Introducing “public annotations” in journalism
How can anyone tell these days what is factually accurate and what isn’t? What is formulated to reveal and what is written to conceal and mislead?
These are increasingly pressing questions, especially as the next historical round of disinformation is upon us and ‘fake news’ is flourishing in all its glory. Can critical readers, viewers and listeners help in improving the reliability of our information?
Yes, they can, and Make Media Great Again (MMGA), a new non-profit start-up, is capitalizing on this idea with a method for delivering constructive criticism.
MMGA is introducing the concept of annotation in online journalism, starting in the Netherlands, but not in the technically elaborate manner of the World Wide Web Consortium’s Web Annotation Data Model, which describes a structured model and format for enabling the sharing and reuse of web content annotations of across hardware and software platforms.
“MMGA annotations are notes to phrases, sentences or paragraphs or to parts of audio or video files in playback. They are practicable suggestions for editors to improve the quality of written and recorded journalistic productions, and which editors are free to implement or not. Because the annotations are immediately executable and based on the principle of journalistic objectivity, they overcome the known issue of lengthy debate due to subjectivity that arises with regular reader comments.
Furthermore, provided these annotations are clear and factually accurate, they provide the necessary motivation for their immediate implementation, given that doing so will only improve the quality of the work in question”, states the organization’s website.
MMGA has received a fairly promising amount of media attention since its launch, and NU.nl’s editor-in-chief, Gert-Jaap Hoekman, was invited to explain the organization’s mission on a journalistic radio show on NPO1 (one of the Dutch public broadcaster’s national stations). But such evidence of interest is neither a cause for immediate celebration nor a guarantee of success. Remember “citizen journalism”, the once popular idea that was predicted to unseat regular journalism?
Citizens with notepads and cameras, it was said, would capture reality in a much more complete way than their professional counterparts had done. In some respects, the predictions were right, because regular members of the public are everywhere. Thus, much of the video documentation of the events of 9/11 was the work of the public. And the same has been true of other major events ever since, such as the terrorist attacks in Madrid, Paris, Brussels and London and the assassination of Theo van Gogh in Amsterdam.
The potential of this new concept to change the face of news reporting attracted the attention of established news organizations, who promptly launched crowdsourced news initiatives of their own, putting citizen journalism on the map in several countries across the world, with attendant press releases declaring the dawn of a new era. Before the idea fell out of favor, for a multiplicity of reasons, and the initiatives either petered out, or degenerated into outlets for the attention-hungry, such as Ik op TV (“Me on TV”, by Endemol), or were taken over for harvesting, as was the case with the Emmy-nominated, multimillion-dollar-funded NowPublic, the Canadian social news website that was taken over by a US investment company before ceasing to be altogether. Or they evolved into the very thing they had set out to replace, such as South Korea’s OhMyNews, which became OhMyNews International; watch out, New York Times, here we come.
In that sense, the naysayers were right all along: journalism is a profession, fortunately, as is publishing.
What about Bellingcat?
News is like any other product. If it fails to meet our expectations, we simply buy something else. And even when it is “free”, we decline its offer if we deem it unattractive or unreliable. Once that happens, advertisers and financiers flee and the outlet collapses. It’s pretty straightforward.
But, is that really the fate of all crowdsourced news initiatives? Not necessarily, and, in response to above examples, fans of the open-source model of development are no doubt ready to counter with: “Well, what about Bellingcat?”
By all accounts, no professional news organization has revealed as much about the shooting down of Malaysia Airlines flight MH17, Syria’s civil war, the Skripal poisoning or, more recently, the murder of the Arab journalist Jamal Khashoggi as Bellingcat.
Yet Bellingcat’s founder, Eliot Higgins, has himself stated that “I do not see myself as a journalist, but more as a researcher.” And Bellingcat is open to everyone, and is not tied to a particular medium.
It does, however, illustrate the potency and validity of the idea of a knowledgeable public contributing to the provision of better information.
And it’s a contribution that offers much more effectiveness when properly harnessed than readers’ letters, cancelling newspaper subscriptions, deleting a news website from our bookmark folder or even reporting inaccuracies to the Press Council.
Inclusiveness in the broadest sense of the word
The potential of this idea goes beyond journalism; in fact, any organization or body that provides information as a ‘public service’ could benefit from it, be they governmental institutions or museums. And it is arguably becoming increasingly important to use the openness of the internet to facilitate the representation and participation of diverse and hitherto underrepresented groups in media and society at large.
Editorships, newsrooms and the army of opinion leaders typically reveal a skewed distribution in their composition with respect to gender and place of origin and residence, among other things. Whereas Make Media Great Again, with its “diversity panels” geared towards the nuanced use of language in journalism and its emphasis on multiple perspectives in reporting, holds the possibility of genuine balance.
True quality is arguably impossible without diversity
“We find it important that our group of annotators is as diverse as possible. Men, women, people from various ethnic backgrounds and minorities of all sorts. This minimises the chance of overlooking particular contexts. A more diverse group can, according to scientific research, improve the quality of news offerings and build trust in the sources of these offerings. Trust, in particular, is now one of the major issues in mainstream journalism. The study that yielded the findings involved globally recognized names such as the New York Times, the Washington Post, the Wall Street Journal, the BBC and, last but not least, The Guardian. We therefore invite anyone who shares our concerns and wants to help to contact us,” said a representative of MMGA’s project management team.
Newsroom diversity is another matter
Thus, the imperative of diversity is clear. Article 7 of the (apparently) forgotten journalistic ethical code, the Bordeaux Declaration, more or less argues for it in its directions regarding more conscious reporting.
But while some elegiac reports see diversity among annotators (which may also be seen as an expansion of editorial staff by the public) as an opportunity for media in general to become “less male and less white”, we mustn’t get carried away by our hopes. As the media critic John Oliviera warns: “Having a diverse army of annotators will not inevitably make editorial offices more diverse. In fact, you run the risk that news organizations will relax any efforts in this direction, and leave the responsibility to other organizations (like MMGA). But let’s hope for the best.”
In accordance with the principles of the Internet Society
MMGA’s mission: “To offer the general public a transparent means of checking news content for the credibility of sources, balance in reporting, validity of facts, arguments, interpretations and conclusions, including grounded formulation of assertions.”
MMGA therefore sees diversity as a means of improving the quality of published content, rather than an end in itself. For the Internet Society, too, “inclusive discussion and decision-making” is a spearhead for the future of the internet. These may sound like lofty ambitions, and some reading this may be wondering about this initiative’s capacity to harness the technology needed for achieving its goals. So, let’s talk a bit about the initiative’s credentials.
Subsidy-funded start-up
MMGA operates under the wing of the Bema Foundation, which aims to promote and raise the level of public debate in the Netherlands (and if possible outside it as well), and which founded the European Press Prize, now managed by an international consortium that includes the Thomson Reuters Foundation and The Guardian Foundation.
Since 2017, Bema’s management has been the responsibility of its founder Adriaan Stoop, member of the Dutch nobility, lawyer, business advisor and o chairman of the supervisory board of NRC Media; Kees van Mourik, a digital entrepreneur and publisher with many start-ups to his name; and MMGA initiator and project leader Ruben Brave, a pioneering internet strategist, founder of valorization incubator Entelligence and social entrepreneur. MMGA partnerships include VU University Amsterdam, the University of Amsterdam, and a variety of relevant media organizations. The foundation operates on a subsidy from the Strong Internet for All Fund (SIDN Fund), and MMGA was recently by the Dutch Journalism Fund (SvdJ) for its 2018 accelerator program for funding and innovation support.
Of course, the backing of such institutions says nothing about the integrity of our annotations. And we have been as guilty as others of banging the blockchain drum in our enthusiasm for a once promising innovation that has now become a solution in search of a problem. That said, it does have its uses, and it functions in our model is as a “control layer’” for the storage and security of our annotations, a measure that accords with the principle of “Who watches the watchers”.
Artificial intelligence (A.I.) is an equally useful tool. Studies on the problem of journalistic inaccuracy suggest that A.I. could “teach” machines to identify the indicators of “fake news”. This would allow annotators to work even more effectively, and making the idea reality is already one of our objectives.
That said, we are neither unaware of A.I.’s potential limitations, nor of the news that the fight against fake news is already has already become a cat-and-mouse game. If we can deploy A.I. as a tool for identifying fake news, so can its disseminators for the purpose of identifying which types of disinformation have the highest chance of escaping detection. There are people in St. Petersburg working on this as we speak.
Thus for us it is likely to remain no more than a tool, and our core activities will stay the same: the screening, training and mobilization of media critics and critical readers to improve the quality of published information and address mis- and disinformation.
Pilot test with NU.nl
The fact that media organizations themselves are beginning to admit the need to fight fake news to maintain their readership’s trust opens the door for collaborations. And this is how we hope to work, too. After all, the idea isn’t to destroy existing organizations but to improve the quality of what they produce. Which is why we are running a pilot study with NU.nl, which with 7 to 8 million visitors a month is by far the most important news service in the Netherlands.
Selected, screened, trained: new or critical readers
Up to 50,000 readers are involved in the pilot, with candidates individually selected from the news organization’s readers’ commentary panel (NUJij).
“An annotation should be viewed as a critical comment,” explains project leader Ruben Brave at the SvdJ. “For example, an annotation might be reader indicating that an item is missing a source, or that the phrasing of a particular line betrays an implicit bias.”
The system differs from the well-known response form, whereby the reaction usually concerns disagreement with the paper’s opinion, or the tenor of the whole article. Annotations focus on specific elements of an article and are structured according to annotation labels.
MMGA has also conducted a number of other tests with NU.nl, which aim not only to test the annotation system itself, but also see how those involved respond to and work with it.
Because, as we found out quite quickly, reliability is about people, not just technology, and the relationship between editors and annotators is a fragile thing.
Fortunately, most editors are well aware that constructive criticism is the best way to improve your offering, and ultimately increase the trust of your readership. But they are and remain responsible for their reporting. And while anyone may have constructive criticism, it still needs to be delivered with respect.
Want to see all MMGA annotations or become a certified MMGA annotator? Download the MMGA Google Chrome extension via: bit.ly/MMGA-tool & register at MMGA.io!
