There have been a couple of articles (here, here) in the past few days about “The Death of the Article”. While it’s gratifying to see thought leaders catch up to what we have actually been doing here at Newslines for well over a year, it’s frustrating to be so far ahead that no-one has understood what we’ve been up to!
Such is life.
In 2012, to understand how events unfolded, I created a news timeline of the attack on the U.S. compound in Benghazi. It rapidly became the number one resource about the attack on Google’s search results, beating every major newspaper.
But there was a problem: I made the timeline of events in a text file. The 100 news events I had added could only be read from start to finish. As new events happened I had to copy and paste them into the document.
To solve this, I spent many months investigating different content management systems, eventually settling on WordPress. With the help of some custom code, I was able to add each news event into a database, which sorts all events on the same topic into a timeline. My news timeline was the seed of Newslines.
As we began to build up the newslines, I began to compare what we had with another crowdsourced site, Wikipedia.
The Web is Not Paper
Skeuomorphism is when digital interfaces replicate real-life interfaces. For example, a diary app that looks like a real-life diary, complete with leather stitching, gives some familiarity and comfort to users who are used to the real-life version.
However, the familiarity comes at the expense of the digital diary’s functionality. A digital tool is not the same as a real-life object.
Wikipedia is what happens when a book-based encyclopaedia is replicated on the web. Despite the innovation of its original killer feature (the crosslinking of articles), the site continues to act like a paper book.
It’s not just that it actually looks like a boring gray textbook, where each page is presented as an over-long text article; it also functions like a book. For example, Taylor Swift’s page has thousands of words, but the page doesn’t include a single video or music clip.
Pages about movies don’t even have trailers. This has big implications for kids brought up using visual media like Instagram and YouTube. As the saying goes: Writing about music is like dancing about architecture.
And there’s the References section, which breaks the web’s standard way of linking, to mimic the kind of linking you see in a textbook: footnotes. There are over 550 footnotes in Taylor Swift’s page!
But, on the web, you don’t need footnotes because you can link directly to the source. I solved the reference problem on Newslines by linking to source articles from the verb.
We Don’t Need No Stinkin’ Archive
Newspaper articles on the web suffer from many of the same problems as Wikipedia articles, but there’s an interesting distinction: a paper newspaper is meant to be thrown away each day. That means that every newspaper article is full of repetition, because yesterday’s paper might as well not exist.
Think about this: until recently you had to go to a library and pull out a huge big book of newspapers to read the archive. Even now, you have to wade through pages and pages of Google search results to pull together a decent archive on a topic.
For example, this timeline of Troy Carter, Lady Gaga’s ex-manager, took over a week of research to compile from hundreds of separate news sources. When researchers have to look up literally hundreds of different searches it calls into question the usefulness of Google search.
So, every time we read a newspaper story about Yoko Ono, we have to be reminded that she was married to John Lennon and that Lennon died in 1980. But on the web we don’t need to be reminded. Ono’s marriage to Lennon, and his death, can be represented as just two news events in a database of news events about their lives.
A Database of News
Newslines is built for the web. Our contributors summarize each news event in 50-150 words, add a quote, a YouTube video (if available) and tag it with the topic and the event type. News event types run from births, deaths and marriages, through interviews, performances, arrests and legal issues and many more. We source from news articles, video clips, and Facebook and Twitter posts.
In effect we are creating a database of news events, organized by topic. We currently have over 10,000 interconnected newslines made up of over 30,000 events.
Because the events are in a database it’s easy to create different views of the data. By default, each page is sorted to show most recent news first, but simply hit the sort button and the news events reverse to show a ‘biography’ view, for example, Freddie Gray latest news, or by biography view.
Data-driven Pages Reduce Bias
The major advantage of building news through news events is not the extra control the user has over the data, but that each news event, and therefore each newsline, has far less bias than traditional articles.
This is because articles are often (consciously or unconsciously) written with a particular result in mind. By contrast, each newsline is built up piece by piece with no goal in mind. The story is revealed to the reader as more events are added. The ability to influence the entire article is minimized.
This is particularly true in cases where the sources are highly partisan. Our Newsline on Emma Sulkowicz (Mattress Girl) was compiled by extracting only the factual information from highly biased articles on both sides of the debate. The result is not only the most complete account of her case on the web, but the most unbiased account.
The Medium is the Message
Marshall McLuhan’s famous dictum, “The medium is the message,” resounds strongly through my experience building Newslines. Each media creates its own kind of communication.
When you have paper, you make books and newspapers, articles, and print photos; TV uses moving images; The web uses databases, interlinking, filtering, and video.
While the final forms have not been developed yet, just as Wikipedia is stuck trying to replicate a paper encyclopaedia, newspapers are stuck trying to replicate the look and feel of a paper newspaper.
The beauty of the web is that it isn’t restricted to paper, and the data can be re-imagined to better suit the new media. The article is just a dying artefact of that past.
Long live the feed.
- Newspapers will switch to news feeds
The article, at least for breaking news, is finished. Instead, each topic will have its own channel, updated with only the latest news as it happens. I see two streams: A live stream for Twitter-like real-time news, and another for the curated archive.
- News gatherers will not need to repeat background information
News gatherers only need to find the most recent events and add them to the database. There is no requirement to present old information unless it is for analysis.
- Analysis will be completely separated from data
A database needs structured data. It doesn’t make sense to mix news, analysis and commentary. Expect to see interesting ways to link news events to create analysis.
- Short raw video clips beat talking heads
Unlike TV news, which is mainly commentary anyway, a newsfeed doesn’t need long video clips. Just some raw footage to enhance the curated text.
- Archives become very important
Traditionally newspapers haven’t cared about their archives. But the archive is the core of an interactive news feed. It brings depth and context. To quote Orwell: Whoever controls the past, controls the future.
- Archives have to be rewritten to match feeds
It’s also not enough to take some old news article, add a topic header and call it “curation”. Each curated item must work as part of the whole. It’s not enough simply to curate a list of existing articles, because those articles are filled with repetition and background information that annoys the reader. The archive must be custom created to avoid repetition.
- Fluid feeds
The feed isn’t a one-way street. It’s not just current news flying by and discarded, but a collection of data that can be manipulated to give a particular view that the reader wants. For example, Bill Cosby’s accusations, Tom Hanks movie roles.
- Better tools
The first stage is to collect the data, then to give users the tools which allow them to get meaning from it. Expect interesting tools for readers so they can analyse the data themselves.