There are few people more committed to explanatory content than Tristan Harris. In July 2007, he helped create Apture, an application that allows users to highlight any piece of text on a website to view contextual videos, search results, and Wikipedia entries, all without leaving the page. In March 2010, Harris joined Jay Rosen, Staci Kramer, and Matt Thompson to lead the South by Southwest panel, “The Future of Context,” that explored the importance of providing background to ongoing news stories.
We recently caught up with Tristan to discuss some of the best (and worst) explanatory practices.
What inspired you to start Apture?
I think when people read the news, they’re briefly inspired to drill deeper and learn about things they’d never expect to be interested in. A lot of times though, we don’t drill down, because it’s “expensive” to get the backstory. It’s expensive in terms of the time, clicks, mouse movements, loading time, and the need to leave the page.
In 2006 when I was brainstorming this problem with the Stanford Knight Fellows, our canonical example was the BBC News website. Every article about big stories like Darfur, or the North Korea Nuclear Debate, included a sidebar with clearly marked links for backstory, e.g. “Background: Darfur Crisis.” It was exactly what you thought you wanted as a reader, but when you clicked on it, it took you off to a new page and into an overwhelming sea of text. As the page loaded more and more, you’d watch the right hand scrollbar shrink smaller and smaller, effectively telling you, “you might as well give up now, because you’re never going to read all this.”
Noticing how these tiny subtle costs made me explanation-averse, I thought there had to be a lighter-weight way to give people context on demand. That’s where Apture started – with the idea that we could remove those costs and empower readers to easily learn about things, and give publishers a super-easy way to provide the functionality.
It’s quite similar to behavioral economics, in that the difference between something costing a few cents vs. being FREE, is massive. If I told you I was selling lattes for $0.20, you’d say “that’s a pretty cheap latte,” but you wouldn’t necessarily be motivated to buy one. But if I said I could give you a latte for FREE and just put one in your hand, you’d likely accept the latte even if you weren’t really thirsty. So imagine if drilling deeper into knowledge and explanation was “free” and effortless, we think people would do it more often.
Why do you think explanatory journalism is important?
For news and journalism to serve its purpose in a democracy it has to create deep comprehension and long-term retention. If you spend an hour reading something but don’t truly understand or retain it, it’s just a waste of time. If you read a book, but can’t remember or deeply comprehend the information, you can only think of it as entertainment – a way to pass the time. Without comprehension and retention, journalism won’t create an informed public.
Ironically, despite the wealth of information available to us, many of us don’t feel informed at all– and that’s even when we try to understand. In 2008, I read probably thirty articles about the financial crisis as it unfolded, but I still had no idea what was going on. Explanatory journalism will be critical in solving this problem.
What’s also working against us is our increasing inclination to value efficiency– we always want to consume more information in less time. It’s all about doing more with less. And yet when we think of efficiency, we only think of one aspect: how quickly can we get from the beginning to the end of an article, or book, or video. So if we read a lot in a short period of time, we feel “good” about having spent the time wisely. However this totally misses (1) the comprehension of that information, and (2) the long-term retention of that information. Both of these qualities are lost, the more efficiently you try to consume things.
That’s where explanatory journalism can be transformative. It can plug the holes in our comprehension and retention.
So what kind of things should we be doing to make sure our explainers promote comprehension and long-term retention?
Well, I think we need two things. First, we need to come up with “best practices” to explain things to people. We need people like you and storytelling experts to invent the best ways to explain large, complex issues to people and prove that they engender better comprehension and retention. The end product will probably be an intersection of best practices in teaching, storytelling, user interface design, and even best visuo-spatial techniques to visualize information.
Once we’ve come up with best practices, we need to take these patterns and make them reusable through tools. We need to systematize and codify those methods. For example, if you come up with a great way of explaining a country’s military, we need to take that pattern and make it into a reusable template to explain other countries’ militaries.
If you think about it, this is actually what Wikipedia does. Wikipedia has templates for every type of article. They have templates for writing about celebrities, biology topics, companies, cities, countries… everything. Unfortunately, I don’t think Wikipedia has done a great job of optimizing their articles around explaining things to people, but instead have optimized for comprehensiveness. I think the opportunity is there to create more of an “Explainapedia” which has templates for how to explain big issues.
Once we’ve discovered best practices per type of information or topic, then we’ll want codify a new set of best practices around how you personalize that information per audience or end user. For example: how much time does the user have to listen and learn? How much do they already know? What aspect are they most interested to learn about?
Other than Wikipedia, who do you think is doing the best work in the art of explanation?
First I’ll mention some things that I think fail to do this. Topic pages on news websites absolutely fail at this. When you link the phrase “Credit Default Swaps” to a topic page that forces the user to choose from 10 related articles, you’re not doing explanatory journalism. Even if the user clicks through a few related articles, they’re bombarded with redundant information due to the way articles each need to provide just enough context to understand it. Topic pages force users to do more work, not less.
But a few sites have done great work. I think Jonah Lehrer does an amazing job explaining how the brain works to mere mortals on the Frontal Cortex Wired blog. BBC has a library of visual explainers like this and some interactive visual features here. The Khan Academy work is also good. So are TED talks.
What are they doing right that makes their content so successful?
The common theme I’ve personally found in good explainers is (1) an aim to be brief and get the main points across (telling you the “three things you need to know”) and (2) using the richest, most engaging format that works for the topic–whether that’s a narrative, a debate, or a colorful data visualization.
For example, take TED talks. I don’t think of them as an amazing source of journalism per se, but their brand comes with a certain promise. First, all TED talks are somewhere between 5 and 20 minutes. Second, they are told by an engaging speaker, usually a leader in their field who has training and experience in public speaking. They are often a celebrity, in which case the star power brings more willingness to listen to what they have to say. So I think these talks don’t do the explaining, but they lay the groundwork for deeper explanation.
Ignite talks are similar. Their motto is “enlighten us, but make it quick.” It’s a similar premise.
Another type of site that doesn’t explain concepts directly, but explains implicitly through their own format is BloggingHeads.tv. Instead of explaining an issue to you, they let you compare and contrast through hearing two leaders in their field debate one another. Again, (there are) some common themes: they take leaders in their field who have some star power, but instead of getting one, you have two or three.
The BloggingHeads format is not different from ways we know help people learn new languages – you use parallel corpora. Instead of explaining grammar directly to people, you just show them example sentences in two columns – one column in the foreign language, and the other column with the phrases in your native language. Through comparing and contrasting examples, you often get a deeper understanding. It’s classic “show, don’t tell.”
What about from a multimedia standpoint? What tools or technologies are being used in the best explainers?
I don’t have a specific answer, because it’s topic-dependent. I think it’s important to know when to use multimedia, and when to not use it. I read a paper recently on the effectiveness of teachers using visual animated explainers like the BBC examples I mentioned, versus using static graphics.
It turned out that while students might enjoy the animated ones more, these often lead to worse retention and comprehension. In an example of explaining how neurotransmitters flow in the brain, they tested one group with an animated explainer showing how the chemicals flowed from one area to the other, and showed another group of students static images of multiple states of the process, and the kids remembered the static ones better.
The argument was that static media causes a student to have to work harder in their own mind to visualize how things transition from State 1 to State 2, versus having it shown to them by an animation. Animations create more passive engagement with media, while visualizing the transitions yourself created active engagement. The animation is “easier” to understand and fun to look at, but you don’t comprehend or remember as well long-term.
I think this is an example of a nuance in multimedia journalism. We can get excited by the whiz-bang of animated BBC data visualizations and guides, but sometimes they really aren’t better than more primitive storytelling using images, text or sound.
In “Context: The Future of the Web,” you talk about how we must provide content in a scalable way. How do we make explainers scalable and repeatable while still making them unique and interesting?
That’s a great question, obviously that’s really hard. It’s like saying, “How do you templatize art and make it scalable and repeatable?” And even if you do succeed, you almost always lose that magical quality of a single great work. It’s like the disappointment of seeing a great local restaurant turn into a chain.
Right. You can’t make a template for the song NPR used in “Giant Pool of Money.”
OK, I think about it like this: instead of the explainers themselves between created off a factory line of templates, I still think it’s all about people painstakingly doing the best work to explain a topic to someone else. It’s still about the art of doing that well.
BUT, when it is done well, no one should have to do that work all over again. Meaning, if you did a great job of explaining the difference in size between the South Korean military power and the North Korean military power, then I should be able to reuse that work for all time. I shouldn’t have to create that explainer again.
And I think that’s where technologies (like Apture) can help, in identifying latent needs where people are looking for more information, and then helping to automatically remember which explainers were most successful so they can be reused for all time.
Interview by David Holmes
Learn More About Tristan and Apture: