The “Illusory Truth Effect”: le monde entier est un cactus

If a statement is repeated enough times, does it become more true? We don’t like to think of ourselves as susceptible to such an elementary tactic of altering our beliefs, but if we were immune to such tactics, why is repetition such a constant feature in marketing or politics? The short answer is because it works. The “illusory truth effect” exists, and the mere presence of repetition suffices to alter what we accept as “truthful.” An overview of this state of affairs can be had in this episode of “With a Side of Knowledge,” a University of Notre Dame podcast featuring researcher Lisa Fazio from Vanderbilt University.¹

Fazio’s research into the relationship between repetition and perceived truth has two main points: the influence of repetition can be observed at a very early age and continues on through adulthood (Fazio’s study includes a group of 5 year-olds); and, prior knowledge of a truth does not prevent its alteration through repetition (a phenomenon known as “knowledge neglect”). From the podcast cited above, Fazio states: “[…] repetition increases truth even when it directly contradicts what you already know. So among people who can tell you that the skirt that Scottish men wear is called a kilt, when they hear the skirt that Scottish men wear is called a sari twice, they think it’s more true than if they’ve only heard it once.” She and other researchers have noted that there isn’t widespread agreement on why this works (explanations range from “ease of processing” to differences in fonts²), but researchers agree that there is sufficient evidence to demonstrate that the effect does indeed exist and that it “generalizes across a remarkably wide range of factors.”³

It is worthwhile to note that the illusory truth effect studied by Fazio concerns what one might call “facts” or very elementary forms of “knowledge”: Is the skirt worn by Scottish men called a kilt or a sari? Is the largest ocean on the globe the Pacific or the Atlantic? Do potatoes grow above ground or below ground? This is not to criticize the validity of such research, nor its effect in areas such as advertising or the spread of political and commercial propaganda. If repetition does indeed enhance perceived truth value, then this is a powerful tool for the manipulation of populations, especially in large, social-media oriented consumer societies.

One might think that the rather obvious question of “source credibility” would counteract repetition’s effect on perceived truth. This turns out not to be the case. In the same podcast cited above, Fazio laments:

Unfortunately, people don’t seem to pay as much attention to source credibility as you might hope. So it seems that just kind of raw repetitions, regardless of source credibility, are what matters. […] And it doesn’t matter that they’re coming from these kinds of disreputable sources.⁴

The general conclusion one can draw from Fazio’s research is a simple one: “facts” are unstable entities. Any fact you may think you know has been influenced by how many times you’ve heard it. It may quite literally be “as dumb as that.” To the extent that our views about the world are composite views which integrate a wide variety of “facts,” one has to wonder how the phenomenon of “illusory truth” colors our picture of the world:

• It’s called a sari.

• The Atlantic is the largest ocean on the globe.

• Potatoes typically grow above ground.

• Le monde entier est un cactus.

 

Perceived Truth and Belief

Is there any knowledge without a knower? What you know about the world has as much to do with you and your beliefs as it does with any object that is thought to exist independently in the world. James Burke puts it this way in the preface to his series, The Day the Universe Changed:

You are what you know. Fifteenth century Europeans ‘knew’ that the sky was made of closed concentric crystal spheres, rotating around a central earth and carry the stars and planets. That ‘knowledge’ structured everything they did and thought, because it told them the truth. Then Galileo’s telescope changed the truth.

As a result, a hundred years later everybody ‘knew’ that the universe was open and infinite, working like a giant clock. Architecture, music, literature, science, economics, art, politics—everything—changed, mirroring the new view created by the change in the knowledge.⁵

Stating the obvious, knowledge is never just knowledge but results from the application of certain lenses which validate the truth as such, as truth. We’ve seen above that “repetition” is part of those lenses. Furthermore, there is always an eyeball peering through the lens; entire cultures may end up peering through only a handful of historically or economically leveraged lenses. Truth creation in such a context involves at least three components: an observer, the object being observed, and the “lenses” which validate the relationship between observer and observed as one of “truth.” Referring back to Burke above, when people went about making “truthful” statements in 15th century Europe (“The sky is made of closed concentric crystal spheres, rotating around a central earth and carry the stars and planets.”) they were applying the lenses of their time (Aristotelian and Ptolemaic authority). The structure for truth creation in the 15th century included a powerful component of reference to authority, which formed part of the optic through which truthful statements could be grounded. If the biggest names in astronomy, supported by theologians and the Catholic church, say it’s concentric spheres, then it’s concentric spheres.⁶

In the Galilean example, the telescope is the tool through which the universe effectively changes: “The universe is open and infinite, working like a giant clock.” Yet it would be incorrect to think that this is a purely objective statement. Galileo’s process for revising the truth makes the following unstated claims: a) Visual examination is the right tool for the creation of knowledge; and b) Measurement (based on visual observation) is a neutral process which has no effect on the object measured and the knowledge thus created. These unstated claims are so much a part of how we still go about seeing the world that it is very difficult to believe that they are not “objective.” Piercing the veneer of this ostensible objectivity requires that one take a step back and try to “see” the act itself of “seeing.” In a post-Galilean world, the telescope and the act of measurement frame truthful statements about the universe. As silly as it is, consider the following: Do the blind have no direct contact with truthful experiences as they go about their daily lives? Or far less silly, this: When one carries out an act of measurement, does this not limit the nature of the object under study to “something measurable”? The object itself is not, in essence, “measurable” until one’s framework for understanding it makes it so.⁷

In Galileo’s world, however, “seeing is believing.” If Galileo sees and measures something that doesn’t fit into the current explanation of “how things are,” then that particular explanation is replaced with another. Or in Burke’s words: “the universe has changed.”

Burke’s phrasing may strike many readers as over-stated. One might object: “15th century Europeans simply had it wrong. The universe didn’t change!” Burke’s thesis is indeed a very bold sounding claim until one considers the following: Between a) knowing “how the universe really is” (does anyone know how the universe “really” is?) and b) culturally relevant beliefs about how the universe actually is (however “fictional” these beliefs may be)… Between these two ways of thinking, which is more consequential in terms of its effect on real human beings living real human lives? However fictional the belief that there are actually witches running around in Salem is completely immaterial. The belief itself that there were was far more consequential to women living there in the late 17th century. The point here is a simple one: we live less in a world than we do in a world of ideas about the world. Truth is a human and subjective phenomenon, and following Burke, the universe does effectively and regularly change.

 

Truth and Sociality

How does it sit with us in today’s “ideas about the world,” right now in 2022? It is not an exaggeration to say that “the universe” has again effectively “changed.” (For reference, I finished graduate studies at Berkeley in 2002, during which I “discovered” email). The dynamic itself of subject to object (i.e. of observer to observed) has undergone significant alterations that have significantly changed the nature of what it means to “know.” How much credibility, for example, does one give to what one sees (or reads) on the internet? Even further, to what extent is the ostensible object of our attention (the world), tainted as it is by ever higher degrees of “virtuality,” no longer an “object” at all? The phenomenon of “sociality” would seem to be of far greater consequence in the creation of our beliefs than accurate descriptions of an object (the world) that preexists our attempts to understand it. Let us not forget that the virtual world is itself entirely of our own creation. As we spend more and more of our time immersed within it, the fact itself of social interaction (rather than scientific observation) has become ever more consequential in the construction of what we believe to be “the truth.”

To once again state the obvious: Popular social media platforms are not the internet equivalent of scholarly publications. They instantiate intersubjective dynamics which are largely unconcerned with observation, investigation, testing of hypotheses or other forms of knowledge creation in the scholarly sense of the word. Yet social media platforms are by no means without influence on a culture’s subset of underlying beliefs. The dangers I see here are two-fold and highly intractable due to their paradoxical natures: extreme fractiousness and extreme uniformity.

One can begin with the fractious side of the equation with a scholarly paper published in the Proceedings of the National Academy of Sciences (PNAS): “The Echo Chamber Effect on Social Media.”⁸ In it, the authors’ working definition of an “echo chamber” is stated thus:

We can broadly define echo chambers as environments in which the opinion, political leaning, or belief of users about a topic get reinforced due to repeated interactions with peers or sources having similar tendencies and attitudes.

After copious statistical analysis, one conclusion of their research is:

Indeed, users online tend to prefer information adhering to their worldviews, ignore dissenting information, and form polarized groups around shared narratives. Furthermore, when polarization is high, misinformation quickly proliferates.

Polarization and misinformation… Even without the grounding in data analysis provided by Cinelli and his fellow researchers, the divisive nature of social media platforms is anecdotally apparent to most inter-nautes, much like the divergent natures of historical (pre-internet) media outlets (CNN vs Fox vs The Guardian vs Vox, not to mention the egregious neglect in the US of information sources in foreign languages). This variety is nothing new. Media outlets offering differing interpretive tendencies have always existed, and quora for substantive debate and the exchange of ideas are absolutely necessary to any functioning democracy. The danger arises when quora for the exchange of different opinions are no longer quora at all but “echo chambers” in which “knowledge” is reduced to a consumer product, which one then either “likes” or “dislikes.” Very little conceptually informed or “hard, research-grounded” reflection takes place on popular social media platforms, but this hardly reduces the quantity of “chatter” that adjudicates the apparent rightness or wrongness of ethical, political or social questions.

For instance, US Americans clearly do not agree on a variety of extremely consequential issues, such as climate change, abortion, gun control or the broad spectrum of issues that fall under the heading of “social justice.” One might casually navigate the existing offerings on social media platforms, but the tendency is to prefer (as Cinelli and his colleagues conclude) forums that already fit your preexisting beliefs. Alternatively, new social media platforms may spring up around beliefs already held by a potential client-base of users. This is of course part of how any “successful” business model works: give them what they want, or go under, which only further exacerbates the “mirror image” effect of consumers placating their own senses of self. Worse still, children may be indoctrinated into belief systems as a mere matter of a generation’s internet consumer habits. In all cases, an “echo chamber” effect is produced whereby, as the conclusion to Cinelli’s paper states, ignorance of alternative viewpoints and rapid spread of misinformation may be prevalent. To this, one can include (as Cinelli and his colleagues do) forms of confirmation bias which further cement static world views into the minds of social media subgroups. What this in effect means is that what the tech industry sold us in the early 2000’s as fabulous new means to “connect” with each other through the wonders of technology are actually productive of societal and cultural divisiveness. Social media may be the single most powerful force today for fractious pockets of “anti-sociality.”

A second danger, extreme uniformity, is a bit harder to see. It is the hidden “sameness” that lies beneath social media divisiveness. This “sameness” is obviously not one of content, but one of structure: whatever the differences in perspective may be in terms of content (competing views about gun control, abortion, climate change and other highly charged social and political issues), there is an underlying (and in my view, “overbearing”) sameness that is the technological infra-structure itself in which all of this communication (or lack thereof) is taking place. The entire system is a closed system, operating according to rules of technological efficacy in their simplest forms: i.e. Does the platform even function as such, as platform, from a purely technical perspective?⁹

This “closed system” is not at all without consequences on content. To schematize: if differences in belief are apparent in the content of various social media platforms (opinions on Twitter differ from those on Facebook, which again differ from those on Reddit, the medium in which these differences occur could hardly show a higher degree of uniformity. They are all “on line” and subject to all the parameters attendant upon the “virtualization” of social interaction.¹⁰ They are all “tech-driven.” They all require a uniform means of access (an internet service provider, a computer, smartphone or tablet, running primarily on one of only two possible operating systems). Most require subscriptions and collect the same types of user information. They all are component parts of a global, technological and economic infrastructure which now “encloses” virtually all forms of intellectual, social, professional and non-professional activity (from LinkedIn to the Stanford Encyclopedia of Philosophy). This litany of similitude is probably not exhaustive.

My point here is that uniformity to this degree is not without its own effects upon the nature and construction of how individuals and societies “see” and operate within the world, regardless of apparent differences in “manifest content” (the reference to Freud is not gratuitous). In a word, one arrives very quickly at McLuhen’s observations about the medium’s potential effect upon its messages, and all the attendant side-effects of applying a strict uniformity of framing to divergent content. What does it even mean to “frame” divergent ideas within a unitary framework? In such circumstances, is there not an underlying and therefore harder to see “sameness” that is tacitly exerting a great deal of pressure on forms of “difference,” rendering them (partially? predominantly?) illusory rather than real? One might say that this uniformity of framing is the “echo” of “the echo chamber” itself, rendering all forms of structurally-based dissent impossible. In a word, in order to buy out, one has to buy in. But how then does dissent remain dissent after “buying in”? And once one has “bought in,” well… One knows how the story ends. There appear to be few real options “outside” of this unitary framework for the production or recognition of substantive alterity, be this latter epistemological, ontological, political, social or even simply anecdotal. In a way, the nature itself of “visibility” has changed: if it isn’t integrated into the preexisting structure of visibility (as measured by “likes” or numbers of subscribers), it isn’t actually visible at all.¹¹

 

A frightening state of affairs

Some may feel that this is a lot “to milk” from a few studies on illusory truth and “echo chambers.” I’d disagree, and by way of re-grounding my analysis in the research cited above, I’d like to draw the following abstraction: repetition (brute, mechanical repetition) is not a neutral component in how we go about creating “a truth,” and that is the surprising deduction surrounding research into illusory truths, for it (re)awakens us to the possibility that “the truth” is not what one might think it is. Who likes to think that “the truth” can be anything other than fully and uniquely “truthful”? Like it or not, mere repetition is at work here; it is at work within “echo chambers,” but it is also a component part of the framework itself of truth creation as a social-media phenomenon. If repetition and echo chambers are themselves part of the uniform infrastructure for truth creation (uniformly technological, uniformly a “closed system”), then the framework contains within itself the sources of its own un-doing: its weakest component is ourselves. We who do not take the time to think, to consider, to reflect and to reevaluate. Or maybe, we just don’t care because, as I often hear, “It works for me.” It is perhaps the case that this “works for me” is the amoral, non-reflective truth of the matter. Truth is a human-psychological phenomenon, and if something as elementary as repetition can affect how we go about creating “the truth” (is it even possible to speak of “the” truth at all?), then one has to wonder what other unexpected forces are at play as we go about creating—literally creating—the worlds in which we live.

 

Thédoc Scorpo

 

Endnotes

1. Fox, Ted. “On Misinformation and Truth Sandwiches—Lisa Fazio, Vanderbilt University.” With a Side of Knowledge, Episode 4.6, University of Notre Dame, 10/22/2020. https://withasideofpod.nd.edu/episodes/4-6-on-misinformation-and-truth-sandwiches-lisa-fazio-vanderbilt-university/#transcript.

2. Fazio, Lisa and Carrie Sherry. “The Effect of Repetition on Truth Judgments across Development.” Psychological Science, 31, 2020, pp. 1150-1160. https://psyarxiv.com/36mqc/.

3. Fazio, L. K., Brashier, N. M., Payne, B. K., & Marsh, E. J. “Knowledge Does Not Protect against Illusory Truth.” Journal of Experimental Psychology: General, 144 (5), 2015, pp. 993–1002. https://doi.org/10.1037/xge0000098.

4. Fox, Ted. “On Misinformation and Truth Sandwiches—Lisa Fazio, Vanderbilt University.” With a Side of Knowledge, Episode 4.6, University of Notre Dame, 10/22/2020. https://withasideofpod.nd.edu/episodes/4-6-on-misinformation-and-truth-sandwiches-lisa-fazio-vanderbilt-university/#transcript.

5. Burke, James. The Day the Universe Changed. London Writers Ltd, 1985.

6. One has to wonder to what degree this is still very much the case and to what extent this “authority” is now subsumed under the idea of “brand name.” Cf, the proliferation of back-window “University of…” stickers in the US. In not part of their function to valorize a human mind by association with a “brand name” university?

7. For further treatment of the effect of “making measurable,” see: Peppin, Aidan, “Make Measurable: What Galileo Didn’t Say about the Subjectivity of Algorithms,” Towards Data Science, July 27, 2021. https://towardsdatascience.com/make-measurable-what-galileo-didnt-say-about-the-subjectivity-of-algorithms-8d1d324253da.

8. Cinelli, M., De Francisci-Morales, G., Galeazzi, A., Quattrociocchi, W., Starnini, M. “The Echo Chamber Effect on Social Media,” Proceedings of the National Academy of Sciences (PNAS), 2021. https://doi.org/10.1073/pnas.2023301118.

9. Economic viability is clearly another factor but not the point of this paper.

10. Obviously: heightened anonymity and reduced accountability, absence of corporeality, a “permanence” and “ethereality” which paradoxically expand time (permanence: nothing is ever forgotten) and contract time (tldr: nothing may ever actually get read), absence of orality and most forms of non-verbal communication, integration of multimedia elements (“hypermedia”)…just to mention a few. Research into the “characteristics of online communication” is so copious that it defies characterization. My own, anecdotal view is that most, non-scholarly forms of online communication fall into two broad categories: advertising, or “performance art” (in which the “star of the show” is “you”).

11. The economic variant of this “invisibility” is actually quite visible: homelessness. The rise in homelessness in the United States may be a symptom of how strictly uniform and costly (both monetarily and moral-ethically) the pathways to economic viability have become. If one cannot, or refuses (on moral-ethical grounds, think: environmentalism, for example) to follow such pathways, one may be hard-pressed to find viable alternatives to ensure even the most basic forms of economic survival.

Sources

Bacon, F. T. “Credibility of Repeated Statements: Memory for Trivia.” Journal of Experimental Psychology: Human Learning and Memory, vol. 5, 1979, pp. 241–252. https://psycnet.apa.org/record/1981-07029-001.

Burke, James. The Day the Universe Changed. London Writers Ltd, 1985.

Cinelli, M., De Francisci-Morales, G., Galeazzi, A., Quattrociocchi, W., Starnini, M. “The Echo Chamber Effect on Social Media,” Proceedings of the National Academy of Sciences (PNAS), 2021. https://doi.org/10.1073/pnas.2023301118.

Fazio, Lisa and Carrie Sherry. “The Effect of Repetition on Truth Judgments across Development.” Psychological Science, 31, 2020, pp. 1150-1160. https://psyarxiv.com/36mqc/.

Fazio, L. K., Brashier, N. M., Payne, B. K., & Marsh, E. J. “Knowledge Does Not Protect against Illusory Truth.” Journal of Experimental Psychology: General, 144 (5), 2015, pp. 993–1002. https://doi.org/10.1037/xge0000098.

Fox, Ted. “On Misinformation and Truth Sandwiches—Lisa Fazio, Vanderbilt University.” With a Side of Knowledge, Episode 4.6, University of Notre Dame, 10/22/2020. https://withasideofpod.nd.edu/episodes/4-6-on-misinformation-and-truth-sandwiches-lisa-fazio-vanderbilt-university/#transcript.

Peppin, Aidan, “Make Measurable: What Galileo Didn’t Say about the Subjectivity of Algorithms,” Towards Data Science, July 27, 2021. https://towardsdatascience.com/make-measurable-what-galileo-didnt-say-about-the-subjectivity-of-algorithms-8d1d324253da.

Stafford, Tom. “How Liars Create the ‘Illusion of Truth’.” BBC Future, BBC, 26/10/2016, https://www.bbc.com/future/article/20161026-how-liars-create-the-illusion-of-truth.