For whatever reason, i forgot to share this on my blog when it came out two weeks ago! If you haven’t seen this, it’s definitely worth watching :)
My Water’s On Fire Tonight (The Fracking Song) (by davidmholmes)
Some of you may know that i’ve been quite interested in the Knight-Mozilla Journalism Challenges (MoJo for short). Journalism currently rests at an interesting point, where grass roots support and interest is coalescing around movements such as Open Data, and technologists are pursuing innovation not just for developer tools but as a means to change fields outside of the typical world of entrepreneurs and tech geeks.
In that context, i happened to have been thinking about games, interaction and journalism, and came up with an idea. A high school friend who works for Mozilla suggested i might be interested in MoJo. So what follows below is the result of that connection, and discussions i’ve had with friends and news people i happen to know.
If you’ve got thoughts i’d love to hear em :) You can also find the entry on the Drumbeat site here: https://drumbeat.org/en-US/challenges/beyond-comment-threads/submission/124/
Summary
Text-based Adventures for Journalism? Many have complained that print journalism has not substantively changed, regardless of distribution medium, and point to games as the engaging medium of the modern era. We can harken back to the original interactive game, text-based adventures, and gain inspiration for how Journalism can innovate.
Long form articles don’t provide users with any opportunity to interact. Interaction is left to comment threads and social media. If instead we implement articles as a container of paragraphs/facts/memes that represent a path through a topic, a user can direct themselves through a topic with an article as a touchstone or guide. Users control their experience but editorial control is still preserved.
Description
Text-based adventures (TbAs) are some of the most stripped down text based interactive systems that one can build. They’ve been around since the early 90s and continue to enjoy niche interest, including modern implementations in browser friendly languages.
At their core, the development of narratives as TbAs and journalistic writing share a similar level of granularity and attention to detail. TbAs require authors to build spaces and populate them with objects, descriptions, and behaviors players can perform on objects. Journalists and Editors, while unfettered by interaction, nonetheless pore over articles paragraph by paragraph, line by line, for the purpose of both verifying sourcing, upholding writing standards, keeping abreast of quickly moving events, and fitting within page/column length constraints. The peculiarity of the Journalistic process however, is that the level of granularity at which Journalists and Editors work is not exposed to users, who instead receive whole articles as monolithic chunks of content, a problem that TbAs do not suffer from.
However, TbAs in their original form may not be entirely suitable for direct adaptation to the journalistic process. Journalistic narratives are typically (though not exclusively) written in the 3rd person, and TbAs are operated by users in the direct 1st person (n.b. There are conceits which may bridge this gap, exploration of a world w/ an explicit narrator mediating and explaining the user’s interaction w/ the world, for instance).
The key narrative feature that both news pieces and TbAs share is an anticipation of what users know and wish to know. However, where Journalism simply attempts to target a safe lowest common denominator which presumes only what all users know in an attempt to cover the broadest swath of readership, TbAs offer users the ability to discover and investigate narrative elements in further depth, should they so choose.
Interactive News Pieces
While TbAs may not be directly adoptable as a technology, they point out elements that could be incorporated into the digital application version of an article. The most conceptually simple is the ability to make articles queryable systems. Instead of presenting users with articles as atomic objects, provide users with articles as a collection of content, statements, research/sources, which can either be consumed in a proscribed fashion, which retains the traditional editorial narrative we are used to when reading a newspaper, or, alternatively, allow the user to direct themselves through the topic of the article with whatever goals and interests the user may have.
Users could drill down a particular facet of an article’s subject they find interesting (particular sources for instance), or they could back-fill knowledge and context which they may be missing, by inquiring further about an individual or group mentioned in the article, or even ask questions which may be tangentially relevant to the current article, but may be the focus of another article.
See the quick mockup below for an idea of what this might look.
Text Based Adventures from the future!
Even TbAs have much to learn from technological advancement since their heyday, much of which is amenable to adoption within Journalism. The makers of the world’s most prevalent command line interface, Google, have built their business around monetizing what they can learn from cooperative interaction with their users via the services and information they have constructed and compiled.
In short, Google learns from their users’ input, via a humble text box, that magically takes their users to what they are seeking. Google uses this information both to make money, and to improve their systems, which in turn benefits their users. Interactive exploration of the contents of a news story opens similar possibilities. We can imagine iteratively improving an article based on recorded user journeys, much the way that Google records what search queries users submit. Perhaps there are questions which a majority of users wish to have answered regarding a particular subject. With an interactive system users can actively interrogate whether an author has directly answered their question, or failing that, get a next best approximation (perhaps via wikipedia, see the 4th mockup below).
Writers in turn get a record of questions asked, and can view aggregate information, or drill down to individual user journeys, to see whether or not their writing has been the springboard for particular questions on the subjects they are writing about, or whether users follow the path that the author has prescribed for them.
There have also been advancements in story telling systems since the Z-machine. Games like SleepIsDeath (intro slideshow) provide a two player mechanical turk (the original one, not the Amazon.com one) game play experience. Player 1 starts the game, perhaps naïve and unaware that player 2 controls the world and the elements which show up within player 1’s experience. From player 1’s perspective the game is a vastly expansive game, played in 30 second turns, able to react to any possible user input.
Thinking of SleepIsDeath as a mechanical turk system is useful for another reason. There are existing tools in the world of online journalism which have sought to bridge the gap between journalists and audiences such as CoverItLive, which serve as interactive Q&A apps between journalists, and users. These Q&As are preserved for later reference, but Q&As are subject to the same sorts of social overhead found in comment threads. They’re not news items that can stand on their own. Systems like SleepIsDeath bridge the gap between joint creation by users and authors. SleepIsDeath sessions are framed in such a way that the journey player 1 takes can be recorded and experienced by a 3rd party while preserving the narrative frame which player 2 has chosen, without focusing on the relationship between player 1 and player 2.
We can imagine, for instance, taking the queryable news object described above, and, instead of just providing the user with existing blocks of information, piping a user’s questions directly to a writer or subject matter expert to constructing answers which fit within the narrative frame of an article, and provide answers back to users, possibly in near-real time, to augment and complement the pre-prepared materials that are part of the article. In essence, turning analytics from TbA based news, into the news into a multi-player Q&A game about a news topic.
Conclusion
There’s a broad space to explore which aligns the interests and goals of users with the content creation process in a cooperative fashion. A lot has been made of the desire to have journalists further engage users during the writing process. However direct engagement with users on a social basis can have its own pitfalls, and while there are certainly benefits to doing so, systems like the ones described above may be able to help defuse some of the problems that journalists encounter when engaging directly with their userbase (which is something i still advocate and recommend they do). If we can find ways that everyone can cooperate to make the news better, it certainly is a worth while idea to pursue further.
Thanks for reading!
-Ted Han (twitter, my email is my first name at knowtheory.net.)
Mockup #1
(apologies for the notes which remain at the foot of the mockup pages. It has been suggested to me that a “like” feature for elements of the story would be a good feature too. Thanks Atul!)
These mockups excerpt content from the following NYTimes article viewed through their Google Chrome App: http://www.nytimes.com/chrome/#/Top+News//www.nytimes.com/2011/05/09/business/economy/09insure.html
Sources and references
- Parchment, an Open Source Javascript interface to the GNUsto Z-machine
- GNUsto and Glulx, Virtual machines for interactive fiction
- SleepIsDeath a 2 person interactive story telling game.
- An article by Matt Waite on evolving at the core of the Journalistic process
- An analysis of the technical issues arising from the use of articles as the basic unit of journalism
- Don’t mistake your CMS for your development platform (or, the joys of loose coupling)
See, i can’t decide which Science Fiction world we’re living in. Is this Ghost in the Shell? Or Megaman Battle Network? Something else? To be sure, technology being employed by people who don’t understand the consequences of what they’re using.
There’s a textural quality to Hanna that stands out as its most prominent feature. Hanna is a film that was made to explore an idea, and that idea is a (barely) scifi, spy drama filmed as a fairy tale. It’s both grittier and more grounded than “Pan’s Labyrinth” but considerably more fantastical than the Bourne Identity.
Hanna’s extent, though, begins and ends with that textural exploration. The film is an extremely pleasant and engaging experience that carries its aesthetic naturally and completely from beginning to end. But unlike fairy tales or other sorts of hero stories, Hanna is missing a moral, message, or psychological treatment of Hanna’s (the character) development. In short, Hanna (both the film and the character) lacks closure.
Irrespective of Hanna’s narrative progression, Saoirse Ronan (as Hanna) does a great job of playing a teenage assassin exploring her world for the first time – in a neutral, german-tinged accent, no less. Cate Blanchett is a fantastic villain and southern belle, with all of the demur faux-kindness you could possibly want out of a heartless control-freak with a drawl.
The soundtrack also bears mentioning, having been scored by the Chemical Brothers. Interviews about the film have noted how well integrated the tracks are with the sound design for the film, and that’s certainly true, weaving the Chemical Brothers’ distinctive sound in and out of scenes quite seamlessly. But more than that, i was pleased with the Chemical Brothers ability to do two distinct things. They produced tracks that are enjoyable both as a film score and as an album, but they did the film score justice in building tracks that smoothly connect both the fantastical fairytale aesthetic of the film, and the upbeat electronic pacing you’d expect out of a spy caper.
All in all, it was definitely an enjoyable experience. I’d be interested in seeing it again and probably getting the dvd.
More succinctly: make all your refusals contingent. When approached w/ an opportunity you have reservations about, recast/renegotiate the opportunity until you have no reservations (like raising your hourly rate from $25/hr to $100/hr).
I’ve been working with the CrisisCommons gang to help put together and catalogue information in response to the quake and tsunami.
What i’ve been realizing is that there is very little public written record with regard to how this is being done, or where it’s being recorded (save for the wiki), so i thought i’d start dropping links in a place that i could link others to.
So here it goes! And more will come as i find more links.
Crisis Commons links
- http://crisiscommons.org
- http://wiki.crisiscommons.org (where most of the action is at)
Nuclear related information
- http://mitnse.com/
- http://twitter.com/#!/tep_co (Tepco’s twitter feed)
Maps
- http://gis.ats.ucla.edu/japan/ (Data regarding disaster response on the ground, free phone locations, shelters and more)
- Blog post about ushahidi for japan, and tweet cataloguing
I can never remember where the BSD plaintext dictionary is (at least, on OSX). It is in /usr/share/dict in a file called “words”. This dictionary is incredibly useful, as you can do all the wonderful things you’d want via text processing tools available on Unix like systems. Want all words that contain “taco”? egrep 'taco' words Or perhaps all words ending with “ca”: egrep 'ca$' words
Just for completeness’s sake here’s the README in full:
# $NetBSD: README,v 1.2 1997/03/26 07:14:32 mikel Exp $
# @(#)README 8.1 (Berkeley) 6/5/93
WEB ---- (introduction provided by jaw@riacs) -------------------------
Welcome to web2 (Webster's Second International) all 234,936 words worth.
The 1934 copyright has elapsed, according to the supplier. The
supplemental 'web2a' list contains hyphenated terms as well as assorted
noun and adverbial phrases. The wordlist makes a dandy 'grep' victim.
-- James A. Woods {ihnp4,hplabs}!ames!jaw (or jaw@riacs)
