Defining Public Knowledge; Accuracy is not Enough
Research transparency is the key to a healthy public sphere
This essay — like the project it proposes — depends on the premise that there is a crisis of trust in the public sphere. A small amount of research (72 minutes worth, to be precise) unearthed multiple studies that support this conclusion.
The richest of these datasets was the Edelman Trust Barometer. This year’s edition actually reported a 2% increase in trust globally year-on-year from 2019–2020. Potentially as a result of societies pulling together during the pandemic — and as they saw American society falling apart. Australia was among those countries that saw trust gains, including in regards to the media, with 51% of respondents saying they trusted it, up from 38%. This was the lowest score, with business, the government and NGOs all seeing higher levels of trust. But other elements of their findings, especially huge drops in trust in China and the US still led them to describe “an epidemic of misinformation and widespread mistrust”.
And the fact that, for an educated country like Australia, 51% is a good annual score, shows how alarmingly low the bar has come in terms of journalists’ efforts to earn the public’s trust.
One thing Australia’s relative success (along with New Zealand, Taiwan and other countries) in controlling Covid-19 has shown is the value of consensus. Only consensus, never coercion, can drive a successful public health response to a threat like Covid-19. The police can catch a few rule breakers, and deter a few more, but if millions of people join in, they are essentially powerless.
Even in Australia the picture is mixed. With trust in both traditional media and search engines falling, while trust in social media and owned media (content produced by corporations as a direct marketing technique — like this blog) rose. It could be that the headline number disguises a troubling shift away from “mainstream” to “alternative” sources of facts.
The Edelman study also showed that, when they conducted their surveys at the end of 2020, only 64% of the global population was willing to be vaccinated against Covid-19. Only 33% wanted to do so as soon as possible.
In America the situation is terrifying. Trust, broadly defined, is collapsing among Trump voters (dropping 13 points in the Edelman index) while only holding steady among Biden voters. In terms of media specifically, both sides of politics saw trust declines. Among Biden supporters, trust in the press had declined by 3 points to 57%, among Trump voters the decline was dramatic, dropping 15 points to 18%. Pew’s long term studies show something similar.
None of this is new. Plenty of people have sounded the alarm before me, and put forth their various calls to action. These calls have been answered, with immense sums of money invested, donated, endowed and bestowed on various initiatives, projects and institutions
If we want to understand the abject failure of these initiatives to date, we might find a hint in the title given to this year’s Edelman report Declaring Information Bankruptcy.
Bankruptcy implies we are running out of information when we have too much, if anything, making it impossible to know which information to trust.
It is the difference between these two things, information and knowledge, which we must first get clear on if we are to make any progress.
What Is Knowledge?
The question of what members of the general public know or don’t know is often reduced to a question of accuracy. If we ask a representative sample of average people something and they all answer correctly, then that is something everybody knows. Or is it?
Epistemology, the area of philosophy concerned with knowledge, requires that the belief be not just true, but also justified. A belief that just happens to be true might count as accurate information, but without justification it is not knowledge.
Our hypothesis, at Stone, is that a lack of justification currently limits the rate at which public knowledge can effectively be created.
When asked to justify the belief, for example, that Joe Biden won the 2020 election, or that vaccines make us safer rather than putting us in danger, many well meaning and intelligent people can offer little more than an appeal to the institutional authority of the media they consume. This approach is, at bottom, intellectually bankrupt. Because CNN said so is not the argument of a self respecting person, and should not be taken as the final word on anything. No offence to CNN — the same applies to any news outlet.
It is to epistemology what “I vas jast fallowving orders” is to morality: a coward’s answer.
That doesn’t mean that reputation is useless. If you have read ten articles on CNN’s website, investigated nine of them and found them all to be true, then you wouldn’t be crazy to bet on the tenth being true, as well. But in reality that check comes rarely, and consists of little more than comparing one outlet’s reporting with another’s, without checking, for example, if both of those outlets are owned by the same parent company, or if they are both just uncritically repeating the same wire service dispatch.
A more in-depth check would take much longer, and many would not know where to start. What then, can we do to make checking easier?
This is what referencing and bibliographies were designed to do, after the explosion in information technology triggered by the printing press. By listing the author, date of publication, and page number when they reference a source, good faith scholars could make it easier for their peers and to check if they have done so correctly. This move to invite falsification provides a kind of strength through vulnerability upon which the amassed knowledge of the academy rests.
How can we apply that principle to the realm of digital media? How are members of the general public, realistically, to check on the media?
Our Solution
What if we pushed the burden of justification where it belongs, onto the producers of public knowledge, rather than the consumers?
That’s what the tool I used to capture my research, and display the highlights from it in a research portal at the start of this essay, is designed to do. It’s a video based bibliographic (or vibliographic, if you will) tool which scoops up the evidence as it is encountered, allowing it to be presented in a short form highlights-reel alongside the news deliverable, or in a longform vibliography which scrutineers will click through to examine.
Only a minority will examine this epistemological audit trail, but these scrutineers can raise the alarm, confronting the authors and publicizing their errors or omissions.
Our primary users — journalists and other public facing researchers — will no longer be mere sources of information (accurate or otherwise) but genuine knowledge creators, as their justifications can be somewhat inheritable by those who consume their content.
When asked how they know Joe Biden won the election, the same hypothetical person might thus be able to say “I saw a journalist from CCN go through all the cases filed by the Trump campaign, reading the verdicts — go see for yourself”.
This will empower the public, but it will also empower the journalists and newsrooms who take the time to check their own facts and get it right, rather than pumping out sensationalist clickbait.
Consider the diagram below, on the left is the subject matter- the actual reality of the situation. Extending away from it is a kind of cone of information, which diminishes in size as it moves towards the general public.
An inverse cone of attention extends in the opposite direction, away from the general public and towards the subject matter. The intersection of these two cones — attention and information — is public knowledge, with journalism historically positioned at its centre.
The digital age has democratised this picture somewhat — but not in a balanced way. Online discourse about news and current events happens after the journalist hits publish. This has dragged the centre of activity away from the journalist and towards the general public. But that has also meant moving it farther, in some cases at least, from reality.
Most of the activity in response to this issue (such as the various “trust” projects - or more recently “Newsguard”) has attempted to reverse this erosion of journalistic authority in entirely the wrong way — by reasserting selective institutional authority. Microsoft’s edge browser, for example, warns users when they access a news website deemed to be untrustworthy — but why should users trust Microsoft over the publication itself, or their own judgement regarding who is to be trusted?
Similar initiatives, like The Trust project or News Guard all end up having a fundamentally similar approach, they curate black-lists, white-lists and sometimes grey-lists, seeking to sort outlets into tiers and build a grand fortress for legacy media to starve in.
In the end, all they do is further crowd the last-mile of the news supply chain with their filtration systems.
Our approach, instead, is to create activity on the other side of the diagram — research content — giving structure and visibility to the journalist’s interaction with the story and source material.
That is why in this early conceptual drawing Stone is positioned beneath and behind the news article, whereas social media (and many well intentioned interventions aimed at fighting disinformation occur) on top and in-front-of the content itself.
We can take this analogy another step. Imagine the original digital news deliverable — a page on a website with a headline and some information, as a primitive canoe, paddled out into the rough seas of the open internet. Social media, then, is like the sail, an innovation designed to move the craft and it’s cargo faster and further. Stone, then, is like a rudder, designed to add stability.
In a public facing use case like news, Stone can pay for itself by placing advertising in the video content it produces, providing newsrooms with an additional source of revenue. This of video-without the pivot provides a short term incentive, and the trust of their readership provides a long term incentive.
We have launched with a beta version of the application for Windows (Mac users will need a virtual machine for now). A little further down the road we will add a phone app, and a cryptographic index, using blockchain or an equivalent technology to prove beyond any reasonable doubt that our archive is accurate and complete, along with a system for scrutineers to review, rate and discuss the research — with those ratings themselves weighted depending on scrutineer behaviour, including how much of the research in question they have actually reviewed.
But all this progress depends on our early adopters, who — as we respond to their requirements — will become a part of the product’s DNA. This is a chance to be part of the solution, and shape the future of news.
Many journalists, once they grasp our proposal, are instinctively repulsed, seeing it as an extension of an ugly populism that would write them off as Lügenpresse. But this is the opposite of our intention, and of the effect our tool will have.
With our tool they can deflect such claims and show they are acting in good faith. They can connect with their audience — and give them a reason to trust beyond ideological alignment. Unless of course they are lying — either about the subject matter itself, or about themselves, and their level of competence or the rigor of their investigatory methods.
The early adopters we attract will be the most talented up-and-coming researchers with something to prove and everything to gain, and the rewards of taking this unique approach will be the greatest for the first to do it. At first this transparency and accountability will be a novelty, and those who use it will stand out. This will encourage more adoption until, as usage grows, it becomes less of a novelty and more of an expectation.
Stone users will initially form an elite tier of high credibility journalists, then slowly become the new normal in terms of public knowledge production, with secrecy and the selective presentation of research artefacts as the exception, rather than the rule.
As my usage in this article shows, even speculative and philosophical works can find touchpoints in reality. But journalism is rarely a good place for big ideas, and the kind of work our system is most suited for — discreet, verifiable, factual reporting, with tidy contained evidentiary proofs — is exactly what we most need more of, to provide the foundation of shared knowledge upon which not just democracy but any civilised society depends.
