[Update on 22 March, 2011: Professor George Brock of City University is advocating the systematic use of footnotes by journalists to acknowledge primary source material.]
[Update on 24 Feb, 2011: in a far more elegant and effective way than I outline below, the Media Standards Trust has released a tool that exposes Churnalism, journalism recycled, or indeed copy-pasted wholesale, from press releases... More from the Guardian.]
A few minutes down the road from me, The Guardian has this weekend been hosting a hack SxSW event – which helped precipitate (in the chemistry sense) an idea for me that has been swirling for a long time, but which has been particularly swirly since long discussions at Newsfoo in December. I’m sure others are already thinking about or working on similar ideas – although my impression at Newsfoo was that perhaps not. Either way, I offer it up here, warts and all, in case it’s got some merit, and might be of some help.
A caveat: I’m not a coder or a technologist, so please forgive any tech barbarisms [update: e.g. "Metadata? Are you crazy? We can do this with basic tagging..." or "This is classic sledgehammer/nut, hammer/nail territory..." or "Uh-oh - here comes the rabbit-hole...".] I work in sectors (journalism, media, development, human rights), however, that are profoundly affected by the work that many technologists are doing, and that are facing challenges in how to make manifest the provenance, authenticity, accuracy, diversity and representativeness of the information they provide.
Five things in particular have prompted me to thrash this post out now, starting at half-six on a Sunday morning, with regular interruptions from the kids…:
- discussions at Newsfoo about how to make manifest a layer of trust and transparency in news content – distinct from a reader’s habituated trust in a certain journalist or publication
– re-reading my friend Ethan Zuckerman’s thinking on media attention, xenophilia, homophily and other obstacles to diversity of news coverage – and wondering if the problem is less that the international media’s resources and attention are unevenly apportioned, and more that local media, closer to the story, are less able to compete, to project themselves or their analysis internationally
– by extension, a bit of media development dogfooding – what’s good enough to suggest to journalists and media outlets in the developing world should be good enough for any media anywhere
– watching immense surges of communication about the #egypt #jan25 #sidibouzid #wikileaks and other unfolding crises through the Twitterverse and beyond (including the Guardian’s Ian Prior incident described here)
– Mozilla’s Privacy Icons project – an attempt to represent at a glance how a web user’s data will be held and used by a service or website (more here and here)
There are some admirable efforts to crack the nut of information overload, especially with the huge increase in individual media production (tweets, facebook updates, and so on). Projects like SwiftRiver take a technological approach to crunching information, including news sources, and rendering it (ideally) both more contextualised, and more filterable, and therefore more useful and less overwhelming. Google News has started asking news orgs to tag their stories to distinguish between original and syndicated content. Some, like Neography, are basically hieroglyphics for the news. It’s a start, but still a fairly modest one.
None of this offers readers enough fine-grained control, nor does it present a new, different prism through which to understand news events, and it’s not designed in a way that communicates the intention of journalism. Google for one is trying to get there, but its solutions are still reliant to some degree on search (and search itself is an increasingly personalised rather than shared experience). This doesn’t yet truly level the playing field for, say, local media in the developing world, or journalists writing in their native language rather than in a global language. [A propos of which, there is still no Global Voices Online equivalent that, instead of tracking and curating blogs about or from every country in the world, systematically tracks and surfaces stories from local and national media around the world.] High-quality curation in more established international media as well as on many blogs, is helping diversify access to a wider range of original sources somewhat, but again it is constrained by individual capacity, networks, inclination.
I believe we need something extra, a reasonably independent searchable layer, that helps extract and highlight some of the things that we both value and seek to avoid in reading journalism – both ends of which may help spark deeper changes in the practice or perspective of journalism as it continues to evolve.
So, rather than having users rely on a combination of searches, aggregators, curators, recommendations, affinities, semantic crunchers and luck to find what they need, what if a broad swath(e) of news organisations, and other organisations involved in the business of journalism, developed a simple (perhaps even visual) system that, before they read an article:
– addresses both the problems of information overload and consumer choice
– surfaces some of the markers that might distinguish high-quality journalism
– provides readers with clear provenance of information
– and in doing so meets head-on the issues of transparency and trust that we discussed at Newsfoo?
How about allowing readers information not just about the author, topic and location of a story, but also about how it was put together, how it was sourced, how many corrections it has? (Might something that looks a little like the Mozilla privacy icons work?) Or giving news seekers an advanced search, both on newspapers’ own sites, and on third-party sites and search engines, that allowed them to search, sort, filter using this information? I’ve started a list below of the kinds of information I have at various times been interested in seeing made more transparent, and filtering by – not necessarily all at once, in every story, but certainly at one time or another. It’s by no means exhaustive… please do have at it.
- sourcing - greater transparency in and enumeration of sourcing would be a significant leap forward in helping understand the scope and weight of a story – both in terms of people interviewed for the article, and primary sources consulted. How many sources were used for the piece, and who were they or what role do they occupy (if it’s safe to indicate this). If the story is about poverty among women pensioners, are they directly affected, are they a policy-maker with direct influence, what’s their gender, age, and so on. At Newsfoo I suggested an indicator of a source’s proximity to the centre of the story (mitigated by a FireEagle for sources, if you like…) Depending on how granular (or anal) a publication wanted to be, they could even include a count of sources consulted and not quoted, or direct, primary, original source interviews, vs re-quoted sources, desk/telephone research vs face-to-face, were the interviews conducted in local languages, or through an interpreter…
– are there pieces of primary source material available, and attached to the story, or using documentcloud or similar? Or even just a list of source material (including other media reports, blog posts and so on) that went into a piece? This might increase the authority of such primary sources, help give a sense of the scope and grounding for a given story, and at the same time provide more accountability for journalists in terms of the range and diversity of sources they rely on.
– number of corrections, highlighted corrections, and perhaps a right-of-reply opportunity tethered to the article in addition to services like ReportAnError (which perhaps could lead to the article changelog, like The Guardian’s Article History links)
– what kind of article is it? a short update to an existing set of information, an original investigation, an adapted wire piece, an evergreen piece, an interview, and so on?
– where is the journalist based? Is s/he in the country, in the city, at the scene, on the same continent, etc? One can imagine eventually being able to correlate Ethan’s analyses of media attention per publication to the miles travelled by each publication’s correspondents.
– a single identifier code for a story, or a news event? Something like a Dewey decimal system for events? (OK, maybe not that one…)
How might this help?
A couple of ideas of what might evolve in the long term…
– redefining trust in the news through a commitment to transparency (where safe to do so) in how and from whom participating news organisations acquire, handle and package information
– grow and strengthen the reputation of journalists and publications that consistently find and use sources at the heart of stories, strong research sources, have a reputation for accuracy, and so on – it might be a badge of honour for niche topic or methodology publications (e.g. long-form, web-only) to differentiate their content from others. What if readers began to decline articles about Africa from a particular publication that quoted only male sources?
– begin to surface, alongside international media stories about new events, local and local-language media that use such a system. As well as pushing local journalists into occasional international circuits, it might push them against a wider variety of readers with different expectations and experiences, which could help strengthen their reporting in the long run
– if a correspondent’s proximity to the story were to become more important as a differentiator for news outlets, it might create an opportunity for local journalists working as “fixers” for foreign journalists, for example, to count more and merit an actual byline
– readers of business, sports, etc journalists would begin to differentiate between stories (and publications) predicated on rumour and those on documented sources… Provenance, by being made a little more transparent, might become an important differentiator.
– beginning to create a clearer economy for being on-the-record; with the amount of secret docs being released through leak sites, and with Matt Thompson’s Speakularity looming into view, some way of classifying, tethering and sorting this information not purely reliant on search, semantic or other tagging becomes important, and voluntarily releasing or sharing information becomes an important symbolic act
– a couple of people at Newsfoo talked about innovations in reading contexts (I mentioned this in a previous post), but the parameters seemed to be more about the form of the articles – length, particularly – than about other facets
That’s about all I can put together on my kiddie-ruled Sunday – what do you think? An idea that has some kernel of merit, or a complete and utter waste of a Sunday? (Apologies for any errors in this – not much bandwidth, ironically, to proofread… –
I’ll take another sweep through to amend tomorrow. Piece swept on Sunday night, and slightly re-jigged. Note that I’ve not used the word “objective”…)