Context: The Future of the Web & Inklings of SXSW

by Tristan Harris on March 10, 2010

This is an intro to my panel next week, exploring the Future of Context w/ Jay RosenStaci D. Kramer, and Matt Thompson. I’m writing this to share some early thoughts and get you involved in the conversation, so pls leave your feedback in the comments and come Monday the 16th!

Say you’re walking into the Metropolitan museum of New York on a sunny afternoon in May. You walk inside and stroll down a hallway of 17th century paintings from Italy. If you’re like most people, you probably don’t know much about the paintings that line the halls, or why a certain piece is particularly notable or revolutionary. You just sort of go along with it. You’re obeying an implicit social contract you have to the museum during that half an hour– “I’m in a prestigious art museum in NY and society says the paintings here are important, so I might as well pay attention for a little while.”

But the reality is, you don’t really know or care much about the paintings on the walls. While you might glean bits and pieces from the tiny yellow notecards appearing next to each piece – year, author, type of paint used – the whole experience is relatively flat. The paintings haven’t given you any reason to care about them. Put another way, if you peered into your brain during this experience, you’d probably see it light up pretty simple, low-order sensory areas: “look, there’s a black brush stroke on a giant white canvas.”

Compare that experience to this one:

Suppose you walk into the Met and an NYU Professor of Art History suddenly appears saying she wants to tell you everything about art in the museum. She grabs you by the hand and leads you through the hallways, enthusiastically explaining the different artistic periods, pointing out the significance of each flourish used by the painter, describing the life and economic status of the artist during the time they painted, and so on.  Equipped with this framework to understand the painting, instead of just seeing colors and lines on top of canvases, you now appreciate detailed information about each piece that you couldn’t have before… it’s almost as if you’re perceiving a different painting than the one before the Art History teacher showed up.

This is how I think of the word context.

Context is information that informs your understanding of the world, literally allowing you to derive more meaning from an experience. In the case of the painting above, it deepens the meaning of your experience in the Met by increasing the # of features, patterns and ideas you’re aware of in each painting. Even though the rectangle of colored brush strokes is the same as the one you saw before the NYU professor told you all about it, you actually see the painting in a different way after you have context.

Context in web-based news and storytelling:

So how does this relate to my panel next week on the Future of Context in publishing?

If you think about it, the way we read news on the web is much like walking down the hall of the Met without having context. Much that we read relates to big topical issues we probably don’t fully understand: the Health Care debate, the standoff between Iran and the United States, the Financial Crisis. We know we should pay attention to these topics, yet we’re massively under-equipped to understand their nuances and complexity of the issue, especially since it comes to us in a daily flood of headlines (e.g. “UN weapons inspectors are visiting Iran today…”). Articles like this are about as helpful as the little “yellow notecards” are giving basic information about each painting in the Met.

Re-inventing publishing around Context

As Jay Rosen put it, the word “context” itself implies something that information that is secondary or supplemental to the “main” text. But you can see from the examples above that that’s wrong.  Context is primary. You actually need context before you can make much sense of what’s in front of you.

The exciting thing is, the web is most incredible tool we’ve ever had to solve this problem. We’re at the point where there is always something out on the web – a video, a background story, a Wikipedia article, a set of photographs, or an idea in someone’s head – that could provide a greater context and understanding for the topic. Go ahead, think of any topic in the world, and there’s something out there.

A new approach to context

At our panel we’re going to explore a new context-oriented approach to publishing and user experience. As I see it, the problem breaks down into a few distinct challenges:

  • Repeatable & leveraged work: how do we provide context in a scalable way?
    • How can organizations leverage existing work and not re-invent the wheel each time they want to explain something or provide context? It bugs me that when NYTimes or BBC creates a fantastic infographic explaining how subduction plates cause earthquakes, they don’t re-use that infographic the next time an article runs about the next big earthquake or tsunami.
    • Coming from an engineering background & culture, you learn that you should never do work you can’t modularize and re-use.  This was the whole revolution of object-oriented programming in the 1980s. But journalism doesn’t follow this practice at all. Articles are written from scratch every time, never truly re-using or building upon previous writing to save time and money. What news needs is object-oriented journalism in which context is a basic building block upon which to create articles.
  • Personalization
    • When walking in the Met, maybe some of us know a lot about 17th century Italian painters, but a lot of us don’t know anything about it. Context needs to be personalized to the audience so we don’t waste time repeating information to those who already know a lot, while still giving fresh audiences a good entry point into the topic. Like Malcolm Gladwell once paraphrased at TED, the answer for creating the best experience isn’t about finding the single best article for everyone, it’s about finding the set of best articles so that there is a version of the story that suits each customer of the news perfectly.
  • Sources of context
    • What sources can we leverage to provide context, and with what guarantees of authority?  What about all the information harnessed in the vaulted archives of media organizations (all previous articles, photographs, videos, or interactive infographics)?
  • Brevity and hairstring attention spans
    • Suppose we were to invent the perfect way for people to be provided with context, as humans we are bound by the physics of shortening attention spans. People have less and less patience for articles that cause the scrollbar in our browser to shrink to a tiny little nib, and we have less and less time to invest in learning something new.
    • How do we design context for the attention-strapped psychology of our minds?
  • Structure & Design
    • What is the ideal structure to communicate context anyway? Is it just another web page, a contextual window appearing within the page, or a wiki? Many news organizations today attempt to solve the problem by providing “topic pages” with links to further materials… but everybody knows a page full of links is overwhelming and too web 1.0. Who has the time for more links to more pages, and inevitably more tabs? We need a content-oriented context architecture, one that aims in the shortest amount of time to give people the information they need, without leaving the page (sounds familiar, probably)
  • Business model and web economics
    • Changing how articles are written & published will inevitably affect how publishers generate revenue from advertising, and how it impacts things the ecosystem of link-love and SEO. We can’t re-invent news around context without respecting the business models that allow publishing on the web to thrive.
    • What we’ve found so far though, is that today’s most popular sources of context (Wikipedia, NYTimes Topic pages, fascinating blog posts smothered in link love) are largely rewarded by Google in SEO. If you Google for just about any topic, how likely is it that Wikipedia is in the top three results? This speaks good things to publishers if they were to succeed in re-orienting their content into SEO-rich, evergreen pages with persistent URLs that describe the bigger picture.

I’ve been fascinated by the role of context in our lives for more than four years and can’t wait to discuss them with you and several leading thinkers in this area next Monday.

What do you think? What else should we address in our panel?

{ 5 trackbacks }

The future of context and the future of journalism
03.16.10 at 10:46 am
Jason Kristufek's We Media Blog » Blog Archive » It’s time to change the rules
03.17.10 at 11:03 am
This Week in Review: Loads of SXSW ideas, Pew’s state of the news, and a dire picture of local TV news | Mark Coddington
03.26.10 at 12:57 pm
July Linkdump Social/Political 1 | Idiotprogrammer
07.13.10 at 4:22 pm
Periodismo orientado a su objeto | +None
09.21.11 at 4:48 am

{ 2 comments… read them below or add one }

1 Mike Coleman 03.12.10 at 6:58 am

Thank you, Tristan. I thoroughly enjoyed this post. I rankle when industry people say content is king…that’s bunk. The industry is stuffed to the gills with content and look what it has gotten us: quip headlines and a couple “related articles” links, usually hand-fabricated in a non-predictive, completely human manner. FAIL! Our CEO, Chuck Peters (, has been quite clear on our direction as an information provider of choice (I know how to spell “media company”, but that’s not what we seek to be). Context is the result we achieve when content is created digitally in a predictable and repeatable semantic framework (read: more automation, minimal human involvement in the linguistics). This is no easy path to travel and it is requiring open, introspective thought on technology, content gathering (newsroom) and products.
Yes, context is king – it governs how we relate to each other, how we perceive our surroundings and it is the dominant factor in our decisions. Thanks again!

2 Tristan Harris 03.18.10 at 5:46 pm

Thanks Mike! Hope you got to check out the rest of the discussion on where more of the debate is getting played out.

Leave a Comment

You can use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>