Recently someone contacted me asking to link an article explaining the difference between “old” and “new” IF. I don’t know of any such article that’s remotely up to date. Besides, this is not as easy to answer as they might have hoped, even if you agree on what we’re counting as “old” for these purposes.
But here is a sort of periodization. I’m intentionally vague about dates because people disagree about the milestones. I’m leaving a lot out because this is meant to be a digestible overview. And it’s also worth pointing out that this is a perspective taken from within a particular community, so there were other communities (adult IF, non-English-speaking IF, gamebook/CYOA enthusiasts, visual novel readers, literary hypertext folks) whose histories could usefully be drawn alongside this.
First, a couple of general histories:
The old Inform Designer’s Manual has a brief history of IF through 2000. Let’s Tell a Story Together covers IF from the beginning through around 2005. Twisty Little Passages was also released in 2005 and covers about the same period. The history section of the IF Theory Reader contains accounts of French and Italian IF, as well as Duncan Stevens on the period 1994-2004, and Stephen Granade on the trend towards shorter works of interactive fiction over the course of the late 90s.
The Brief History of Interactive Fiction was an awesome (paper! full-sized for your wall!) timeline that ran through the year 2000; I don’t think there’s a great graphical version of it online, but the list of events included can be found here.
Ca. 2011 Nick Montfort and I wrote an account of IF communities that is arguably already way out of date: at the time, Twine was only barely on our radar. Meanwhile, ifwiki has a glossary that defines a number of terms that might be relevant to reading some of the linked reference work.
Pre-commercial. Text adventures exist, but they’re circulated for free and played on mainframes. Crowther and Woods’ Adventure is typically cited as the first text adventure, though recently we’ve learned that Wander might be a contender for the title. Dennis Jerz has published some excellent original research about Adventure. More generally, the Renga in Blue blog sometimes writes about games of this era. These games are characterized by limited and capricious parsers, very terse text (data storage was a serious issue at the time), and the expectation that the player will be willing to try many, many times in order to win.
Commercial era (very late 70s-very early 90s). Interactive fiction means text adventures, produced by Infocom, Scott Adams, and a number of other houses of various degrees of fame and success. Games tend as a rule to be exploratory and puzzle-focused, and frequently difficult or unfair in ways we have now largely grown away from. This era of IF is being very thoroughly and thoughtfully documented over at the Digital Antiquarian. Within that space, one could draw some additional development history — recognizing, for instance, that some of Infocom’s work (especially Trinity, A Mind Forever Voyaging, and Plundered Hearts) focused on story, theme and character in a way that prefigured generic and technical features of later interactive fiction.
At the same time, text adventures were also being written in other countries and other languages; Italian text adventures were particularly numerous, for instance, but distributed differently from English-language adventures.
When people talk about “golden age” IF, they sometimes mean IF of this period.
Early Hobbyist (early to mid-90s). Overlapping with the end of the commercial era, people began writing their own games with freeware (and sometimes shareware) tools, and distributing them through Usenet and elsewhere. Some of the very earliest games were themselves shareware, and they tended to replicate features of commercial IF, including (sometimes) copy protection features such as feelies that you got in exchange for registering your game. The development of TADS and Inform, and the growth of rec.arts.int-fiction, shaped a lot of this period.
When used ca. 2000 or later, the phrase “old-school IF” tends to refer to works belonging to the early hobbyist era or earlier. These works are generally parser-based (CYOA certainly was around in this period, but wasn’t generally considered IF at the time). Graham Nelson’s Bill of Player’s Rights and Craft of Adventure and Roger Giner-Sorolla’s Crimes Against Mimesis (also available in the IF Theory Reader) provided a critique of the failings of old-school IF: extreme difficulty and unfairness; the need to save and restore frequently or even replay large parts of a game; unpredictable sudden death; unrealistic puzzles at odds with their settings; a tendency to take place in completely abandoned locations and rely on tropes like amnesia, treasure-gathering, etc. Similar complaints were also leveled against the puzzles in classic graphical adventure games. Curses is possibly the apotheosis of old-school IF — grand, sprawling, ridiculous, full of incongruous set pieces, but already starting to show some design impulses towards more aesthetic unity.
“Middle school” (mid-to-late 90s). This is a term largely applied in retrospect, in distinction to New School, but it tends to refer to games that still have substantial geographies and lots of puzzles, but that have started to pay more attention to plot arc and setting. They might have at least some active NPCs with personalities, and they are more likely to feature well-defined protagonists. (Adam Cadre’s now-disowned I-0, with its sexy college girl protagonist, was considered a landmark in that area.) Typically middle school games show a higher level of polish than typical old-school IF, with more scenery objects per room, fewer empty or pointless rooms, better feedback on puzzle failures, etc. Christminster (1995) and Anchorhead (1998) are the pieces I most immediately think of as middle-school IF, though I would also include Jigsaw.
Edited to add: XYZZYNews no longer releases new content, but continues to archive many articles published in this period. They give a sense of the concerns animating people at the time.
“New school” (late 90s-early 2000s). Like the Pont Neuf, “new school” IF is not that new at this point: it tends to refer to puzzle-light but still parser-driven games. The traditional date to begin recognizing new school games is 1998, with the release of Photopia, though arguably Photopia is not puzzleless and other, earlier games actually were puzzleless. However, this period corresponds to a lot of theoretical discussion on the newsgroup and a trend towards more experimental games; the introduction of the IF Art Show, which provided a home for goalless, scoreless IF; and works like Aisle, Galatea, Rameses, The Baron, and a lot of other short pieces designed explicitly for the purpose of testing the boundaries of what IF could be. Victor Gijsbers experimented with IF’s ability to test the player’s moral beliefs; Slouching Towards Bedlam and a few others with narrative structural approaches that went beyond simple branching.
This period saw the original inception of the IF Theory Reader itself, though it wasn’t published until 10 or so years later, for reasons that are largely my fault. This is also probably the last time there was any real consensus about how to categorize IF. The labels I stick on subsequent periods are not necessarily universal.
Inflection point (ca 2005-2009). Several things happened in the mid-2000s. Inform 7 came out, which drew in some new authors and displeased some old ones. There was a certain amount of grumbling that this new tool was so easy to use it would result in an influx of garbage from unskilled users; this is a theme that recurs at several points in IF history.
The IF community moved off Usenet and onto bulletin boards, and with that change came some loss of connection with previous theoretical discussions (though many of these are archived and meticulously indexed at ifwiki). At the same time, a significant part of the former IF community aged out of it: they had kids or their job responsibilities ramped up, and they stopped hanging around so much. The demographics may also explain why there was a bump in child-suitable IF in these years: Aotearoa, Mrs. Pepper’s Nasty Secret, The Lost Islands of Alabaz.
Indie gaming became more visible, with indie game review blogs and download sites, which meant that there was now a larger ecosystem IF could fit into, and IF started to get more visibility on sites such as Jay Is Games. (If you want a cute time capsule of what I erroneously thought this meant for IF, here’s my completely wrong 2008 post on hardcasual IF.)
(ETA:) The IF community also became more interested in outreach in this period — how to find more players, how to make more IF that would appeal to those players. This meant an increased emphasis on tutorial elements, hints, and improved parsing. Blue Lacuna goes to extreme effort to provide helps and guidance for the player, and other games (Bronze, Dreamhold) were written specifically to try to offer novices an easier route into IF. Others explore ways of making the parser more adaptive and friendly, like Dead Cities.
To my mind, this period has less of an obviously identifiable character than the periods before or since. A lot of easy and obvious experiments had now been done, and I personally was interested in games with more AI-driven agents, as seen in When in Rome and Mystery House Possessed (something I would come back to with the Versu project). Attack of the Yeti Robot Zombies is sort of a counter-New School experiment, a game in which examining is discouraged and the player is told to play only once, in an attempt to capture forward movement and a sense of adventure. Other people felt that the time for experimentation was now over and we should concentrate on consolidating our craft learning; Jimmy Maher said that that was his aim with The King of Shreds and Patches, a huge and ambitious work in some respects but in others using a pretty conservative design.
At my most unkind, I find some of the work emerging from 2008-9 or so kind of stodgy: technically correct and polished, reflecting a lot of lessons learned, but bland, inoffensive, and unmotivated by any particularly compelling creative vision or experimental question. But there are also some really significant exceptions to that: Jim Munroe’s wonderfully written and illustrated Everybody Dies, Sean Huxter’s inventively designed and IMO underplayed Piracy 2.0.
Twine Revolution (ca 2010-2013, with ongoing effects). howling dogs in the IF Comp, the evangelism of Anna Anthropy in Rise of the Videogame Zinesters, and an energetic community of (especially) queer authors writing about their life experiences led to a huge burst of new work. This was hypertext, which the IF community had been ignoring for years, but now so visible that it was impossible to disregard.
Occasional choose your own adventure or even other types of choice-based game had been turning up for a long time before this. I think in particular of David Brain’s single-player ARG Sun and Moon, which might have been quite interesting but was totally hammered in the 2002 IF Comp for being Not What We Do Here. (I bet it would have fared a lot better in the 2015 comp.) But the number, quality, and interest value of the new Twine games was so great that it wound up shifting the self-definition of the community instead, to the frustration of some of its members.
Not all Twine work is alike; though the stereotype is that it’s always short, easy, lightly branching or linear narrative with an autobiographical bent, in practice there are puzzly Twine games, long Twine games, Twines with significant procedural generation components, etc. There’s also a fair amount of non-Twine hypertext heavily influenced by Twine; and it’s fascinating to look at how Twine macros and link conventions compare with conventions from older literary hypertext such as work in Storyspace, but that could easily be a long article of its own, and I know other people are working on that history. (I have chatted a little with Mark Bernstein about it, though, and at some point I hope to put some of that discussion here.)
For specific types of mostly linear, mostly experience-focused work in this space, we’ve also started to talk about dynamic fiction or dynfic.
Commercial Revival (the last several years). The growth of mobile gaming, the support for indie games in general, the proliferation of gamebook adaptations, and the successes of inkle studios, Choice of Games, Failbetter, Tin Man Games, et al, mean that there are new sales venues for interactive fiction, especially if that IF is polished in appearance and reasonably friendly to play… which tends to mean avoiding the parser.
These studios have staked out their own territory and have very different brands, but in many cases their success has led to the growth of small surrounding communities creating freeware emulating, or using the tools of, these studios. So Failbetter released StoryNexus, which I think it is fair to say did not meet their commercial aims, but which did allow some IF authors to experiment with quality-based narrative; Choice of Games has a large forum full of people writing ChoiceScript games.
While these categories tend to be described as choice-based (in contrast with parser-based IF), they’re actually significantly different from Twine hypertext. The player is also making choices, but typically at the end of a passage rather than in-line; they generally don’t use Twine-style text morphing; and they’re often using stats in a way inspired by gamebooks and RPGs.
The Lifeline series was created by people who so far as I know weren’t especially invested in IF history generally, but who saw Twine as a prototyping tool for a particular experience that would suit a new piece of technology (the Apple Watch). Their extreme financial success has of course meant more imitators, though as far as I can tell not all of those imitators are particularly experienced in branching narrative design generally, so the results are mixed.
Even failures in this space have been influential. Varytale and several other projects represented attempts to make interactive fiction in collaboration with traditional publisher models and directed primarily at self-identified readers. While that mostly hasn’t taken off quite as well as writing IF aimed at gamers, the UI developments and formal experiments about how to make IF for a reading public have influenced the look and feel of other IF. In particular, I’ve seen more willingness to experiment with visible plot maps, stats and mechanics that reflect how a work is read rather than the movement of a protagonist through the narrative space, and so on.
Diaspora (now). IF is now being written by many types of people, for many reasons, as freeware or for sale, and distributed in a wide range of places. It is harder than ever to follow what is going on, and also there is a greater than ever variety to the works and players.
There are lots of implications to that: diversification and fragmentation of genre; removal of gatekeeping because there is no longer just one arena in which IF can succeed; the corresponding loss of a lot of community received wisdom, or a lot of people who simply haven’t picked up on it.
On the one hand I am really excited by the diversity of what is being created, and by the fact that I can barely keep up with releases in this field because they’re happening so fast; on the other hand, I find that there are fewer venues for the kinds of advanced craft discussions we used to have around IF in the New School period, because there just isn’t the common ground about what we’re even talking about to support that kind of conversation. (This again isn’t universal; it’s just a trend.)
There’s a lot of platform experimentation. One of the major features of the 1995-ca. 2005 era — maybe even through 2010 or so — was that the IF community relied heavily on a handful of the same platforms and languages. TADS and Inform and sometimes Hugo were spoken of as “first tier” IF development languages; this (in retrospect) rather condescending distinction meant that they’d been refined enough that, in the hands of a skilled author, they could offer the kind of polish that a parser aficionado had come to expect. A lot of other languages and tools existed, but were generally considered not as good, and anyone entering a game with a homebrew system into the IF Comp was regarded with extreme suspicion.
If this tendency has not completely gone away, it has at least changed a great deal in the past five years. Thanks to the relative ease of development and distribution these days, the extra resources (or hope of profit) brought by the commercial revival, the shift towards non-parser IF, there are now many more toolsets available, and it is no longer considered quite so hubristic to enter something in a system you wrote yourself into the comp. Tools and creations from adjacent fields, like procedural text generators and bots, are affecting what we write and how we think about what IF might be.
The IF Demo Fair (2011) showed off a lot of experiments with interface, procedural generation, interactive poetry, and NPC behavior. Some of those themes I explicitly encouraged, but others arose from people’s spontaneous interest in particular areas.
Meanwhile, recent parser experimentation often (among other things) plays at new ways of combining narrative and spatial representation, again perhaps influenced by what is going on in non-parser parts of the interactive fiction world.
In addition to all the caveats I’ve written above, I should also say that (as anyone with an art history background will recognize) periods are rarely neat and tidy.
In a curious way, the pieces that characterize a period in histories may not really be all that characteristic: during the era I’ve labeled New School, for instance, there were lots of traditionalist, spatially-oriented, parser-based puzzle games — more released in a given year than there were experimental puzzleless pieces. So this history is partly about what people were talking about at the time, not what the mean or median work was. Often a period is defined by whatever-it-was that was currently provoking a backlash. In the 2000s, it was puzzleless games that were “ruining” IF; then it was Inform 7; more recently it’s been Twine. And through all that time, there have always been alternatives available to those works.
Then, too, there are often scouts years ahead of their time, or archaizing works that intentionally look backwards, or works of outsider art by people who are either unaware of or uninterested in current trends.
Twine itself was a scout: the system existed for some years before anyone took it up. First Draft of the Revolution was a deliberate attempt to look forward and consider what IF might be like in an age of interactive epubs. I’ve been told that its mechanic suggested the text-cycling macro designed for Twine; if that’s true, it’s something I’m rather pleased about, in that it’s one case where my attempts to be prescient actually sort of worked.
Meanwhile, there have always been some IF authors, in every era, who wanted to write IF like what they played as kids, and that means a steady trickle of Infocom-likes or pieces based on British commercial IF of the 80s. Sometimes those works are almost indistinguishable from the work of twenty or thirty years ago; the author of Axe of Kolt has apparently ported his work fairly directly into a modern system, having originally written it for much older computers. (He has expressed some frustration about its uptake, though; I’m not sure how much of that reflects modern players’ indifference to older-style games, and how much has to do with how the game advertised itself.)
Sometimes, too, archaizing works take elements from a past period and apply a few contemporary design features to them. Just last year, Daniel Stelzer came out with Scroll Thief, which is a piece of significant technical and design ambition that nonetheless positions itself aesthetically right next to Infocom’s Enchanter series.
My own Savoir-Faire and Counterfeit Monkey are both consciously archaizing in their way — Savoir-Faire, as its initial launch page suggests, is an attempt to recapture the sense of challenge and mystery I loved about old-school IF, but to implement that with more fairness and polish. Counterfeit Monkey, meanwhile, owes a lot structurally to middle school IF, particularly in the way the plot is gated. There are a number of narrative set scenes where the player is trapped with an NPC, very much as Gareth Rees designed such scenes into Christminster. Endless, Nameless is pretty much a history of the old-to-new-school evolution in game form, with gameplay elements reflecting each stage of that development.
Finally, pornographic IF tends to be outsider art, and although it’s one of the largest works of interactive fiction ever written, most people I talk to are completely unaware of Nuku Valente’s furry adult IF Flexible Survival. Paul Allen Panks, may he rest peacefully, was a defiantly outsider artist who would submit two or three BASIC IF RPGs a year to IF Comp during the height of the New School period, even though they almost always placed right at the bottom of the list. Arguably Iain Pears’ Arcadia would count as well — undertaken with institutional backing, but also largely ignorant of conventional IF and defiant of the examples set down by literary hypertext, it belongs comfortably neither to the territory of IF Comp nor of the ELO.