Overnight from 3rd to 4th July 2018, the two new laws proposed by the French government to tackle Fake News were adopted by the National Assembly. This nocturnal discussion, during a season heavy on sports news, suggests some reluctance for a law that claims to verify information to be exposed to verification itself.
This self-contradiction is ironically heralded by some coincidences of dates: the French government historically enshrined in law the idea of a defined truth, controlled and authorised by the State, on the day when, 242 years previously, the founding fathers of a young nation had declared its independence on the principles of individual freedom, directly descended from John Locke and the legacy of the Enlightenment.
The logical pitfalls of self-referencing are merciless: whoever claims to be able to objectively draw a line of demarcation between the true and the false will sink into the worst lies, whoever claims to use unambiguous criteria to establish the boundary between manipulation and sincerity will turn out to himself be the worst manipulator.
In search of an impossible definition
Our new truth vigilantes encountered a problem from the outset which they had not realised was fundamental: that of defining exactly what fake news is.
The first draft of the law included the following wording: “any allegation or imputation of a fact without verifiable elements likely to make it plausible”.
There are two shortcomings in this definition. The first is that the verification of a fact is never clear cut, except in basic cases such as announcing a death or when a disaster happens. Even the “verification” of an economic statistic (unemployment, inflation, etc.) is not a simple operation, as the criteria for measurement can be open to discussion and interpretation. It is impossible to consider facts – or figures – in isolation in economics.
The second is that a truth or even just a likelihood does not consist of a collection of accumulated facts. Revisionist discourse, for example, very often relies on verifiable and confirmed facts, but with only the convenient parts selected. Moreover, their proponents do not present them alongside competing narratives which order the facts according to a chronology that might be more plausible than their own.
The only merit of this first definition is to have highlighted the relevant ideas – verification and plausibility – to show that a law like this plunges us from the outset into the depths of logic and the scientific theories of verification that have preoccupied the last two centuries of modern epistemology.
However, the result clearly shows that there can be no objective and a priori criterion for the veracity of a statement.
Our only resource is to compare competing statements and rank them from most to least likely, an assessment that can only be done a posteriori, that is, without any prior censorship. Any student of contemporary logic or epistemology could have warned the government about this impossibility.
They must have been aware of the inadequacies of this definition however, as a second version appears in the text voted on 4 July: “fake news” is “Any inaccurate or misleading allegation or imputation of a fact”
Our budding legislators, at least the more thoughtful ones, realised that it would be extremely difficult to find an objective criterion of demarcation. So, they’ve tried a remedy used for over a century – incorporating the author’s human intention in the judgement on the statement.
Once again this attempt fell flat. Apart from mathematical truths, any statement is “inaccurate” to some degree of precision. Who will set the a priori threshold for permissible inexactitude? As for the “deceptive” nature of the statement, how can an error made in good faith be distinguished from an obvious deliberate lie? Except in the very rare cases where proof of the production of a forgery has been provided, for example in the Dreyfus case, it will be impossible to prove there is a deliberate attempt to lie. Most authors of fanciful conspiracy theories are firmly convinced of their outlandish ideas, they can be perfectly sincere in their claims.
The erratic course of our legislators, not knowing what definition to adopt and groping around desperately for one would make any student of logic smile, for the same reason that a physicist would be privately amused at a handyman in search of perpetual motion or a professional mathematician who was shown scribbles supposed to demonstrate the squaring of the circle. When an elite who take themselves very seriously is fussing every which way round an intrinsically insoluble problem, someone who knows the impasse well may enjoy the ridiculous spectacle.
One might object that many European countries have proposed similar legislation. And yet, only one country has taken the leap: Germany in January 2018, with the NetzDG, which has a very different content: the excesses targeted are explicit calls for hatred and murder. This criterion must in fact be much more provable in a court of law: a call to murder is not speech, but an offence, which must be punishable. Italy and the United Kingdom are working on legislative proposals, but face the same fundamental difficulties as their French counterparts. Only the European Commission seems in a hurry to legislate on the subject. It is true that in the past it has shown a great appetite, or rather a great thirst, for undemocratic means of control.
Small minds lost on the shoulders of giants
The problem of an a priori criterion to distinguish between true and false has been approached in different ways, but can be summarised by a three-way discussion between some giants of contemporary epistemology: Karl R. Popper, Willard V. O. Quine, and Thomas Kuhn. Addressing this problem seriously calls for an understanding of the nature of their dialogue. It can be explained in a few paragraphs to non-specialists:
Karl Popper pointed out that there can be no neutral observation and although he was far from being the first to do so, his demonstration remained one of the most striking. Knowledge does not consist of an accumulation of observations, as a simplistic model would represent – we would ingest the world through our senses and spit out a thought at the other end. The predictive and generalising power of science does not consist of repeating previous experiments, precisely because scientific knowledge makes it possible to predict a phenomenon that has never been observed before but that subsequent experiments will confirm. A scientist’s theory is not the same as a reporter’s testimony.
Any observation, even the most simple, passes through a considerable number of sensory and cognitive filters. We only observe what is within the limits of the frequency of sight and hearing, as well as at the scale that is available to us, and in fact variations of scale can bring contradictory information: a barrier which is impenetrable at a certain scale becomes crossable when considered at another.
Finally, cognitive biases are even more important than sensory biases. When we observe a scene, the highlighting of important versus inconsequential points depends to a great extent on the observer and what he seeks to perceive, mainly because of the conditioning of natural selection. When we observe nature, we always have a question in the back of our minds that we want to ask of it.
This in no way makes Popper a relativist, it was he who established a criterion of demarcation of scientific knowledge. Popper only said that our knowledge was not the direct result of “neutral” observations.
The scientist is by no means someone with a superior ability that makes his observations more objective. The theories he puts forward are in no way free of emotion or affects. It is people far removed from science who believe in such a stereotypical image.
What allows scientific knowledge to extricate itself from the mire of passions and prejudices is not to be free from them at the outset, but to use a test procedure that will pass them through a ruthless sieve, which will burn off any errors of judgement in the crucible of correction. Scientific knowledge is not born pure, it is forged by incessant verification forcing us to look again and again at our theories. An initial statement may be as fanciful as a fable; it will constantly be reinforced by successive corrections and errors.
A scientific statement offers itself up voluntarily to refutation: it invites bombardment from opposing arguments. An ideologue would see this as a sign of weakness, as they do not tolerate being contradicted. On the contrary, someone who voluntarily exposes himself to contradiction is strong, because it allows him to constantly improve.
Popper thus dispatches with equal alacrity both the relativist who holds that all ideas are equal and the fundamentalist or ideologue who claims to possess the one sole truth.
While Popper’s model was the framework of contemporary epistemology for several decades, it would itself be subjected to the refutation test.
Willard V. O. Quine highlighted a flaw in Popper’s principle. The same argument that allows Popper to show that there is no neutral observation can turn against the disproving facts that are used to test the theory. Just as one cannot have a direct experience of reality, one cannot summon an isolated and pure piece of it to be used for our verifications. Every theory includes in itself an interpretation of its disproving facts.
The Soviet model, for example, lasted a very long time despite the emergence of facts that constantly contradicted it, but which were “explained” as the result of not having proceeded far enough in Sovietism. When something does not work, either the principle is called into question or the failure is considered to be due to the fact that not enough of something has been added. Worship of the European Union is an example of this logic.
Any theory can eventually be refuted by facts, but these do not act as external attacks bombarding the model: they induce internal contradictions forcing the model to rejustify its own coherence. A theory eventually collapses not because the facts have destroyed it but because it crumbles under the weight of its own inconsistencies, which facts have caused to resonate in sympathy. The disproving fact has to be seen as a vibration which makes the model vibrate until it bursts, more than as a particle impacting on it.
Eventually, Thomas Kuhn pointed out that Popper’s principle assumes that all opposing theories are comparable, allowing them to be arranged in order on a scale of likelihood. But a scientific theory is also a representative model of the world, a universe in itself, a “paradigm”. These universes do not permit the same metrics for testing them. Relativity theory and quantum mechanics, for example, represent the world according to opposing and contradictory theories, to the extent that to compare them simply no longer makes sense. It is only their final result which separates them, as the history of science has not decreed either one as the winner, relativity or quantum mechanics both being very effective in non-overlapping fields of physics. Two high-level scientists can thus be of diametrically opposed opinions without one of them being wrong. The dialogue between Einstein and Bohr is the best illustration of this: their totally opposite visions of physics in no way prevented the two men greatly respecting each other.
What is the consequence of this epistemological framework for “fake news”? With Popper we know that there is no one point of view above all others until it has proven itself. Intellectual honesty is defined as having the humility to recognise that our own idea is tainted with bias and imperfections and the courage to submit it to the acid test. There is no other means of proof and verification: it is part of the human condition.
Banning opposing ideas, however odious we may find them, deprives us of the chance to refute them and pushes us into an even worse lie: the prohibition of a theory does more than deceive our fellow men, it would have us aspiring to “God’s point of view”, according to an ironic logician’s expression.
With Quine we see that factual refutation is much more complex than it would seem: “fact checking” is a simplistic version applicable in a small minority of cases. Either it is a straightforward matter, such as rumours about the death of a person or the occurrence of a natural disaster, which can be refuted by simple basic facts. But it is not this kind of news that pollutes social networks, but much more complex theories or news whose means of verification are unclear. Fake news broadcasters are smart enough to choose news that cannot be refuted by a basic check. Consequently, “fact checking” always misses its goal. It only clarifies straightforward events, with anything warranting involved debate just making its shortcomings obvious.
The intermingling between a theory and the facts to disprove it requires a much more complex verification procedure than a simple checklist. The basis for testing the solidity of a theory lies in four procedures
- Check that it is not contradicted by simple elementary facts (fact checking).
- Check that it explains as much of the facts as possible: many outlandish ideas pass fact checking but only present the facts that suit them.
- Order facts according to a chronological narrative: Consistency does not consist only of checking each fact in isolation: the sequence of the facts and the causal explanations put forward are part of establishing plausibility.
- Compare the theory with other competing theories in order to select the most solid, the one that explains the most phenomena.
Some examples from concrete information:
- Some conspiracy sites denied the collapse of the twin towers on the basis of “scientific” arguments, namely that the temperature of a fire triggered by the impact of an aircraft could only be lower than the melting point of the steel holding up the structure. Any physicist can easily refute this argument: at the impact temperature, steel’s pliability is significantly increased. It doesn’t have to be liquid for the towers to collapse, if it already has the consistency of a cheese at room temperature. If we ban a conspiracy theory, we miss the opportunity to ridicule it by showing the pseudo-scientific nature of its argument. On the contrary, the ban will give it an illicit appeal leading to the opposite effect. Rebuttal is always a hundred times more effective than censorship, because it allows everyone to form their own opinion.
- When Aleppo was taken back by Loyalist troops, many newspapers presented this military operation as the crushing of courageous resistance fighters located in East Aleppo by Bashar El-Assad’s evil brutes, supported by Russian troops from West Aleppo. And they focussed on the bombing of hospitals housing injured children, which is factually true information. This presentation neglects to specify that the military action in the districts of East Aleppo was conducted by leaders claiming to be Daesh, that they had carried out similar bombardments on West Aleppo fifteen days earlier, including on hospitals sheltering children, in strikes directed at Christian districts. It was also proved that the shelling of hospitals in East Aleppo was as a result of the use of human shields. So some condemned the brutality of the Russian strikes despite the use of human shields … before realising a few weeks later when Mosul was retaken that their own camp had not managed to do any better. The need to include all the facts and to place them in a chronology modifies the meaning of each unitary fact. It is possible to propose a competing analysis without in the least being a supporter of the bloodthirsty Bashar El-Assad, but one might also point out that an explanation couched in terms of good and bad has very little explanatory power.
- During his candidature for the 2017 presidential election, Francois Fillon’s legal and moral failings were proven and verifiable facts. Their significance may vary, however, if you consider that a majority of political leaders, particularly among his rivals, had committed comparable offences. And as for the chronological account, it ought to mention that in living memory no legal procedure has ever been conducted with such unseemly haste. Once again this is not a question of denying or glossing over the faults committed. But they take on a different meaning when you contextualise them amongst the other facts and reconstruct their history.
- Economics is an inexhaustible source of “factuality” traps. For example, it is a fact that inflation has remained very low for years. Some economists have been surprised at this, because monetary creation has never been so strong, given the practice of “quantitative easing” implemented by central banks to “support the economy” or to put it bluntly, to avoid a financial collapse. Several studies have shown that liquidity injections into markets have created hidden inflation, noticeable on the prices of financial assets or on the property market. But it is not obvious, as inflation is measured by a consumer price index that only covers current goods and services, well downstream of the monetary exchange circuit. Fact-checking in economics seems simple, because what could be more elementary than putting one figure next to another? Should those who claim that there has, in fact, been inflation be prosecuted for broadcasting fake news? Any economic figure raises the question of its measurement criteria and the other indicators which it is interdependent with: the economy is the ultimate area where an isolated fact makes no sense. The same applies to unemployment, poverty, demographics, etc.
- One argument often heard by proponents of “the joy of globalisation” is that while it has impoverished the lower classes in developed countries, it has raised the average level of wealth globally. But suppose that on a scale of 0 to 50, the lower classes in developed countries dropped from 12 to 9, while those in the world as a whole rose from 5 to 7. And that below the level of 10, we were to see the emergence of the political structures of non-democratic countries (lawless zones, communitarianism, organised crime), as democracy relies on there being middle classes which are strong enough. The preservation of civil liberties and those of century-old democratic traditions are seriously threatened, which is something that no economic index measures, and with the political structures of unstable countries spreading to the whole planet, it could trigger a widespread step backwards in growth. The problem with reasoning based on an average is that it ignores the threshold phenomena, activating or deactivating barriers that are commonplace in economics. Will anyone who dares to criticise the benefits of globalisation as it is practised be condemned by the “fact-checkers” who will limit themselves to placing one figure next to another?
Are ethics dissolving into infantilisation?
After Popper and Quine, what was Kuhn’s concrete contribution? We could say that this is a fifth principle added to the other four: it is only individual free will which can consider which theory stands up best to reality. Kuhn showed that each theory is a mental universe which cannot be directly compared to others. As a result, contradictory or even totally opposed theories may coexist. As in the case of Einstein and Bohr’s dialogue.
The time is long gone when the quality press could present two or three opposing but all equally respectable ideas on the great social topics. The obsession with chasing fake news relates to an impoverished concept of journalism which holds that there is only one respectable and truthful line on economic and social subjects, all others being tainted with bias. There would be only one objective point of view and any idea challenging it would be pitched in amongst the ideological points of view. By an extraordinary coincidence, the acceptable interpretation is always that of the constituted powers, at a national or European level.
Behind the belief in “fact checking”, the standardisation of information is rapidly taking shape through a system of notations. The green, orange and red “flags” of the “decoders” at the newspaper “Le Monde” make this leap from being sole possessors of factuality to attribution of value points for any opposing idea. How can one not draw a parallel between this infantilising rating system and the “scoring” set up by the Chinese government, making it possible to calculate a “good or bad citizen” rating accessible in real time by the government and constantly updated.
It would be laughable, were it not that such infantilism reaches the very highest levels of state. The “decoders” are light years away from imagining the counter-truths and critical threats to freedom that their actions entail. Any discussion of verification and evidence plunges us directly into the works of Karl Popper, Wittgenstein, Bertrand Russell, Tarski, or Gödel.
The “decoders” thrown in at the deep end with Western analytical philosophers amount to much the same as French TV comedian Cyril Hanouna up against Einstein and Poincaré or the parachuting of jolly tourists in flip-flops and t-shirts to the summit of Anapurna.
Major and minor totalitarianism
Admitting that there may be fake news is a collateral damage to our freedoms, just as a biological body will never be perfect and will always include impurities. Does this mean nothing should be done and we should let potentially dangerous gossip invade social networks? Of course, we must act, but we will need tireless patience to refute, present counter-arguments, and propose alternative explanations which are convincing because of their consistency and the seriousness of their documentation. You cannot fight wrong-headed thinking with bans because thought needs to wander, to grope its way forward, to confront itself in order to progress.
The only way to show that your point of view is better than the others is to put it into the ring with other arguments, to bring it through the acid test. Equality in the face of death and equality in the face of argument are two principles of the human condition that circumscribe merit, that avoid any arbitrariness or abuse of power. To consider oneself naturally above others, especially on the pretext of the common good, is the worst form of intellectual dishonesty, one concealing an absolute power lust concealed behind the service of others.
To equate a process of verification with a “checklist” of supposedly more objective points is deeply fallacious: such “verification” hides the fact that it is itself a theory, in order to avoid competing interpretations.
In this respect, our era sees the opposition of the two faces of minor and major totalitarianism. Their natures and roles should not be confused.
Minor totalitarianism is the danger of demagoguery and populism. Of pub chat, beery and short-sighted. While it is not always the source, it is often the vector of propagation of fake news.
Major totalitarianism is much more dangerous, in the sense that it comes from people who think they are enlightened and have got it into their heads to re-educate the populace and set them right. Exaggeration? Here is what the French Minister of Culture said on the occasion of the reading of the “fake news” bill by the cultural affairs committees:
“Do not give in to demagoguery, by only relying on citizens’ capacity for discernment”
It would seem that the government is granting itself the right to judge the discernment of citizens and is already indicating which press is “authorised” to put them back on the right path, the “decoders” and “fact checkers” having been highly praised by Françoise Nyssen a little earlier in the same text.
A pub bore can sometimes be dangerous, but a Lysenko is much more dangerous. Françoise Nyssen should remember that in the former Soviet Union, the black market far exceeded the official economy, because what you ban is immediately endowed with an irresistible attraction. Those who rail against populism do not imagine for a second that they are the first to be responsible for it, in forbidding any opposing ideas out of an appetite for power.
The recent scandal which the French government has become bogged down in is the best illustration of this, and the valiant defenders of “true information” took this opportunity to engage in a veritable orgy of “fake news”, providing a prime example of the infinite possibilities of transforming a “factual” discourse into a series of self-interested lies. They are falling apart under a hail of blows from opposing viewpoints, all equally factual, but more plausible. QED
Contempt for the people is not the sign of the powerful and the enlightened but of the weak. Truly enlightened people, those whose point of view stands out, are always very simple, because they know all too well that we are all equal in the face of argument and that we have to bring out our proofs every time there is a new debate. George Washington retired from politics when he realised that his influence and aura made him a reference and that his point of view was no longer robustly contested.
This was the heritage of true freedom of expression from the founding fathers of Philadelphia. 11 years after the Declaration of Independence, they wrote a preamble to the Constitution beginning with the words “We, the people”. Populism? No – the “Common Decency” that both the most humble and the most powerful can show when they are committed to an ideological debate worthy of the name, worthy of the rights of man – every man.
This post is also available in: FR (FR)DE (DE)