Gamebooks, interactive fiction and hypertext


Gamebooks were developed roughly in parallel with text adventures, and became fairly popular during the 1980es and 1990es; I remember reading quite a few of them before getting my first computer. (I would have continued, but lost access to my source.)

One would have expected the genre to migrate naturally to computers and especially the Web. But when I got on the Internet, there was little to be found apart from whimsical, quickly abandoned addventures and the high-brow experimental stuff typically promoted by the Electronic Literature Organization. And then there was the modern interactive fiction community, which I discovered around 2004. But the IF community had little love lost for the genre.

Now, that may seem like an odd claim when there are things like Andrew Plotkin's famous The Space Under the Window or the LOTECH Comp, which ran (in the small) for four years. More recently, acclaimed titles such as Whom The Telling Changed and Blue Lacuna added keyword-based navigation on top of the usual parser, while Walker & Silhouette and Starborn did away with traditional commands entirely 1. But generally, CYOA works (choose your own adventure, after the famous book series), tend to be frowned upon. And frankly, I find the usual reasons why unconvincing.

See, the general consensus appears to be that the appeal of interactive fiction comes at least partly from a sense of exploration. You don't know what's there until you look, and you don't know what's possible until you try. But that can also be a problem: when you can try absolutely anything, what do you try? What is valid? What is relevant? A skilled author will gently steer the player in the right direction, but that's tricky at best, as one person's obvious is another's "I would never have thought of that". And the genre conventions -- which greatly help a skilled player -- are baffling to newcomers.

Worse, to dismiss games where all the options are laid out explicitly is to dismiss the vast majority of games in existence, beginning with chess. Knowing your options doesn't mean you know what will happen further down the road... or even that the road is clear at all. Machinery-based puzzles, anyone?

As an aside, menu-based conversation systems are similarly despised, even though the ask/tell system boils down to the same thing... except you have to guess at the available options. And we all know how well that goes most of the time.

I've also been told that it's not really interactive fiction unless it has a fully developed world model, regardless of user interface. But consider an example: say at one point in a keyword-driven game you find a jacket. You can type "jacket" to examine and take it... then you never get another chance to refer to it again. Does it still matter if it's fully modeled? Might as well be just a binary flag.

Where does hypertext come into the picture? Well, think about it. What difference is there between choosing a numbered option from a menu and typing one of a few highlighted keywords? It's a purely cosmetic distinction. And you know what else is equivalent? Following a link on a web page. It's no surprise, then, that people have tried creating gamebook-like stories on the Web, using wikis at first. But wikis, just like normal Web pages, are stateless. For a while, I even pondered augmenting a wiki engine with a scripting language, to allow for dynamic pages. Then I hit upon the right idea. Which led me to discover not one, not two, but three competing systems, all in active development. (And the IFWiki lists even more.) There is interest in this sub-genre. So why the generally negative attitude?

My guess is, beginning authors may be tempted to go for a CYOA piece instead of a parser-driven text adventure on the idea that it's easier. And because the former are relatively uncommon, a rush job will be all the more noticeable. Also, with little in the way of mechanics to distract the player's attention, the writing has to carry the burden, even more so than in regular IF.

There are subtler issues, too. Simply deciding what choices to offer the player can be difficult, and you can't just require the same kind of low-level manipulation -- in CYOA you need a higher level of abstraction. I seem to remember one game where you had to solve a complex puzzle with some scales and weights... using only numbered options. Suffice to say, I gave up quickly. Then again, the converse is just as possible: I still think Andrea Rezzonico's Hello Sword should have been menu-driven, the way it restricted your actions at every turn.

To conclude, player interaction in IF is far from a solved problem. Emily Short's rant about the parser still echoes a year later 2. But the same Emily Short will tell you that "it's not IF unless it's text in, text out". Perhaps, as Jimmy Maher points out, "What we need is something with the ease of use of CYOA and the flexibility of the parser." Trouble is, it would take some serious experimenting to find that something, and as far as I can tell the community rewards conformance instead, in ways as subtle as having a "Best Use of Medium" category in the XYZZY Awards and none for innovation. (EDIT: turns out, now there is one. Thanks, Emily!) But maybe I'm reading too much into that. After all, this spring there has been such a thing as the IF Demo Fair.

In the mean time, IF is moving to the browser, browsers are increasingly mobile, and modern mobile devices tend to have touchscreens. Which aren't exactly good for extended typing. 2 + 2 = 4? It remains to be seen.

P.S. This article ended up sounding like it’s about menus versus parsers. The issue goes deeper than that, but I didn’t want to spread myself thin. Graphics in text adventures suck… until Future Boy comes along. Combat in text adventure sucks... until King of Shreds and Patches does it right. I’ll admit that we applaud successful experiments, but we’re way too harsh with failed ones, though they are just as important. And if an experimental work falls flat on its face, it may not even be because of the experiment, but general poor quality. But guess which cause is normally blamed.


1. Correction: W&S does retain support for full commands
2. Heck, it still echoes seven years later.