanted to get back to that choice again. That investment did make me care about the choice somewhat, but it was much more important that Black Closet sold me the choice narratively: here was an opportunity to express a final reflection about who my character was as a person. As it happens, in the event, I knew absolutely which of those three things my particular character would be likely to do: not an agonizing decision, but a solid one instead.
In other words, you can make a game with interesting interactive morality without necessarily pinning the player to a chair. In fact, often the games that have occasioned the most thought for me are not the ones that forced me to experience just one ending. I think I wouldn’t now say that Slouching is a morally amazing piece of work – I feel pretty confident about the right course of action and I don’t think there’s a strong counterargument to be made – but I would say that it showed some of the tools for that kind of work. That was pretty important at the time. And Slouching is still fun, and still doing something comparatively uncommon.
Anyway, there’s a lot else in the Radio K podcast than just the remarks that sent me on this tangent. Adam and Jess Haskins also go on a riff about the value of editors for interactive fiction, and I totally agree with that; in fact I’ve got most of a blog post written on the topic. Adam also talks about the challenges of writing good IF for beginners, IF prose writing, and various other goodies.
10 thoughts on “Morality and Diegetic Agency in IF (via Radio K on Slouching Towards Bedlam)”
I think there’s more to the problem with branch-at-the-end without much effort other than the presence of UNDO. I thought The Secret Vaults of Kas the Betrayer from IFComp 2014 had an annoying end choice that was equally trivialized by the ability to just hit the button and see what the other ending would do, whereas with Switcheroo from IFComp 2015 even if the ending choice had an easy-undo I would not have used it because I felt like there was a “right choice” for my story and there was no use checking the other one.
Put more simply, Secret Vaults didn’t have enough of the character built up for me to care which ending I got, whereas Switcheroo did.
I didn’t explore all the endings of Map, either; I went through a couple and felt like I could settle on one because that was the “right” story.
Slouching feels more like an “all the endings exist in a quantum state simultaneously” type scenario, although I haven’t seen all the endings.
Also, close your em mark after “Black Closet”, your italics are running.
Doh, fixed. (WordPress has once again introduced a “new” “improved” edit mode that is worse than the old one, and breaks up my standard practice for these things…)
I hate the new mode with the passion of a thousand burning suns. I’d be good if I was just writing a narrative book but when I have to flip back and forth with links and images it is just terrible.
Also, what I do is start the post, save a draft, switch to myblog.wordpress.com/wp-admin, click on the draft name, and then I go back to old comfortable mode. I don’t know if there’s a faster way to switch to that.
When I’ve done user experiments with frictionless undo with CoG games, players can’t seem to resist mowing the lawn. A, back. B, back. C, back. D, back. Hmm. Let’s stick with C. OK, A, back. B, back…
And then they have the gall to tell me how boring it is to try every option!
My feeling is that there is usually a “right amount” of friction for undoing narrative choices, and it’s usually non-zero, but having too little friction doesn’t necessarily totally undermine the choice.
When players are curious to learn about the full possibility space, reducing that friction actually increases the fun. (Hence a lot of visual novels have a “fast forward” mode where you skip content you’ve already seen.) A number of folks on our forum are fond of reading the game source code after playing a few times, to get the full experience that way.
You know you’ve written a good choice when it still feels meaningful even when the player can see the game source.
Possibility/Problem space. It’s not so much the endings. It’s the progression of the model. In parser, in particular, branching narrative causes an exponential growth in development, and, in particular, testing. Map was supposed to be considerably less linear than it ended up, but I could see no way of modelling an open-ended parser world that reflected non-directed narrative choices by the reader. The choices are Yes and No. Even introducing a ‘maybe’ required more development-hours than atoms in the universe. I exaggerate.
I’m struggling to put this into words.
When I read a really wonderful ChoiceScript game like “Choice of Robots” I simply feel like I’ve read a book that was magically tailor-made for me, at that age, in that year, and even in that particular moment. I can see and feel that things could have turned out differently if I’d made different choices, and there’s an eerily lifelike mixture of regret and peace there. After that first read, I set aside games for a long time, like favourite novels – all the best novels change between readings, too. When I read “Synesthesia Factory” it was an intense personal experience. I’m literally afraid to read it again, in case I see the wires holding it together and all the magic falls away. When I read a print author I utterly adore, I’m afraid to read another book of theirs for very similar reasons. (Not sure if any of this is relevant, of course. Um… I was thinking about multiple endings, that’s right!)
It’s worth mentioning another game that has touched on the “reset moral choice and replay” recently is Undertale — while it’s not interactive fiction, it’s another game with important choices, and (SPOILERS) the game records meta-flags separately between resets so it can call you out if you try to reset and undo major decisions. In particular, one ending has consequences that apply permanently through all future playthroughs (unless you perform registry-hacks to undo it, of course.)