第 1 章
The Trust Paradox: Why We Need It, Why It Fails, and How to Make It Work
Imagine you're at a crossroads, facing a decision that could change your life. The job offer that requires relocating, the investment opportunity promising remarkable returns, or the romantic partner proposing a deeper commitment. Each scenario hinges on one fundamental question: Can I trust? This question occupies our minds to a degree few other concerns can, because trust isn't just about facts-it's about predicting behavior based on competing interests.
David DeSteno's "The Truth About Trust" has garnered attention from business leaders like Richard Branson and psychologists worldwide for its revolutionary approach to understanding trust. Unlike traditional self-help books that offer simplistic "trust your gut" advice, DeSteno's work draws from cutting-edge research across disciplines to reveal trust's complex mechanisms. The book has been featured in The New York Times' science bestseller list and has influenced corporate training programs at companies like Google and Microsoft, where understanding trust dynamics directly impacts team performance and innovation.
第 2 章
Trust: The Essential Gamble
At its core, trust exists because we're vulnerable. Our needs and desires aren't entirely under our control-they depend on others' cooperation. Whether it's business partners potentially embezzling funds, spouses possibly having affairs, or confidants potentially betraying secrets, our well-being hinges on others who have competing needs of their own.
Trust becomes necessary precisely where our needs diverge from others'. If everyone's goals aligned perfectly, trust would be unnecessary, but such alignment rarely occurs naturally. This makes trust fundamentally a bet containing risk. Though humans are generally risk-averse-with losses hurting more than equivalent gains feel pleasurable-we trust because we must.
The potential benefits from cooperation substantially outweigh potential losses on average. Human society's complexity, technological advancement, and economic resources all depend on trust. From NASA mission control teams to everyday banking, education, and household management, more can be achieved through cooperation than isolation.
Not all trust is well-placed, as the 2008 financial crisis demonstrated. Yet statistically, trusting pays greater dividends in the long run, creating a dynamic tension that shapes how our minds solve the trust equation throughout our lives.
The alternative to trust-transparency through verification-is often impractical due to its costs in time and energy, and the reality that exchanges often involve time lags between giving and receiving. Without delayed reciprocity, cooperation would be severely limited. This commitment problem represents one of human life's central dilemmas; without trust and honored commitments, society would collapse.
Computer scientist Robert Axelrod conducted tournaments where different strategies competed in iterative prisoner's dilemmas. The superior performers shared two properties: an initial willingness to be trustworthy (never defecting first) and being provokable (responding to untrustworthiness in kind). The simple tit-for-tat (TFT) strategy emerged victorious-start fair and then copy your partner's previous action.
However, human interaction differs from computer simulations because we're imperfect and sometimes break trust unintentionally. These "noisy" interactions can doom TFT when used against itself, as unintended slights spiral into mutual defection. This led to the discovery that generous tit-for-tat (GTFT)-which occasionally forgives defection-outperformed regular TFT by overcoming noise. The key insight: trust exists in an ever-changing equilibrium between selfishness and cooperation.
第 3 章
The Myths That Mislead Us
Despite decades of research on trust, our understanding requires revision due to four fundamental errors in how we conceptualize it. First, we mistakenly assume reputation represents stable character traits-that honest people are always honest and cheaters always cheat. Scientific data contradicts this view, showing human morality is remarkably variable.
In experiments using the Give Some Game, participants received tokens worth $1 to themselves but $2 to their partner. When researchers induced feelings of gratitude in participants, they became significantly more trustworthy and cooperative, offering more tokens to partners-regardless of whether they knew their partner or not.
Similarly, social stress increased trustworthy behavior by approximately 50 percent in other studies. Even something as simple as wearing counterfeit designer sunglasses versus authentic ones dramatically affected honesty-71 percent of knockoff-wearers cheated on a math test versus only 30 percent of those wearing authentic brands.
These experiments confirm that trustworthiness can't be reliably predicted from past actions. The question shouldn't be "Is he trustworthy?" but rather "Is he trustworthy right now?"
Second, most people view trust as a battle between selfish and selfless motives-a devil and angel on opposing shoulders. But this view misses a crucial evolutionary insight: humans prosper through cooperation. We engage in "selfless" acts because they ensure we'll be beneficiaries of others' resources when needed.
The real competition isn't between selfishness and selflessness but between short-term and long-term gains-more like Aesop's ant and grasshopper than an angel and devil. Untrustworthy behavior can yield immediate rewards, but jeopardizes long-term outcomes if discovered. Our minds continuously balance these competing mechanisms, with momentary states tipping the scales.
Third, traditional trust models often end with the words "on average"-cold comfort when you're about to entrust someone with your life savings. These models offer two solutions: either use past interactions to calculate trustworthiness, or simply trust strangers unless you're certain you'll never interact again. But I've shown that past actions and reputation aren't reliable predictors of future trustworthiness.
Finally, trust isn't just about integrity-competence matters just as much. As Adam Russell from IARPA cleverly illustrated: you might trust someone's honesty and intentions, but would you let them perform brain surgery without medical training?
Most trust research focuses primarily on integrity, overlooking competence. This happens partly because experimental trust games like the Prisoner's Dilemma make competence irrelevant-anyone can give or take money. Yet our minds rapidly assess both qualities.
第 4 章
The Biology of Trust
On a summer evening in Telluride, at a conference on compassion, I found myself surrounded by scientists studying virtuous behavior who were unusually open and trusting despite barely knowing each other. My dinner companions, Stephen Porges and Sue Carter-experts on the biology of social behavior-reinforced two central points about trust: it's driven as much by physiological states as by rational calculation, and humans are built for both trust and betrayal depending on what serves our interests.
Our feelings-stomach pits, racing hearts, calming touches-all influence trust decisions, often occurring automatically rather than following conscious analysis. We frequently feel first and decide whether to trust afterward, guided by ancient biological systems that predate our capacity for rational thought.
Understanding our biological trust mechanisms requires an evolutionary perspective. Our nervous system evolved through three stages: the ancient unmyelinated vagus nerve (causing freezing/immobilization when threatened), the sympathetic-adrenal system (fight-or-flight response), and the mammalian myelinated vagus nerve. This newest system acts as a "cardiac brake," calming the body and reducing stress hormones while connecting to facial muscles and vocal apparatus-creating a physiological state conducive to social engagement and trust.
When we feel safe, increased vagal tone allows us to connect socially, accurately read others' emotions, and act compassionately. Our bodies start with the highest system (social engagement) and move backward if trust isn't warranted. However, balance is crucial-ultra-high vagal tone can make someone overly trusting and unreliable, promising more than they can deliver. Our minds constantly perform "neuroception," scanning for safety or threat and adjusting our physiology accordingly, which influences our trustworthiness without conscious control.
The ability to detect cheating is fundamental to trust across species. Humans reject unfair offers even at personal cost, as seen in the ultimatum game where people refuse splits worse than 80/20. This aversion to inequity isn't uniquely human-Sarah Brosnan's research shows many primates share this trait. Capuchin monkeys and chimpanzees recognize unfair treatment and react with anger when seeing peers receive better rewards for equal work.
Oxytocin's role in trust is more nuanced than initially believed. While it increases trust in economic games and maintains trust even after betrayal, research reveals its dual nature. It doesn't simply make people trust anyone-it specifically strengthens bonds with ingroup members while potentially fueling distrust toward outsiders. Rather than being a "moral molecule," oxytocin functions as a team-building chemical that optimizes social bonds along lines of affiliation.
Humans are built both to trust and distrust, to be trustworthy and to betray. This apparent contradiction makes sense when we understand that evolution optimizes for success, not virtue. Our physiological systems are designed to build bonds when appropriate and be selfish when necessary-to trust when repayment is likely and exploit when we might be cheated.
第 5 章
Trust and Learning: The Child's Perspective
Learning is the essence of human existence, with our minds constantly scanning and analyzing information. This thirst for knowledge is particularly intense during childhood, when developing brains must rapidly acquire survival information. Children's relentless questioning, while sometimes exhausting for adults, reflects this biological imperative to learn.
For children, trust is fundamentally linked to learning because they have limited options for acquiring knowledge. Unlike adults who can research information or conduct experiments, young children must often rely on others' testimony. This makes evaluating trustworthiness-particularly competence-crucial for their development.
Children are remarkably selective about whom they trust for information. Harris's research shows that even with unfamiliar adults, children direct questions three times more frequently to individuals who previously named familiar objects correctly versus those who made errors. This preference persists over time-when brought back a week later, children continued to prefer learning from the previously competent individual.
Perhaps most profound is the discovery that children actually remember information better when it comes from trustworthy sources versus untrustworthy ones-the same information is better retained when delivered by someone they trust. This has enormous implications for education and exam performance.
When competence conflicts with familiarity, older children prioritize expertise. While three-year-olds remain loyal to familiar teachers regardless of competence, four- and five-year-olds will abandon their usual teacher in favor of a competent stranger when learning new information.
Beyond examining how trust impacts learning, we must understand how children learn to trust in the first place. Contrary to outdated parenting philosophies suggesting babies are manipulative, research shows infants enter the world with innate prosocial tendencies.
Harvard psychologist Felix Warneken demonstrated this by observing eighteen-month-olds' responses to adults in need of help. The toddlers readily assisted adults who appeared distressed-retrieving dropped pens or showing adults how to access items through hidden compartments. Before most can string two words together, the instinct to cooperate is already active.
However, children possess dual natures regarding trustworthiness. By age three, they struggle with fairness when self-interest is involved. In experiments where children worked with puppets on coin-retrieval tasks, three-year-olds would adjust rewards based on performance, but never gave puppets more stickers than themselves. While they recognized the principle that more work deserves more reward, they maintained a self-serving bias.
This tension between fairness and self-interest evolves with age. By eight, children show strong aversion to inequity in public settings, rejecting unfair distributions that favor themselves. Yet five- and six-year-olds who share equally when resources are public will hoard prizes when they believe others won't know.
These findings reveal that trustworthiness emerges naturally in children-they don't need to be taught to act fairly, though moral education certainly helps shape their conscious minds. Children arrive with a fledgling morality that, like adults', is more complex than typically recognized.
第 6 章
Trust in Romantic Relationships
Romantic relationships create unique vulnerabilities that make betrayal particularly devastating. While all adult relationships involve some joint dependence, romantic bonds create a universal vulnerability that's difficult to escape. The burning desire to connect, share, and depend on a partner makes "moving on" after relationship failure extraordinarily difficult. Few experiences match the emotional turmoil of betrayal by a trusted spouse or significant other.
This emotional intensity stems from the magnitude of costs and benefits attached to romantic relationships. Evolutionarily speaking, the trustworthiness of a co-parent directly impacts reproductive success-disloyal spouses historically meant lower odds of passing on genes. But benefits extend beyond reproduction: people in satisfying long-term relationships enjoy better physical and psychological health across all metrics. A good relationship provides a constant cooperator who buffers against life's challenges-stress, finances, and social dilemmas.
At heart, romantic relationships operate on give and take, with the most satisfying ones maintaining roughly equivalent costs and benefits. The specific distribution doesn't matter-one partner might provide more income while the other handles family needs, or one offers emotional support while the other manages finances. What's crucial is that each provides benefits the other values, and the subjective value feels equal.
This balancing act between self-protection and gaining benefits from intimate bonds becomes unmanageable if we try tracking every exchange. The human mind simply can't maintain accurate tallies of all promises and cooperative acts. Trust functions as a cognitive shortcut, removing the need for detailed tracking. It's a bet that our partner will, on average, hold up their end of the relationship.
When trust emerges, relationships transition from "exchange-based" to "communal" styles. People stop calculating "if I do this, she'll do that" and simply uphold responsibilities while assuming partners will do the same. This frees mental capacity previously used for monitoring exchanges.
Trust does more than reduce mental fatigue-it actively strengthens relationships. Jeffry Simpson's research using "strain-test paradigms" demonstrates this by having couples discuss goals requiring major sacrifices from their partners.
His findings reveal that trust functions as a relationship buffer that smooths rough spots. Individuals with higher trust levels show greater accommodation and collaboration during difficult discussions. At each decision point where conflict might escalate, trust tips the scale toward compromise by making us consider the long-term relationship value over short-term victories.
Trust also positively biases our interpretations of a partner's actions. Those with higher trust levels regularly overestimate the accommodations they receive-seeing their partner's concessions as noble sacrifices. This creates a self-reinforcing cycle: more trust leads to seeing greater sacrifice, which in turn builds even more trust.
When that nagging feeling arises that something's amiss in your relationship, you face a choice between trusting your head or your heart. This decision isn't simply about choosing intuition or reason-both systems work together to determine your response.
Intuitive trust operates outside conscious awareness through automatic calculations, while reasoned trust relies on deliberate analysis of a partner's behavior. How these systems interact becomes crucial when trust is challenged by ambiguous situations.
Jealousy manifests as a complex constellation of fear, anger, and sadness-each reflecting different aspects of broken trust. Fear stems from anxieties about the social and economic costs of losing a partner. Anger derives from the belief that relationship rules were violated. Sadness comes from realizing someone whose judgment you value sees another as more desirable.
Despite common perception, jealousy isn't fundamentally about sex but about trust. It only emerges in triadic situations where a partner might choose someone else over you. This explains why swingers experience little jealousy-they've established that recreational sex doesn't breach trust in their relationships.
第 7 章
Power, Status, and Trust
The relationship between social class and trust reveals surprising patterns contrary to common expectations. While one might assume those with fewer resources would prioritize self-interest for survival, the opposite proves true. Trust functions not as a luxury but as a survival tool for those who must depend on others.
Frank Capra's "It's a Wonderful Life" crystallizes this dynamic through George Bailey and Henry Potter-the trustworthy common man versus the untrustworthy wealthy elite. Those lower in socioeconomic status need mutual support since complete self-reliance isn't possible, making trust essential. Meanwhile, the wealthy possess resources allowing them to rely on money rather than people, potentially making them less invested in trustworthy behavior.
Even crossing streets involves trust-we must believe approaching drivers will stop for pedestrians. Paul Piff's research at UC Berkeley examined this trust dynamic across social classes with striking results. At a San Francisco intersection, researchers found that while 100% of drivers from the lowest socioeconomic class stopped for pedestrians, nearly 50% of upper-class drivers broke the law by cutting off pedestrians.
This pattern extended beyond traffic behavior. In a negotiation experiment, upper-class participants were significantly more likely to conceal that a job would be eliminated in six months when interviewing candidates seeking long-term employment. Similarly, in a dice-rolling game, wealthier participants more frequently inflated their rolls to receive undeserved money.
The idea that power corrupts isn't new, but what's remarkable is how quickly it happens. Changes in trustworthiness don't require slow enculturation-they occur rapidly when one's perceived need to rely on others shifts. Class and power differences are malleable and situational; what matters is where one stands relative to those nearby.
Paul Piff demonstrated this with a clever experiment involving candy meant for children. Participants who had just compared themselves with others of lower socioeconomic status (feeling relatively powerful) took significantly more candy from a bowl meant for children than those who had compared themselves with higher-status individuals. Crucially, these participants didn't actually differ in true socioeconomic status-just in how powerful they momentarily felt.
The simple presence of money triggers a preference for self-sufficiency over social connection. Kathleen Vohs found that mere reminders of money caused people to physically distance themselves from others-sitting farther apart and preferring to work alone rather than cooperatively on difficult tasks. Money primes essentially turned people into lone wolves who rejected interdependence, preferring isolation and self-reliance over social connection.
Power focuses on short-term gains without fear of repercussions. Powerful individuals can order rather than trust others, but maintaining power through coercion creates risk of eventual retribution. Evolutionarily, this makes sense-if power enables reproductive success and resource acquisition, even temporary advantages benefit gene transmission.
However, there's another leadership path. As Dacher Keltner's research shows, socially engaged individuals with high emotional intelligence often rise to power initially. Those few who resist power's self-indulgent impulses maintain their positions with less stress. These "benevolent leaders" who maintain fairness and trustworthiness may accrue smaller immediate profits but achieve greater long-term stability.
第 8 章
Detecting Trustworthiness
Until now I've established trust's central role in life, implying that detecting trustworthiness would offer powerful advantages. Despite decades of research by academics, businesses, and governments, most deception detection methods remain ineffective-people are only slightly better than chance (54%) at identifying liars.
This seems counterintuitive. If partner trustworthiness is integral to success, why can't the human mind read it? Either trust signals don't exist, leaving humanity flying blind, or they do exist but we're not looking correctly. Given trust's evolutionary importance-offering great benefits while posing significant risks-it seems unlikely this ability would be left to chance.
Unlike most communication designed for quick, clear information exchange, trust signals must be subtle. Cornell economist Robert Frank recognized why: if trustworthiness were easily detectable (imagine a giant "T" on trustworthy foreheads), the trustworthy would be constantly targeted for exploitation.
Even if everyone's trustworthiness were visible, we wouldn't achieve utopia. Initially, trustworthy people would partner together while the untrustworthy would be ostracized. Over time, as most remaining people became trustworthy, attention to these signals would decrease-creating opportunity for cheaters to mimic the signals successfully. This would trigger an evolutionary arms race: cooperators developing secret codes to identify each other while cheaters attempt to crack these codes.
Identifying trust signals is challenging because they must be subtle to avoid exploitation by cheaters. Unlike simple nonverbal cues that supposedly have fixed meanings (smiles mean happiness, furrowed brows mean anger), trust signals require both configural and situational context. Individual gestures are unreliable-a smile can indicate happiness or malice depending on who's smiling, and athletes' victory expressions often look like anger when viewed in isolation.
To tackle the challenge of identifying trustworthiness signals, I assembled an interdisciplinary team starting fresh. David Pizarro (social psychologist), Robert Frank (economist), and I combined expertise in nonverbal behavior, emotional decision-making, evolutionary models, and behavioral economics. We needed precise control over nonverbal cues, which humans can't provide since most nonverbal behavior occurs unconsciously. Our solution: robots.
Our experiment revealed that face-to-face conversations improved trustworthiness prediction accuracy by 37% compared to instant messaging. Using time-synced video recordings, we identified a set of four nonverbal cues that strongly predicted untrustworthiness: crossing arms, leaning away, face touching, and hand touching. The more frequently someone displayed these behaviors, the fewer tokens they shared with partners. Importantly, people used these signals unconsciously-no participants could articulate what influenced their judgments.
To definitively prove our identified cues actually cause perceptions of untrustworthiness, we employed Nexi, a humanoid robot from MIT's Personal Robots Group. Using a "Wizard of Oz" setup, one researcher controlled Nexi's basic movements and speech while another independently triggered either our target untrustworthiness cues or neutral gestures. The results confirmed our hypothesis perfectly: participants who saw Nexi display the four untrustworthiness cues reported less trust in her, expected her to share fewer tokens, and were less willing to share with her themselves.
I've emphasized that trust isn't one-dimensional-it requires both integrity and competence. Unlike signals of fairness which benefit from subtlety, competency signals are meant to be visible. These include expanded posture, head tilted upwards, arms held open or akimbo, and decreased gazing during interactions-all nonverbal expressions of pride and status.
第 9 章
Trust in the Digital Age
The term "cybertrust" often evokes foreboding feelings about internet dangers, but the basic issue of trusting technology has existed for thousands of years. What's new is technology's encroachment into the human domain of sociality, opening vast new ways trust can be manipulated. When we rely on technology, we make ourselves vulnerable to its competence, just as ancient sailors did with astrolabes or modern drivers do with mapping software. Despite occasional failures, we continue trusting technology because benefits outweigh costs and misfires are quickly corrected through competition.
Unlike humans, machines typically don't have desires or intentions-they're made to accomplish tasks. This removes integrity from the trust equation, making technology increasingly seductive. Our minds unconsciously accept markers of scientific sophistication as cues for competence. The "MRI effect" demonstrates how people find claims more persuasive when accompanied by brain images, despite neuroscientists knowing brain imaging is still in its infancy.
Social media, virtual agents, and online games have us using technology to be social like never before. When accepting electronic information, we're vulnerable to the intentions of those who sent it, and increasingly subject to manipulation of social cues. By 2011, over half a billion people spent more than twenty hours weekly interacting as avatars-digital representations "given life" through operator interfaces. Beyond gaming, corporations like IBM expect most employees to attend meetings via avatars by 2015.
Technology can be harnessed to build rather than exploit trust. Facebook's team is exploring how similarity metrics might help bullied users find trustworthy helpers. Since similarity increases compassion and willingness to help, computational systems can analyze vast user data to identify friends most likely to provide support. The key is identifying appropriate candidates who will trigger the mind's similarity sensors while avoiding problematic connections like ex-partners.
Timothy Bickmore's work with virtual nurses demonstrates technology's positive potential. His team created Louise, a virtual nurse designed to help low-health-literacy patients (36% of American adults) understand discharge instructions. Patients reported higher trust with the virtual nurse than human counterparts, spent more time engaging with it, and showed greater intention to follow medical regimens. One patient remarked, "She treated me like a real person!" Another noted the agent's patience compared to rushed doctors. The success came from patients feeling empowered through a relationship with an entity they could trust.
第 10 章
The Challenge of Self-Trust
Trusting yourself presents a unique challenge-you're placing a bet on what your future self will do, whether resisting financial temptation, remaining faithful in relationships, or avoiding unhealthy treats. While we might expect self-trust to be straightforward since we know ourselves better than anyone else, we often misjudge our own trustworthiness. This isn't due to lack of intelligence or introspection, but rather a built-in limitation of the mind. How much we value different rewards depends not just on what they are, but when they occur-creating a disconnect between our intentions and actions.
When asked if they're fair and honest, most people confidently answer yes. However, laboratory experiments consistently reveal a disconnect between this self-perception and actual behavior. The author's research demonstrates how people can be self-serving while simultaneously convincing themselves they're being fair and honest-a fascinating illustration of our limited self-knowledge and our tendency to give in to temptation while maintaining positive self-image.
Our ability to predict our own trustworthiness is hampered by two mental illusions: forward-looking myopia and rearward-looking whitewash. We focus more on the present than the future, making our predictions about resisting temptation wildly inaccurate. Dan Gilbert's research shows we're terrible at forecasting our future feelings-we don't account for contextual factors like stress or hunger that will influence our decisions. Additionally, our minds value immediate rewards disproportionately over future ones.
Kathleen Vohs's research demonstrates that willpower functions as a limited resource-once depleted by resisting one temptation (like dieters avoiding nearby snacks), we're far more likely to give in to the next temptation (consuming triple the ice cream). Similar patterns emerge with thought suppression experiments and impulse shopping, proving that trusting ourselves requires understanding these psychological limitations.
We rationalize our failures through what I call the rearward-looking whitewash. As Ayn Rand noted, rationalization involves making reality fit our emotions rather than perceiving it accurately. Our innate desire to view ourselves as competent and upstanding leads us to explain away our failures. Gilbert's research with Boston commuters demonstrates this perfectly: those who missed trains blamed external factors like broken turnstiles rather than their own tardiness, while those predicting how they'd feel anticipated self-blame.
Our mental time-travel limitations are partly flaw and partly design. The flaw stems from our relatively new evolutionary ability to simulate the future-our enlarged frontal cortices enable this, but like new software, it comes with bugs. We're good at predicting concrete outcomes but terrible at forecasting our emotional responses. The whitewash mechanism, however, serves a purpose-without it, we might never trust ourselves again, destroying motivation toward long-term goals.
The solution lies in bridging the gap between our present and future selves. Modern technology offers tools that allow the present you to communicate with or even control the future you-from diet apps to programs that prevent drunk dialing. While humanity has always used reminders to solve this problem, today's technology makes this communication more seamless, helping us make more informed decisions across time.
第 11 章
Trust as Social Resilience
When disaster strikes, trust becomes the engine of resilience. After Hurricane Katrina devastated Gulfport, Mississippi in 2005, Hancock Bank made an extraordinary decision: they distributed $40 million to anyone who would write their name and information on paper scraps-without verification of account holdings. Bank employees even laundered money retrieved from flooded vaults to make it usable. This trust-based approach worked remarkably well-all but 0.5% was repaid, and the bank's assets grew by 20%.
Similarly, after the 2013 Boston Marathon bombing, residents opened their homes to stranded strangers. These aren't anomalies-research shows neighborhoods with higher trust levels recover faster from disasters. Trust enables social and economic capital to flow when most needed. Yet blind trust isn't optimal-some people will always exploit vulnerability. The key is having a system that wants to trust but knows when to pull back-exactly what evolution has given us.
In a perfect world where everyone could be trusted, cooperation would flourish beyond imagination-though such certainty would eliminate the need for trust itself. Our imperfect reality requires our minds to constantly balance immediate selfish gains against longer-term communal benefits. Evolution didn't design us to be saints but successful adapters, making our trustworthiness responsive to changing conditions.
Trust extends far beyond handshakes and promises about monetary exchanges. First, it's not solely about integrity-competence matters equally. Intent means little without ability, so assessing trustworthiness involves evaluating both expertise and reliability. Second, trust isn't limited to major transactions or potential infidelity; it infiltrates seemingly mundane aspects of daily life. Trust in teachers affects how children learn, trust in partners influences daily well-being, and trust in technology determines information quality and security.
Past behavior poorly predicts future actions-that's why relying on reputation to judge trustworthiness is problematic. Ethics are situationally determined, which explains why "I thought I could trust him" is such a common lament. The best trust assessments focus on inferring a person's motivations in each specific instance rather than relying on long-standing labels.
Trust decisions, more than almost any others, hold the keys to both individual and societal resilience. Traditional moral education emphasizes top-down approaches-telling children to be trustworthy and use willpower to honor commitments. This is valuable but incomplete, as willpower inevitably fails. What's needed is a more effortless, bottom-up source of virtue that doesn't require willpower but flows automatically.
Moral emotions-what Adam Smith called "moral sentiments"-constitute an ancient mechanism guiding virtuous behavior by restraining immediate self-interest. Parents often resist allowing children to experience shame and guilt, yet these states, along with gratitude and humility, are necessary for building emotional intelligence. Learning to manage these feelings enhances understanding of transgressions and modifies future behavior.
To improve individual and collective welfare, we need both principled thinking and emotional mechanisms that increase empathy. We must not only think about trust but feel it. This approach won't create utopia, but will nudge the world in that direction.
