Understanding Context

by Andrew Hinton

Rules and digital agency impart the ability of the map/territory to make more of itself. One way this happens is through something called procedural generation, which refers to the algorithmic creation of structure and content. Procedural generation is especially popular in video games, in which it can make a unique map for each played instance. Unlike older video games, where each level was the same each time and part of mastering the game was about remembering the patterns of each level, procedurally generated games make players rely on perceptual-reactive skills more like those needed for new territories in the real world. Sim City and The Sims creator, Will Wright, made use of procedural generation in the game Spore, and it’s also behind the chunky-but-weirdly-natural environments found in the poignantly immersive Minecraft, shown in Figure 17-16.[338] Entire unique continents can grow with little or no user input. Minecraft can randomly generate terrain, or the user can specify parameters that the game uses as a “seed” to create maps with preferred features, such as a snowy mountain range or jungle biome.

  Figure 17-16. Minecraft world generation is procedural, but can be “seeded” to create essentially the same structures from one world to the next, with slight variations in surface detail

  Most “maps that make themselves” aren’t so literally self-generating, but more dependent on user activity. Algorithms can provide structures and functions that tap into the collective work of the user base to grow rich semantic topologies.

  For example, some websites have contexts that are defined entirely by algorithm, such as Flickr’s photo-browsing context based on Interestingness (Figure 17-17). Here’s how Flickr’s describes Interestingness:

  There are lots of elements that make something ‘interesting’ (or not) on Flickr. Where the click-throughs are coming from; who comments on it and when; who marks it as a favorite; its tags and many more things which are constantly changing. Interestingness changes over time, as more and more fantastic content and stories are added to Flickr.[339]

  Figure 17-17. Flickr’s Interestingness facet, in the current version of the platform[340]

  Even though they have some procedurally generated features, environments such as Flickr, Craigslist, and Wikipedia are more rule-frameworks than static web structures. Such frameworks can channel user content creation into rapidly generated cyber-landscapes. In 2006 alone, Wikipedia grew by over 6,000 articles per day—each one a new page adding to its vast corpus.[341]

  This organic perspective reminds us of the question: where does the system end and the user begin? Ecologically, we are part of our environment; this principle doesn’t stop being true just because an environment is made largely of language and bits. In a sense, these environments are morphing and evolving through the efforts of several resources at once: digital algorithm-agents scouring and crunching raw information into new content; armies of product development teams updating their functionality and architecture; and millions of inhabitants creating new content, tagging it, linking and commenting, adding layer upon layer of semantic material that feeds the system’s appetite for new information.

  Whether the software is making more of itself, or users are generating more territory through publishing, commenting, discussing, and uploading, the effect on user perception is pretty much the same: the environment grows and grows, expanding, reshaping, and shifting under our feet. As Resmini and Rosati put it, pervasive information architectures are “evolving, unfinished, unpredictable systems.”[342] The map is also territory, complete with its own ecosystem of species migration, stormy weather, continental drifts, and the occasional exploding volcano.

  Metamaps and Compasses

  These ever-expanding, unpredictably evolving maps become territories in their own right at such scale that we need “metamaps” for understanding them. We also need the equivalent of compasses for finding our way through their more uncharted wildernesses.

  Although it is well past its day at the top of the hype curve, Second Life (see Figure 17-18) is an interesting object lesson in how navigation of our surroundings might become more about the semantic information landscape than the physical one.

  Even though Second Life does its best to mimic the ecological nature of the physical structures we have in our “first life,” in actuality these structures can change at any moment. Giant buildings can be moved, deleted, or just “put away” into an owner’s personal inventory. Whole islands can shift or disappear from day to day.

  Figure 17-18. A now-decommissioned virtual home of the IA Institute, in Second Life

  The information we rely on in the physical world to be invariant can actually be quite variant in Second Life. So, the information its dwellers come to rely on most is the meta-information: the dynamic, semantic information that informs players where activities are happening or where friends are located, all on the Cartesian grid of Second Life’s highly variant, digital topology. On a trivia group’s regular Tuesday gathering, it might be meeting in a cloud city, a jungle, or a skyscraper. The members don’t rely on the faux-physical landmarks, but the semantic ones, because they can be searched and found even more quickly than a user can “fly” or “teleport” to the location to check it out in person.

  Perhaps the most prescient part of Second Life isn’t its virtual-reality landscape, but its navigation interface, the Viewer, whose menu bar is shown in Figure 17-19. Its residents use this tool to basically “fly by instruments” the way a real-life pilot must use only the cockpit’s instrument panel to fly in a fog.

  Figure 17-19. From a tutorial on how to use Viewer version 2, on the official Second Life Wiki[343]

  More and more, we are now navigating our world by the language we put into the stratosphere with digital technology rather than by physical landmarks, even though physical objects are much more stable in nonvirtual life. We want to know which store has something in stock, which theater has the movie we want to see at our preferred time, or which bar has more of our friends hanging out in it on a Friday night. It’s this dimension of ever-shifting semantic information that we want to track and navigate. In that sense, our smartphones are now acting as “Viewer” devices for navigating the “First Life” world around us.

  Yelp (Figure 17-20)acts as a sort of compass in the sense that you can use it to filter your environment by various factors—choosing your “true north” by facets such as average review, distance, or price.

  Increasingly, our tools for navigating our environment are becoming more personalized, as well, keeping track of our preferences so as to create new “true norths” that are unique to us. The Google Now service is one of a number of newer technologies that aim to tell us what we need to know before we have to ask. Before it was acquired, a similar service called Donna (see Figure 17-21) explained that it “pays attention for you so you can focus your time on the people you’re with.”

  Figure 17-20. Yelp’s Filter and Sort interface is a sort of sextant for navigating local resources

  Figure 17-21. Donna, now acquired and discontinued by Yahoo, was a pioneer in the personal-compass app category

  By knowing what’s in our calendars and address books and tracking our email activity as well as (increasingly) our general activity patterns, these services aim to be digital agents that do some of the heavy lifting for us, narrowing and distilling the essentials of the information in our environments—creating bite-size maps-of-the-map and the equivalent of arrows pointing us in a direction, saying “wear warm clothes today” or “leave early due to traffic.” They do their best to understand our context, and in turn to improve our experience of that context.

  These wonderful metamaps and compasses don’t work effectively all on their own, though. They need to use information from somewhere, and that information has to have the right qualities to be trusted and comprehended.

  Let’s recall that these navigational aids and digital agents are only as smart as the structures provided to them in the environment to begin with. Google’s search ranks sites higher when they’re well-structured and
effectively made places. Donna can tell us to leave early because someone created an API with which the service can know about traffic, weather, and geolocation—all of which required effective labeling and metadata. A pricing service can send a text message alerting us that there’s a sale in a store we’re visiting because someone took the time to define and structure the frameworks that make those connections possible. The rules these agents use for behavior have to be defined as part of their architectural infrastructure. Digital agents need structural cues, too. No matter how smart they get, there’s work to be done in the environment for bridging between physical, semantic, and digital information.

  And we shouldn’t forget: no matter how enabled by artificial intelligence, such metamaps and compasses tend to become less accurate as they try to be smarter and more richly relevant to context. The bigger the gap we’re trying to bridge, the more it’s subject to the fog of ambiguity; this is especially the case when the environment’s information involves tacit familiarity versus explicit definition.

  We should also remember what we learned about maps: that they’re always interpretations of the environment, and they serve some set of interests. The same is true of mediating, augmenting tools. What they choose to show us is always driven by an agenda, intentionally or not. As users, we should always be wary of what that agenda might be. As designers of these entities, we should provide enough transparency to let users in on the priorities and interests being served by the way the elements and rules are assembled.

  * * *

  [318] Photo by David Fiorito.

  [319] Dibbell, Julian. My Tiny Life: Crime and Passion in a Virtual World. New York: Henry Holt and Company, 1999:50.

  [320] Most of this story line is covered in a chapter that originally was an article in the December 23, 1993 edition of Village Voice, called “A Rape in Cyberspace.”

  [321] I cover some of the more relevant things about Quake for information architecture, in particular, in a 2006 article for the ASIS&T Bulletin. “We Live Here: Games, Third Places, and the Information Architecture of the Future,” (http://bit.ly/1wma8BC).

  [322] Found at http://www.tomsdevshack.be. Images borrowed under terms as seen here: http://www.tomsdevshack.be/legal-information

  [323] Guilbert, Juliette. “How I use Rooms on Windows Phone 8” Posted in Blogging Windows November 29, 2012 (http://bit.ly/1yVcD2f).

  [324] Screenshots courtesy of Dan Klyn.

  [325] Richtel, Matt. “Young, in Love and Sharing Everything, Including a Password,” The New York Times, January 17, 2012 (http://nyti.ms/10iHFTy).

  [326] Images from Avocado.io.

  [327] http://techcrunch.com/2012/06/20/avocado-mobile-app-for-couples/

  [328] “Most Efficient Health Care 2014: Countries,” Bloomberg.com, http://bloom.bg/1yfGKON.

  [329] Jeffries, Adrianna. “One small hotel’s long nightmare with Expedia (update),” The Verge (theverge.com) December 29, 2012 (http://bit.ly/1oqphDg).

  [330] “Expedia: Bad for the Traveler, Bad for the Hotel” Posted in Luna Blue’s PlayaZone (playazone.wordpress.com) December 4, 2012 (http://bit.ly/1t9EBnF).

  [331] Photo by Ryan Somma. http://bit.ly/1Fx7LCg CC license: http://bit.ly/1vI2YJR.

  [332] “The next 10 years in mobile” techcentral.co.za November 25, 2011 (http://bit.ly/1AoObqO).

  [333] Images from Flock promotional video for work done with Tesco/Home Plus (http://vimeo.com/75586377).

  [334] Luca, 2011.

  [335] Promotional image from Google’s Glass website: http://bit.ly/1pxlcry (retrieved 2014-09-07)

  [336] Morville, Peter. Ambient Findability. Sebastopol, CA: O’Reilly, 2005:41.

  [337] Wurman, R. S. Information Architects. Graphis Inc., 1997:16.

  [338] At least I find Minecraft to be this way. Played in survival mode, without assistance of plug-ins or cheats, it’s one of the most meaningfully engaging video game experiences I’ve ever had.

  [339] http://www.flickr.com/explore/interesting/

  [340] https://www.flickr.com/explore/interesting/7days/

  [341] Wikimedia Commons: http://bit.ly/1AoOnXh

  [342] Resmini, Andrea, and Luca Rosati. Pervasive Information Architecture: Designing Cross-Channel User Experiences. Burlington, MA: Morgan Kaufmann, 2011:60.

  [343] http://wiki.secondlife.com/wiki/Viewer_2_Quick_Start_Guide/Interface_overview

  Chapter 18. The Social Map

  Talk is essential to the human spirit. It is the human spirit. Speech. Not silence.

  —WILLIAM GASS

  Conversation

  IMAGINE A BUSY DINER ON A SATURDAY MORNING AND ALL THE CONVERSATIONS GOING ON THERE: a family at a table where parent and child negotiate about eating the eggs before the waffle; a couple making plans for the rest of the weekend; a man on a cell phone reconnecting with whomever he met on a date the previous night. There are also visible gestures, facial expressions, and body language woven into the activity of talk. Not to mention the newspapers and magazines being perused over coffee—conversations mediated by publishers and writers. There are also people texting via SMS, reading news and sharing stories via email, checking their “feeds” of friends and family on Facebook and Twitter, or gazing at pictures on Instagram and Flickr. They’re using these threads of information in their table talk, showing friends at the table what is on their phones as part of the topic at hand, and using it the other way around—taking pictures of food and friends, and posting them to the cloud. Conversation is still about people talking with people, but it’s now an everyday thing to not just be in one place at a time, but in two or more simultaneously.

  We tend to think of conversation as something that fills the gaps between actually doing things. By now, though, it’s hopefully clear that talking is actually a quite tangible form of “doing,” and that we wouldn’t be doing much at all without language, which exists because of the need to converse.

  Conversation is made manifest all around us. All human-made environments are conversations in action. Cities and their structures exist because people had conversations, and they instantiate the meanings of those conversations in stone, concrete, glass, and steel. Before print and literacy became widespread, buildings such as cathedrals were the primary medium for telling stories and broadcasting messages. Later, libraries became the meta structures that house the published artifacts of the slow-moving print-based conversations of our culture.

  Structures are conversations, but of course, conversations also have structures. Watch two people talking, and you’ll pick up on tacitly informal patterns of tonality, inflection, facial expression, posture, and gesture that add up to a sort of punctuation, signaling whose turn it is to speak and the structural relationships between statements. There are formal structures for conversations as well, such as Robert’s Rules of Order, a system invented in the nineteenth century for better organizing discussions by deliberative bodies such as parliamentary gatherings or board meetings.

  There used to be a fairly clear distinction between real-time, spoken conversation and a written or published one via postal mail or other printed media such as books and newspapers. But now, that distinction is dissolving under our button-pushing fingers, as we publish our communiques without need of paper, printing presses, or supply-chain distribution.

  Conversation doesn’t exist in a vacuum; it is inextricable from the properties of its media. Just as the joints in a creature’s limbs evolved to suit the invariant structures of the environment, the way we converse also evolves within—and is shaped by—our environment. To paraphrase Marshall McLuhan, the medium really is an intrinsic property of the message.[344] So, when digitally enabled environmental properties are added to our environment, those shape the message, too.

  One thing the emergence of persistent, pervasive digital networks has done is remove a lot of the environmental context that conversation evolved in, changing the nature of how conversation works. We need to learn new properties of the environments we use to communicate; as we�
��ve seen, many of those properties don’t necessarily work like the world used to.

  Social Architectures

  Those of us who use Twitter a great deal have become accustomed to occasionally committing a faux pas known as a “DM Fail.” This is when a user means to send a direct message (DM) but instead posts a tweet publicly. For example, Figure 18-1 shows one in which two information-architecture community members had an exchange they allowed me to capture. Austin Govella accidentally tweeted “Love you”: a message he meant to be a private communication to a family member. Of course, part of the now-familiar tradition of the DM Fail is to gently poke a bit of fun at the mishap, as Tanya Rabourn happily did in this example, with a delightful “#win” hashtag celebrating the moment.[345]

  Figure 18-1. A garden variety “mistweet” and a gently snarky response

  This is not an error we would normally commit in the nondigital environment, where so much ecological information is available to let us know both for whom a message is intended and who else might be able to hear it. Even in a phone conversation, there’s an assumed context of privacy that’s specified in the environment: a phone receiver pressed to your ear whereby you hear only one other person’s voice, close to the microphone of his own receiver. What is it, then, about the Twitter environment that confuses the context enough for people to make this error?