Calling Bullshit: The Art of Skepticism in a Data-Driven World

In a world drowning in data and misinformation, are you tired of being misled? "Calling Bullshit" offers a refreshing dose of skepticism, arming you with the tools to dissect the deceptive practices lurking in everything from news headlines to scientific studies. Prepare to embark on an intellectually stimulating journey where you'll learn to distinguish correlation from causation, identify selection biases, and decode misleading data visualizations. This isn't just about spotting falsehoods; it's about understanding the *nature* of bullshit and how it permeates our data-driven society. With wit and rigor, Carl T. Bergstrom and Jevin D. West empower you to become a savvy consumer of information, capable of critically evaluating claims and confidently calling bullshit when you see it. Get ready to sharpen your mind, challenge assumptions, and reclaim your intellectual autonomy in an age of rampant deception.

In this chapter of *Calling Bullshit*, Carl T. Bergstrom and Jevin D. West dissect the pervasive nature of bullshit, tracing its roots from animal deception to sophisticated human manipulation. The authors begin by illustrating how even creatures like mantis shrimp and ravens engage in deceptive behaviors for survival, setting the stage for understanding human bullshit. The key difference, they argue, lies in humans' advanced cognitive abilities, particularly our theory of mind and complex language, which allow us to craft elaborate deceptions. The authors reveal the first core insight: bullshit thrives because everyone is trying to sell something, be it a product, an idea, or an image. Think of it as the modern world's echo chamber, where truth and falsehood are often indistinguishable. Bergstrom and West then delve into the nuances of paltering—misleading without outright lying—and the use of weasel words to evade responsibility, showcasing how language can be twisted to create false impressions. They present the second core insight that the gap between literal meaning and implied meaning is a playground for bullshitters. Consider the image of a politician skillfully dodging a question, leaving a trail of ambiguity in their wake. Moving beyond intentional deception, the authors explore how even seemingly innocent storytelling can become a form of bullshit when the focus shifts from truth to self-promotion. This reveals a third core insight: much bullshit isn't about deceiving others, but about constructing a desired self-image. The authors introduce Brandolini's principle, highlighting the asymmetry between the ease of producing bullshit and the difficulty of refuting it. The image of truth struggling to pull its breeches on while falsehood gallops around the world encapsulates this challenge. They then analyze the spread of misinformation on social media, using the example of the false story about an eight-year-old girl killed in the Boston Marathon bombing to illustrate how quickly rumors can propagate despite efforts to correct them. The authors introduce a fourth core insight: cleaning up bullshit requires significantly more effort and resources than creating it. They underscore the rapid changes in information sharing, from newspapers to social media, which have created fertile ground for the proliferation of misinformation. The authors present the fifth core insight that the speed and reach of modern communication technologies amplify the spread of bullshit, making it harder to contain. Ultimately, Bergstrom and West paint a sobering picture of a world inundated with bullshit, driven by self-interest, cognitive biases, and the dynamics of information dissemination, yet equip the reader with the beginnings of a critical eye. As the authors conclude, the fight against bullshit is an uphill battle, but one that is essential for navigating the complexities of our data-driven world, and the sixth core insight is this: skepticism, critical thinking, and awareness are essential tools for combating bullshit in all its forms.

In "Calling Bullshit," Carl T. Bergstrom and Jevin D. West dissect the modern infodemic, starting with a bold paradox: smartphones, meant to be bullshit detectors, have instead become bullshit amplifiers. The authors trace this problem back to the printing press, noting how each technological leap democratizes information but also floods the market with fluff. Like Filippo de Strata lamenting the printing press's 'brothel,' critics now decry the Internet’s endless stream of verbal excrement, a torrent that overwhelms our capacity for discernment. Bergstrom and West argue that the click-driven Internet prioritizes sparkle over substance; headlines, once concise summaries, now contort themselves into clickbait promising emotional experiences rather than conveying facts. They reveal how partisanship exploits this vulnerability, with hyper-partisan news amplified by social media algorithms that prioritize engagement over truth, creating echo chambers where tribal epistemologies reign. Judith Donath’s communication theory highlights that sharing information often signals allegiance, reinforcing community bonds even at the expense of factual accuracy. The algorithms, those silent bullshitters, learn to feed us increasingly extreme content, hijacking our attention and wasting our minds. The digital landscape becomes a hall of mirrors, where misinformation and disinformation thrive. The authors paint a vivid scene: a family in India, wrongly accused based on WhatsApp rumors, becomes a victim of mob violence. They explore the rise of digital counterfeiting, where bots and deepfakes erode trust and swamp genuine voices. Jenna Abrams, a fictional persona created by a Moscow propaganda outfit, demonstrates the effectiveness of disinformation spread through trusted networks. Ultimately, Bergstrom and West offer a multi-pronged defense: technological solutions, governmental regulation, and, most importantly, education in media literacy and critical thinking. The authors leave us with a call to action: to triangulate information, seek independent witnesses, and become vigilant bullshit detectors in this age of digital distortion.

In this chapter of *Calling Bullshit*, Carl T. Bergstrom and Jevin D. West dissect the essence of bullshit, distinguishing it from mere lies or simple inaccuracies. They begin by noting how the term is often used loosely, as a synonym for anything disliked, before honing in on a more precise definition. The authors clarify that bullshit isn't just about deception; it's about a speaker's indifference to truth, prioritizing persuasion or impression over factual accuracy, like a shimmering mirage in the desert of discourse. Frankfurt's perspective is highlighted: bullshit arises when one tries to impress without regard for truth. Cohen adds that academic bullshit is often 'unclarifiable unclarity,' so dense and convoluted that critique becomes impossible. Bergstrom and West emphasize that the bullshitter manipulates rather than communicates, employing rhetorical flair or statistical 'snake oil' to overwhelm the audience. They define bullshit as language or data presentation intended to persuade or impress, blatantly disregarding truth and coherence. Latour’s work on power dynamics between author and reader underscores how authors often create 'black boxes'—complex jargon or impenetrable methodologies—to shield claims from scrutiny. The authors illustrate this with an example of genetic variants associated with bullshit susceptibility, where the complex biotechnology becomes a black box. Lies, they argue, become bullshit when concealed behind rhetorical artifices, like a magician's smoke and mirrors. The chapter then turns to a case study: an algorithm claiming to detect criminality from facial images. Bergstrom and West demonstrate how, without even delving into the algorithm's mechanics, one can expose the study's flaws by examining the biased training data—criminals' ID photos versus non-criminals' professional headshots. The algorithm, rather than detecting criminality, was merely detecting smiles. The extraordinary claim of linking facial structure to criminal tendencies crumbled under the weight of a more plausible explanation: situational facial expressions. Ultimately, the authors reveal that spotting bullshit often doesn't require technical expertise but careful consideration of data and results, urging readers to question the data's biases and the plausibility of conclusions, transforming the reader into a discerning detective of data.

In this exploration of causality, Carl T. Bergstrom and Jevin D. West dissect the pervasive human tendency to confuse correlation with causation, a cognitive leap that often leads to misinformation. The authors begin by illustrating how easily we assume causal relationships where only associations exist, such as the link between self-esteem and kissing, urging us to recognize that correlation does not imply causation, a mantra against simplistic narratives. They introduce the concept of linear correlation, emphasizing that while associations can be informative, they rarely reveal the underlying causal mechanisms, like the suggestive, yet ultimately misleading, correlation between housing prices and birth rates. Bergstrom and West then navigate the philosophical complexities of causation, acknowledging the instrumental need to understand cause-and-effect for practical purposes, even when definitive proof remains elusive. The narrative tension rises as the authors expose how media outlets frequently misrepresent research findings, particularly in health-related stories, where correlations are inflated into causal claims, painting a vivid picture of how easily we are swayed by compelling stories over rigorous evidence. Bergstrom and West highlight the importance of scrutinizing prescriptive claims, especially in medical journalism, where advice is often based on mere associations, further cautioning against the post hoc ergo propter hoc fallacy, the assumption that because one event follows another, the first caused the second. They dismantle the misconception surrounding the marshmallow test, revealing how parental socioeconomic status, a common cause, influences both a child's ability to delay gratification and their later academic success, challenging the notion that delayed gratification directly causes success. The chapter crescendos as the authors unmask spurious correlations, those chance alignments that lack any meaningful connection, such as the whimsical correlation between the age of Miss America and murders by hot objects, cautioning against data dredging, where endless comparisons lead to coincidental similarities. They critique arguments that deny causal relationships by conflating probabilistic cause with sufficient or necessary causes, using Mike Pence's flawed logic on smoking as a stark example of how statistical truths can be twisted. Finally, Bergstrom and West offer a resolution: manipulative experiments, when ethically feasible, provide the strongest evidence of causality by isolating the purported cause and controlling other variables, referencing experiments on fever to illustrate how scientists can tease apart correlation and causation, ultimately empowering us to become more discerning consumers of information.

In a world saturated with data, Carl T. Bergstrom and Jevin D. West cast a critical eye on the seductive power of numbers, revealing how easily they can be manipulated to mislead. The authors highlight that while numbers project objectivity, their interpretation is always subjective, shaped by context and presentation. Bergstrom and West illustrate how even seemingly exact counts can be flawed due to miscounting, sampling errors, or biased procedures, pointing out that the choice of summary statistics—mean versus median, for example—can drastically alter the perceived narrative; politicians, for instance, might tout average tax cuts that overwhelmingly favor the wealthy, obscuring the reality for the median family. The narrative tension builds as the authors dissect how indirect measurements, like radar guns or whale population estimates, rely on models that introduce potential inaccuracies, stressing the importance of calibration and acknowledging the inherent guesswork involved. The chapter illuminates the concept of 'distilling numbers,' showing how choices in representation—percentages of starting versus final volume, annual versus total loss—can create vastly different impressions, like the whisky industry's 'angel's share' evaporation being framed as either a minor or major loss. The authors caution against 'pernicious percentages,' where figures are presented without meaningful comparison, such as a '99.9% caffeine-free' label on cocoa that's as informative as labeling strong coffee the same way, or Breitbart's misleading statistic about DREAM Act recipients. Goodhart's Law enters the stage, cautioning that when a measure becomes a target, it ceases to be a good measure, illustrated by the Hanoi rat bounty program and colleges gaming ranking metrics. The authors then introduce 'mathiness'—formulas that look mathematical but lack logical coherence, exemplified by the VMMC Quality Equation and the Trust Equation, which promise precision but fail to deliver, often lacking dimensional consistency. The chapter concludes with a warning about 'zombie statistics'—outdated or fabricated numbers that persist through repetition, like the myth that 50% of scientific articles are never read, emphasizing that without source and context, statistics are nearly worthless. Like shadows playing on a wall, numbers can distort reality if we forget to question their origins and interpretations.

In this illuminating chapter of 'Calling Bullshit,' Carl T. Bergstrom and Jevin D. West turn our attention to selection bias, a pervasive yet often invisible force distorting our understanding of data. Like skiers at Solitude convinced it's the best mountain because, well, they're *at* Solitude, we often fail to recognize that where we look profoundly shapes what we see. The authors caution against extrapolating findings from one group to another without considering whether the sample truly represents the broader population. They expose how selection bias arises when the sampled individuals systematically differ from the eligible population, leading to skewed impressions, such as the auto insurance paradox where every company claims substantial savings for switchers because only those poised to save actually switch. Bergstrom and West then reveal how this bias hides within seemingly innocuous statistics, like university class sizes, where the 'average' class size touted by administrators clashes with the 'experienced' class size felt by students, because larger classes disproportionately impact more students. This mathematical sleight of hand extends to the friendship paradox, where most people's friends have more friends than they do, not due to personal unpopularity, but because social butterflies inflate the average. Like traffic during rush hour where one is more likely to find themselves in the slower lane, the presence of the observer becomes intrinsically linked to the observed variable, skewing perception. The authors introduce Berksons paradox, illustrating how selection for certain qualities, like niceness and attractiveness in dating, can create negative correlations where none exist in the general population. Finally, Bergstrom and West dissect data censoring, exemplified by the misleading graph of musicians' death rates across genres, where the premature deaths in newer genres like rap skew the data because the genre is simply not old enough to account for a full lifespan. To disarm selection bias, the authors advocate for randomization in studies, citing the example of employer wellness programs, where randomized controlled trials reveal the ineffectiveness of programs that observational studies had previously lauded. Only by understanding and actively mitigating selection bias can we hope to navigate the data-driven world with a clearer, more skeptical eye, ensuring that our conclusions reflect reality, not just the echoes of our own skewed perspectives.

In this enlightening chapter, Carl T. Bergstrom and Jevin D. West turn their skeptical eye toward the world of data visualization, a realm where clarity should reign, but often confusion and deception lurk. They begin with a striking example: a graph about Florida's Stand Your Ground law, its inverted axis transforming a rise in gun murders into an apparent decline, a potent reminder that even unintentional design flaws can mislead. The authors then trace the history of data visualization, from William Playfair's pioneering charts to the complex interactive graphics of today's New York Times, noting a critical gap: our education hasn't kept pace with this visual deluge, leaving many ill-equipped to interpret these powerful tools. Bergstrom and West introduce the concept of 'ducks'—graphs where aesthetics overwhelm data, like USA Today's whimsical but ultimately confusing charts, arguing that such cuteness undermines understanding. They warn against 'glass slippers,' visualizations that force data into ill-fitting forms, such as endless, nonsensical periodic tables of everything imaginable, losing logical coherence in the process. The chapter then sharpens its focus, revealing how axes can be manipulated to distort the story, like the truncated bar chart that exaggerates differences in trust levels or the dual-axis graph that falsely links vaccines to autism. It is a world where choices, even subtle ones, shape perception. The authors present the 'principle of proportional ink,' a guiding star for ethical visualization: the area of a shaded region should directly reflect the value it represents, violated by 3D bar charts, donut charts, and other visual gimmicks that prioritize form over function. Bergstrom and West urge us to be vigilant, to question the stories that data visualizations tell, and to recognize that behind every graph lies a designer with the power to shape our understanding, for better or for worse. They remind us that data visualization, when done right, illuminates; when done wrong, it obscures, manipulates, and ultimately, calls bullshit.

In this exploration of big data, Carl T. Bergstrom and Jevin D. West cast a skeptical eye on the promises of artificial intelligence, starting with Frank Rosenblatt's early, grandiose predictions about perceptrons, which echo even today. The authors note how little has changed in the AI hype cycle, despite advances being driven primarily by increased data and processing power rather than algorithmic breakthroughs. They reveal the core principle of machine learning: feeding computers labeled training data to generate new programs, emphasizing that flawed data inevitably leads to flawed outcomes—garbage in, garbage out. Bergstrom and West caution against focusing solely on algorithms, urging critical examination of the data itself, and they illustrate this with the U.S. Postal Service's effective handwriting recognition system, which thrives on high-quality labeled data, contrasting it with the complexities of identifying fake news. The narrative tension rises as the authors dissect a study claiming AI could detect sexual orientation from facial features, revealing how easily interpretations can be skewed and extraordinary claims can lack solid evidence. They expose the myth that machines are free of human biases, explaining that algorithms often perpetuate biases present in their training data, which leads to unfair outcomes in areas like criminal sentencing and lending. The authors then discuss algorithmic accountability and transparency, noting the difficulty of eradicating biases and the challenge of understanding how complex algorithms make decisions; as an example, they describe how Google Flu Trends failed due to overfitting and spurious correlations. Ultimately, Bergstrom and West advocate for a balanced perspective: acknowledging the value of data while remaining vigilant against hype and uncritical acceptance, reminding us that big data is not inherently better, just bigger, and it certainly doesn't speak for itself.

In this exploration of scientific integrity, Carl T. Bergstrom and Jevin D. West reveal that science, humanity's greatest invention, is also a human endeavor, prone to both accidental and deliberate missteps; it’s a system jury-rigged atop the evolved psychology of humans, driven by curiosity, status-seeking, and the pursuit of recognition. The authors underscore that scientists, while driven by a quest for truth, are also motivated by career advancement and peer recognition, a tension that can lead to the priority rule, where being first overshadows being accurate. Scientific papers, the currency of this world, are assessed through peer review, a process intended to uphold standards but not immune to human fallibility. The announcement of cold fusion by Fleischmann and Pons serves as a stark reminder that even revolutionary claims must withstand the scrutiny of replication, a cornerstone of scientific self-correction. Bergstrom and West then dissect the concept of p-values, illuminating how easily they can be misinterpreted, leading to the prosecutor's fallacy, where the probability of a match given innocence is confused with the probability of innocence given a match. Like a defense attorney skillfully revealing the flaws in a seemingly airtight case, they caution against mistaking statistical significance for definitive proof. The replication crisis, a storm gathering on the horizon of scientific research, highlights the alarming rate at which published studies cannot be reproduced, not always due to fraud, but often due to subtler issues like p-hacking and publication bias. The authors illustrate how researchers, under pressure to publish, may inadvertently manipulate data to achieve statistical significance, a practice that undermines the integrity of the scientific process. To further clarify, they introduce the base rate fallacy, emphasizing that the prevalence of a condition must be considered when interpreting test results, lest false positives lead to misguided conclusions. Ioannidis's provocative assertion that most published research findings are false is examined, revealing the potential for publication bias to skew the scientific literature, creating a mirage of certainty where doubt should prevail. As science reporting amplifies these biases, sensationalized headlines often overshadow the preliminary nature of findings, misleading the public and eroding trust. Bergstrom and West argue that science should not be viewed as a collection of definitive facts, but rather as a series of arguments, each contributing to an ongoing dialogue. Finally, they expose the market for bullshit science, where predatory journals exploit the pressure to publish, polluting the literature with unreliable articles and undermining the credibility of scientific research. Yet, despite these challenges, the authors conclude on a note of cautious optimism, highlighting the cumulative nature of science and its inherent drive toward self-correction, a testament to its enduring power to illuminate the world around us.

In a world awash with misinformation, Carl T. Bergstrom and Jevin D. West cast themselves as guides, arming us with skepticism in their chapter, "Spotting Bullshit." They open with a stark example: a fabricated image of a football player desecrating the American flag, a digital phantom born from cultural division. This sets the stage for their core argument: spotting falsehoods requires cultivating specific habits of mind. Like a driver scanning for danger, one must learn to question information at its source, mirroring a journalist’s interrogation: Who is telling me this? How do they know it? What are they trying to sell me? The authors illustrate this with Colleen McCann, a crystal healer, and her claims about crystals absorbing information, urging us to scrutinize not just the message, but the motives behind it; everyone, they remind us, is selling something, be it a product, an idea, or a perspective. Bergstrom and West then caution against unfair comparisons, spotlighting media’s penchant for sensationalism, for instance, the claim that airport security trays are germier than toilet seats—a misleading comparison focusing only on respiratory viruses. The authors suggest a deeper look; ranked lists, they argue, are only meaningful when comparing truly comparable entities, and they dissect the pitfalls of "most dangerous cities" lists, revealing how arbitrary city boundaries skew crime statistics. A key insight emerges: question the metrics. Shifting gears, the instructors introduce the "too good or too bad to be true" heuristic, recounting the dubious claim of a 40% drop in international student applications due to Trump's policies. The authors urge us to dig to the source, revealing how a tweet distorted the original survey's findings, reminding us that in the age of social media, extreme claims often mask deeper inaccuracies. They advocate for thinking in orders of magnitude, offering the Fermi estimation technique—a mental tool to quickly assess numerical claims. They dismantle the claim of nine billion tons of plastic entering the ocean yearly, a figure dwarfing the entire human population, revealing it as a thousandfold exaggeration. The authors then caution against confirmation bias, the tendency to embrace information aligning with pre-existing beliefs. They dissect a viral tweet about gender bias in recommendation letters, revealing that the data actually contradicted the initial, emotionally resonant interpretation. Finally, they urge us to consider multiple hypotheses, illustrating this with the example of Disney's stock drop following the Roseanne Barr controversy, highlighting that correlation does not equal causation. Bergstrom and West conclude with practical steps for spotting online misinformation, urging us to corroborate claims, trace information to its origin, and be wary of deepfakes. They paint a vivid picture: the internet as an information superhighway, urging us to stop littering it with unverified claims, and reminding us to think more and share less, cultivating a cleaner, more discerning information environment.

In this culminating chapter, Carl T. Bergstrom and Jevin D. West shift from spotting bullshit to actively calling it out, framing it not just as an intellectual exercise, but as a performative utterance—a powerful act demanding prudence. They dissect the anatomy of "calling bullshit," distinguishing it from mere skepticism, emphasizing its public nature and its potential to protect communities. The authors caution against careless accusations, highlighting the importance of accuracy and respect. Like a seasoned debate coach, they introduce key strategies for refutation, starting with *reductio ad absurdum*, turning flawed arguments into laughable extremes, as seen in the critique of overly simplistic models predicting women would outrun men in the Olympics. They then advocate for the power of counterexamples, a single tree standing defiant against sweeping claims about immune systems, illustrating how one well-chosen instance can dismantle an entire theory. Bergstrom and West champion the use of analogies to reframe arguments, like comparing Seattle traffic improvements to baseball team investments, urging us to trust our critical thinking. The authors emphasize redrawing misleading figures, revealing hidden truths in data, and deploying null models, creating simplified systems to expose flawed assumptions. The chapter gently guides us through the psychology of debunking, advising us to decouple identity from the issue at hand, simplify the narrative, and find common ground. Navigating the minefield of misinformation requires ethical precision: be correct, be charitable, admit fault, be clear, and crucially, be pertinent, distinguishing the insightful "bullshit caller" from the annoying "well-actually guy." Ultimately, Bergstrom and West present calling bullshit not as a mere intellectual game, but as a moral imperative—a necessary act to safeguard truth, science, and democracy in a world drowning in deception, urging us to start with self-reflection, recognizing that the most potent source of bullshit we must confront resides within ourselves.

“Calling Bullshit” equips us to navigate a world saturated with misinformation. Beyond identifying falsehoods, it fosters critical thinking about data, visualizations, and even scientific claims. The book reveals how easily numbers can mislead, biases skew results, and algorithms amplify existing prejudices. It underscores that skepticism isn't cynicism, but a vital tool for informed decision-making. The emotional lesson lies in recognizing our own susceptibility to manipulation and the responsibility we bear to challenge misleading narratives. Practical wisdom emphasizes questioning sources, scrutinizing data presentations, and understanding the limitations of AI. Ultimately, the book empowers us to become discerning consumers of information, promoting a more truthful and transparent world.