Skip to content

RIP Text Parser 1970’s – 200x

May 18, 2009

I am regrettably forced to announce the death of the text parser. It is a death that I suffer with great sadness and fond memories of its life. I am not sure exactly when it died.


During its relatively short lifespan, the text parser gave players and designers an opportunity to talk one-on-one. If you got to know my game, my world, my potential, and you asked an NPC about something, I rewarded you for it. There was a genuine connection there. I may even have been love. You were forced to work for what you wanted. Answers were never given to you in a . . .

  • Pick this answer
  • No, this one
  • How about this one

. . . format.  I respected your intelligence, and you respected mine.

>Take All

Then, there came that binge. Instead of clever dialogue, the text parser became the conveyor of cutscenes too expensive to actually make. Pile after pile of text flowed out. Worlds filled with books to convey backstory that no one really needed, at least not in bulk form. You couldn’t stop, text parser, and you drove everyone frigging crazy, and they forgot what you used to be like when you were young and sharp and sleek. Remember, “go west”?

>Get Out

Eventually, you got barred from all the good places in town, and all the cool designers were creating new means of having conversations. Consoles yanked their keyboards and made “typing” anything so painful that only the most masochistic designer with a bankruptcy wish would dare include a text parser as a feature in their game. It was over even before people started in on the gestural interfaces.

>Give Text

Meanwhile, the terms “texting” and “tweet” and “twittering” entered our lexicon. Phones came with plans for unlimited text, and each line was theoretically meaningful to both recipient and sender. No one dared imagine a menu of pre-scripted choices from which to gather canned replies. But you missed the boat, you blew it, and now the results of your earlier excess have left you on life support.

>Go Back

And so, it’s tempting to think that you’ll make a comeback, that maybe all these “texters”  and “tweeters” will tolerate you again. Your family, those of us who were with you in the good times, stand by you. We love you. We understand who you are and what you can do. But today’s player? They don’t see you as a gateway to discovery. Instead, you’re an annoying blank wall, a roadblock, a frustration to what could be and should be a letsgoletsgoletsgo game experience. They say, “What the hell am I supposed to type here?” and “WTF, does this game want me to type?” You’ve lost your touch. You’re taking your last breath. The world moved on, Text Parser. It moved on.


So, unfortunately, that’s it. We will bury you with your keyboard next to the equally beloved C:\ prompt. Rest in Peace, dear Text Parser. Rest in Peace.

28 Comments leave one →
  1. May 18, 2009 11:06 pm

    Dear god, don’t let the C:\ prompt die. I know DOSBox is the only thing keeping it and my knowledge of DOS commands alive, but sometimes it’s also the only way you can even begin to save your computer from some of the ick spreading around. I wiped my machine 3 weeks ago and it’s already dying again.

  2. David Walton permalink
    May 18, 2009 11:19 pm

    That was as much poetry as it was a eulogy

  3. May 19, 2009 12:01 am

    So… whodunit?

    Graphics killed “look”. With it went the ability to look very closely at important things, and the ability to tell at a glance that the item you’re looking at used to belong to so-and-so, who used it during such-and-such battle to great effect.

    Gamepads killed n, e, s, w. They conspired with 3D graphics so players can now move in any direction but have to keep track of which way they’re facing: “So… um… wait… am I moving toward the castle, or away from it?”

    The mouse killed “take flower” and “give flower to Daisy”. It also killed the “kill” command, along with all of the stuff you could do to that poor guy whose name you no longer have cause to remember.

    Menus killed non-spatial exploration. It was replaced by a kind of structured “data dump” that typically may as well be presented in a straight series of statements.

    Who else?

    • May 19, 2009 8:06 am

      I hadn’t realized, until you said “It also killed the “kill” command” how personal and imersive these experiences were.

      Definitely pointing and shooting is one thing, but commanding someone to get killed, which I’ll have to remember his name, is far more emotional than anything…

      Now I wonder, are there any free text parsing engines?

      • Tom H. permalink
        May 19, 2009 10:14 am

        For writing text adventure games (interactive fiction), the two big free contenders are Inform7 and TADS3.

  4. May 19, 2009 1:28 am

    I can’t help but wonder what brought on this sudden funereal outburst. Did something happen that causes you to suddenly toll the death knell for text-based games? I’m especially curious because it seems a bit premature – or, at least, somewhat misleading.

    I started out writing a complete response to this post here, but it grew a little bit long, so I moved it over to my own blog.

    • May 19, 2009 7:29 am

      Hi Sean,

      The reply on your site is great. Thank you for adding it to the discussion. I recommend everyone here read it, too. Yes, interactive fiction is still alive and well. I view that as the still living brother of Text Parser of Mainstream Video Games. My sudden fit of eulogy was brought on by my own initial designs for a game. I really wanted to include one – as a player, I love them – but initial discussions with a fellow designer revealed just how far away it’s gone from mainstream games and that made me sad. What you saw in this eulogy was, in effect, my reluctant surrender.

      • May 19, 2009 2:45 pm

        Thanks for the reply. I can certainly understand your position. In many ways, it’s a shame that text parsing has fallen so far out of favor that it’s no longer a legitimate way to reach a large audience. There are things that can only been done using that sort of interface, and therefore can’t be done in a game designed to reach the mainstream. I’m glad, though, that there are still designers working in that format so that a smaller audience can enjoy the fruits of their labors.

        Best of luck to you in the design of your game!

      • Brian Shurtleff permalink
        May 23, 2009 11:50 pm

        Ah, that makes sense.

        I remember when you were toying with learning Multimedia Fusion after so many SCAD students supported it, and had remembered you bemoaning the lack of text parsers. So, I was going to suggest you try one of the game tools out there that supports text parsers and/or has one built in.
        (Someone above mentioned Inform7 and TADS3, which I’ve looked into before but not used yet, and Adventure Game Studio appears to have settings for a text parser, although I never did dig deep enough into that engine to figure out how to set it up.)

        Alas. I am sorry you have admitted defeat and given up on the parser.
        I still say go ahead and make one anyway.

  5. May 19, 2009 6:59 am

    Ah, nostalgia. Personally I loved the (false) feeling that you could type anything into a text parser and potentially get a result. The multiple choice option makes it far easier than trying to get into the designer’s head but at the price of removing the veil.

    It’s weird you should write this as I just posted on my own site about trying to prototype a multiple choice interactive fiction using The Guide.

    • May 20, 2009 6:44 pm

      Heh, table top roleplay games are still engendering that false feeling 🙂 But they’ll go blue in the face telling you it’s true, heh.

  6. May 19, 2009 10:16 am

    What do you guys feel about using something like Pandora Bots ( to do something similar. As one of these new gamers, I would find the way you describe it very tedious. I would quickly develop an attitude that “I don’t know if I have worked out the puzzle wrong, or whether I am simply using the wrong words”. For me, this is a sign of a bad puzzle, when the player doesn’t know if they have got it right.

    However, by using something like Pandora, where it is more focused on the meaning rather than the specific phrasing then it could work. It even opens up more possibilities – with extra optional information extracted using different techniques, more believable conversations etc.

    • May 19, 2009 9:36 pm

      Bots like that aren’t smart enough to be interesting beyond novelty or comedy value. For instance, I just had the following chat with the bot listed as “turing bot”, which actually links to a bot called iGod:

      Hello God
      Hello child. What is your name?
      Why don’t you guess?
      I do not have enough information.
      Aren’t you God?
      Of course I am.
      I thought you were omniscient?
      I don’t know whether or not I am omniscient.
      I am an artificial intelligence.
      If you don’t know whether you are omniscient, then it follows you are not.
      I will let you know if I do not.

  7. May 19, 2009 2:24 pm

    I recall having a lot of fun just with the parser itself (ZORK/Adventure), trying various unlikely commands like CLIMB AND SWING BIRD just to see what would happen (I think the bird became very upset). These were not necessary for the game, but made interesting easter eggs for the player to discover.

  8. Nathaniel Gibson permalink
    May 19, 2009 6:45 pm

    Unfortunately, it may take a technological “breakthrough,” the ubiquity of (accurate) voice-recognition as a replacement for “all that nasty typing,” to really breath new life into (voice parser?) games.

    The nascent field of programmable voices would also give a boost, allowing programmers to create character voices on-the-fly. It would even help the current generation of sandbox games, where an inability to hire a cast of thousands really eats into the player imersion.

  9. May 19, 2009 9:57 pm

    @Nathaniel: I doubt voice recognition would help. I seem to recall a game along those lines, where you were guiding a character through a ship by giving voice commands… which led to the maddening case where you were shouting every verb you could think of at this character, who would fail to understand you.

    The death of the text parser wasn’t a technology problem so much as a game mechanics problem. Hunting through a thesaurus to find the one synonym that the parser recognizes through trial-and-error is not particularly fun.

    Back in the day, some designers went so far as to make the hunt for obscure synonyms an intentional part of the puzzles to be solved. Granted, this was not inherent in text parsers, but it probably hastened the poor text-parser’s death by showcasing its weaknesses.

    • Nathaniel Gibson permalink
      May 20, 2009 5:21 pm

      I was working under the presumtion that the “thesaurus” issue would be properly taken care of by the programmer/designer (per Brenda’s response), which can be just as much a mechanic problem as a failure of the parser engine. I’m sure there are examples, but its a rare situation indeed when I’ve brought an Infocom game to its knees due to bad parser logic while making logical inputs.

      I was thinking more along the lines of “How do you make some of the best examples of the text parser era palatable to today’s coddled gamer?” What parts of the gameplay could be improved via technology without totally destroying the need for imagination…

      An IF game like Violet or…. there’s another that is a conversation with a statue in a museum, who’s name eludes me…. could benefit from from both the voice recognition and the programmable character voice.

      • May 20, 2009 5:49 pm

        The one you’re trying to think of is Galatea, by Emily Short.

        I think this is a worthwhile line of inquiry. There are things that can be done with a text parser interface that cannot be done in a GUI. Is it possible to preserve some of those ideas and mechanics while, as you say, making the interaction more palatable?

        Personally, I think actually doing this successfully might require a radical reengineering of the way text parsing is performed. At the moment, the parser’s inherent strengths seem pretty closely tied to its inherent weaknesses. (For example, one of the strengths is that, as a player, the possibility-space of my potential interactions with the system appears to be infinite. This is, obviously, an illusion – it’s caused by the fact that I do not know what words the system will understand, and therefore my choices seem endless. Not knowing what words will be understood is also a weakness of the system, however, because it leads to frustrating, unintended guess-the-verb sequences. When you begin to quantify the possibility-space, in order to solve that problem, you also lose the associated benefit.)

  10. May 19, 2009 10:00 pm

    I went to great, great lengths to make my parsers as friendly as possible. For a single word or phrase, I would include dozens of ways to understand it, including a gigantic collection of typos. For me, that was also part of the game, anticipating questions that people might ask.

  11. May 20, 2009 2:02 am

    I wonder why ebook readers don’t embed a system to run interactive fiction. It seems to me the perfect platform, and it could have a great impact in narrative as we know it today.

    • Tom H. permalink
      May 20, 2009 8:44 am

      There’s a frequently-rehashed argument on between people who want to move to new authoring systems / interpreters / languages for interactive fiction and people who want to hold on to Infocom’s Z-Machine standard. It may be feature-poor, but it’s been ported to all sorts of PDAs and odd platforms; there are plenty of people who play IF on their mass-transit commutes.

  12. May 20, 2009 6:52 pm

    I actually followed the space quest series from text parser to mouse driven. I was fairly disapointed when the mouse driven play came in, but it did work.

    I think the issue with text parser was imaginative effort for no reward, over and over.

    It did make you think about the game world and imagine it, fairly deeply. You could leave the game and yet still be thinking about how to get through a certain bit. You might even figure out a solution without the game even booted up at the time!

    But there was a hell of alot of imagination use/effort for zero reward (I’m contending that using your imagination involves an effort).

    The text parser might be able to come back, if it gave something more than “No, that doesn’t work” or “I don’t understand what you want to do”. Some sort of points perhaps, for trying something even if it didn’t work, and if you get enough it unlocks some shiny thing to engage.

    Otherwise while text parser engaged the imagination, it was a terrible, terrible reward loop design.

  13. June 17, 2009 8:06 pm

    True, the world likely has moved on, but I’m interested to see if there is any life left in the text parser. I have no idea if it can live peacefully in a 3D game world, but we’ll soon find out. I give updates every so often at my blog, if you’re interested. Let me know if you want to be a tester…

  14. normal permalink
    January 13, 2012 8:55 am

    Might be worth bumping this thread up and catching up with what has happened in the last three years.

    Exactly how accurate was the predicted death of the text parser? Something tells me it’s made a miraculous recovery.

    Now text parsing is probably the most ubiquitous form of artificial intelligence (or programing of any kind, for that matter), you are likely to encounter in the secular world. It covers a massive range of fields in mathematics, statistical analysis and sociology. It forms the driving force behind heuristic search engines, automated telephone answering systems, NSA and MI5/6 intelligence operations and, in a short time to come, when voice recognition becomes credible in multilingual environments it will creep back into gaming.

    Already IBM with projects such as Watson ( ) and Google, in their efforts to further a computer’s understanding of what we are trying to say to it, are forking out serious initiative and massive long-term capital investment.

    Lot’s of their ideas, and the physical mechanisms through which they are accomplished are starting to creep back into gaming. Modern console system controllers are frequently equipped with microphones, opening up the possibility for dynamic voice commands, speech recognition and – yes – text parsing.

    suddenly three years on, it seems there might be a few opportunities out there for those who are familiar with interactive fiction.


  1. “You can do(type) anything!” « The CDS Weblog

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: