Who is the guardian of the truth?

On the same topic:

Once upon a time, in Savatthi, there was a king. He gathered all the people born blind in his kingdom and showed them an elephant. Relying on their sense of touch, each tried to discover what was before them. Those who touched the tusk said that an elephant is like a plow. Those who touched the trunk retorted that an elephant is like a snake. Those who touched the side argued that it is like a wall. Those who touched the tip of the tail argued that it is a broom. And so on, first disagreeing, then arguing more and more heatedly ("This is the truth!", "No, it's something else!") until they ended up punching each other. The king was greatly amused by the spectacle. The Buddha (Udana 6.4) tells this parable to comment on the disputes between followers of different religions. We will find it convenient to use it as a common thread in what follows. The truth is not in the elephant, nor in the people who examine it, but in the conformity between the elephant and what is said about it. Human truths are partial (and sometimes biased). Call them, if you like, conjectures (or convictions). If there is a higher truth like the elephant, we must conform to it. Seeking truth is the name we use to honor the blind's effort to ensure that their knowledge conforms to the elephant.

This article revisits the parable of the elephant and the blind men, moving it to the global village of the internet . A blind man might create a post, crafted with the goal of going viral. Another would exploit his role as an influencer, which, after all, has some affinities with that of a prophet. A third, instead of seeing with his own hands, would rely on the opinion of a generative intelligence, nourished by the statements of other blind men. Without forgetting that, at the end of the parable and ignored in the summaries, there is a king who watches with amusement. The premise is that the path to truth requires three crucial steps. The first is the production of knowledge, as the blind men do when they touch the elephant. The second is its validation, which is a necessary condition for others to trust it and make it their own. The third is the sharing of validated knowledge, which consolidates the wealth of knowledge for the benefit of all. The thesis is that in the global village, connections have increased explosively: a collateral casualty of the information bombardment is the validation of knowledge, which changes the meaning of sharing. If there were an internet in Savatthi, the cacophony of opinions surrounding the elephant (the "data traffic") would be deafening. The king, perhaps, would be even more amused, while the elephant would remain, as always, unperturbed.

The mechanisms for producing, validating, and sharing knowledge depend on the domains considered. We can distinguish three: rational truths, factual truths, and conventional truths. Rational truths are necessary and universal, like mathematical truths. The statement 1+1=2 is unquestionably true. (Jokes speculate that in the binary number system the correct statement is 1+1=10, but the meaning remains the same.) The etymology of the word "mathematics" refers to what can be learned, or discovered, by oneself: the blind need not use touch; their mind's eye suffices. Natural numbers are infinite: no one could have taught us the name of every number; yet, we know which number comes after 2025. As of October 12, 2024, the record for the largest prime number has exceeded 41 million decimal places; yet, we are certain that the record can be broken because we know that prime numbers are infinite. Mathematicians have differing attitudes toward the elephant they study. Many believe the elephant exists and that their work progressively reveals its features. According to this view, truth (aletheia) is the unveiling of ideas that exist beyond the sensible world. Others reduce mathematics to a formal construction, manipulating symbols according to rules that guarantee correct conclusions. From this perspective, a mathematician invents axioms and studies their consequences. In the fantasy novel "The Glass Bead Game," Hesse describes a system of abstractions where the best players share a sense of elegance and beauty without reference to any elephant. The mathematical model par excellence is Euclid's Elements, governed by the triplet: axiom, theorem, problem. Axioms are premises accepted as true a priori. A theorem is a proposition logically demonstrated by axioms. A problem is a question to be addressed by exhibiting a constructive procedure that leads to a solution. Theorems are the body of knowledge, which feeds the techniques applied to problems. The production of rational truths is determined by their provability. When a mathematician proves a proposition, he recognizes it as true. Validating a mathematical proposition requires that its proof be examined and deemed correct by other experts (peer review). On June 23, 1993, Andrew Wiles announced that he had proved Fermat's Last Theorem, which had remained open for over 300 years. Three months later, a fallacy emerged in the proof that took a year to resolve. When the complete proof was published in 1995, Wiles's discovery was validated and shared. Validation and sharing are essential for a proposition to be considered true by a community. I cannot verify Wiles's proof: yet, since it has been validated and shared by experts I trust, I admit the statement of Fermat's Last Theorem among my true knowledge. Experts are the custodians, protecting the quality of sources. Truths of fact are contingent and empirical, as in the physical and natural sciences. They cannot be proven beyond doubt, but they appear to be the most coherent explanations for the collected sensory evidence. While truths of reason are evaluated for their consistency with axioms, truths of fact are required to conform to the external world (or to what we can measure of it). In the parable, each blind man provides his best version of truths of fact. Each blind man produces knowledge, but there are no guardians to filter and organize it. If a blind man with a tusk were to take the hand of a blind man with a trunk and lead him to caress it, the hypothesis that the elephant is like a plow would be falsified and discarded. On the other hand, if all the blind men touched only the tusk, the community of the blind would consider the hypothesis true (until proven otherwise) that the elephant is like a plow. Factual truths often fall within an interpretative framework, which must ensure a good overall degree of consistency with empirical data. This framework is subject to incremental changes or can undergo drastic upheavals, as was the case with the Copernican revolution or the theory of relativity. The blind produce knowledge, the guardians validate interpretation. The empirical method developed by science aims to build solid interpretative frameworks, despite human limitations. Francis Bacon had already described some of these in Novum Organum. We tend to favor data that confirm what we believe (confirmation bias) over data that force us to question it. The search for truth can be distorted by spurious motivations such as power or prestige, or subjugated by authority or tradition. Bacon classifies systematic errors in the search for factual truths into four categories, astutely called idola (simulacra) because they replace the elephant the blind are called upon to examine. Idols are the tendency to imagine spurious causes or intentions; individual stereotypes and prejudices; errors attributable to language and communication codes; the uncritical acceptance of philosophical or religious systems that influence the interpretative framework. Society has entrusted the production and validation of factual truths to the scientific community. The term "scientist" was coined in 1833 by William Whewell following the growing professionalization of this community of experts and custodians. It was proposed ironically as a calque of similar forms such as artist, journalist, or economist. The expression was in opposition to the now obsolete sciolist, someone who speaks confidently about something they are not an expert on: in Italian, we could translate it as know-it-all or, with poetic license, as opinionist. The success of science and the benefits it has brought to humanity are evident. However, contemporary production suffers from overproduction. A report commissioned in 2020 by the National Science Foundation estimates that the number of scientific articles published annually exceeds 2.5 million. Such a high production raises many questions: no one can read more than an insignificant fraction of them—what are the mechanisms that direct attention and consensus to initiate their validation? Validation confirms the production of knowledge through independent replication. A second blind man who touches on the same point examined by the first blind man must reach similar conclusions. However, the validation process is slow and unrewarding. Confirming another's thesis enhances the reputation of the discoverer, not that of the custodian. Furthermore, validation takes time away from production, which is the primary indicator used to allocate funding and professorships. The king incentivizes the blind to compete rather than to seek the truth together. Conventional truths are beliefs widely shared by a community, yet lack sufficient evidence to validate them using the empirical methods of natural science. Conventional truths guide the actions of blind men when the elephant is elsewhere or even absent. The production of conventional truths is facilitated by the malleability with which facts are interpreted and by the human propensity to invent alternative explanations. Their validation is easily questioned because there is no consensus on methods or custodians. The main difficulty is described by the poet Coleridge in his Biographia literaria, where he comments on the paradoxical effect by which the growing circulation of books leads to the prevalence of opinion-formers (sciolists). When the growth of readership increases the proportion of those interested in being entertained or confirmed in their opinions compared to those who wish to know the elephant, success smiles on the authors who indulge them. If conventional truths are shared so quickly that they preempt their validation, mere popularity becomes an alternative criterion of truth. If they're being talked about, there must be some truth to them; if they're being talked about so much, even more so: vox populi, vox dei. Those who hold different positions or question prevailing convention are denied the trust that underpins the validation process. The public distances itself from experts, ending up believing what others believe without a preliminary validation phase. Conventional truths are self-fulfilling: they are believed to be true because others believe them. Remember hydroxychloroquine? On February 25, 2020, a French research institute uploaded a video to YouTube claiming that Chinese researchers had observed spectacular improvements in Covid-19 patients. Within about a month, a chain of random events culminated in a tweet from the US president declaring the drug a potential game-changer for the pandemic. The public, despite skepticism from the scientific community, agreed that the claim was true: sales of generic drugs skyrocketed, along with the number of commentators claiming its efficacy without evidence. For a couple of months, a significant portion of the public followed a group of blind people who, without ever having touched the elephant, told them what they wanted to hear. Today, the scientific consensus is that hydroxychloroquine-based Covid-19 treatments have no, if any, effects. Let's recap: the production-validation-sharing sequence is essential for organizing a community's knowledge, enabling it to understand, evaluate, and act. The explosion of connections enabled by the internet has accelerated the speed of sharing to the point of challenging validation and calling into question the custodial function entrusted to experts.

“Better to know nothing than to know many things halfway! Better to be a fool on one's own initiative than a wise man according to another's opinion” (Nietzsche, “Thus Spoke Zarathustra”).

Marco Li Calzi is Professor of Mathematical Methods for Economics at Ca' Foscari University of Venice. His article continues Il Foglio's summer series dedicated to truth. Each week, a different author will examine this fundamental concept from the perspective of a specific discipline: law, mathematics, astrophysics, economics, politics, information, or theology. "Truth, in Practice" by Michele Silenzi was published on July 15th, "Truth in the Bar" by Giovanni Fiandaca on the 22nd, "What Truth for the Polis" by Flavio Felice on the 29th, and " It Takes a Bestial Physics" by Marco Bersanelli on August 5th.

More on these topics: