Book Summary: Talking to Strangers by Malcolm Gladwell

This is a summary of Talking to Strangers by Malcolm Gladwell, which argues it is incredibly difficult to understand other people. It is well-written and contains many engaging examples. Unfortunately, the examples do not illustrate Gladwell’s points well and the analysis is not very rigorous.

Buy Talking to Strangers at: Amazon | Kobo (affiliate links)

Key Takeaways from Talking to Strangers

• Most humans have a tendency to “default to truth” and assume that someone is being truthful.
  • Although this tendency will occasionally lead us astray, defaulting to truth will give you the right answer more often than not.
  • Overall it is good that society is trusting, so we should not penalise each other for defaulting to truth.
• People are not transparent, like they are in TV shows. But most people mistakenly assume people are transparent. This can lead to misunderstandings.
  • Facial expressions are not universal. Even within a culture, people tend to overestimate how expressive they are.
  • We see ourselves as complex and nuanced, but jump at the chance to judge others. This is the illusion of asymmetric insight.
  • Some people are mismatched. Some honest people act the way we think liars act, while some liars act the way we think honest people act. Even seasoned law enforcement officers are terrible at telling whether someone is telling the truth when they are mismatched.
• Alcohol makes us myopic (i.e. focus on short-term consequences at expense of long-term ones).
  • When social rules are unclear, alcohol can increase the risk of misunderstandings.
  • Alcohol makes it really hard to tell just by looking at someone whether they’ve blacked out. (“Blacked out” here does not mean they’re unconscious. It just means they won’t remember that moment.)
• We need to accept that there are limits to our ability to make sense of strangers.
  • When people are subjected to extreme stress (e.g. in an interrogation or torture), they are less able to recall information accurately.
• People behave differently in different contexts. Some actions are coupled to particular place or context.
  • Suicide seems to be coupled to particular suicide methods. We could therefore reduce suicides by reducing the number of suicide methods unavailable.
  • Crime seems to be extremely concentrated in certain street segments (“hotspots”) in a city. We could reduce crime by aggressively policing in those hotspots.
  • However, aggressive policing outside of a crime hotspot is too great an intrusion on people’s liberty. It will erode trust between the police and the community.

Detailed Summary of Talking to Strangers

This summary departs from the order of Talking to Strangers. First, I set out the 3 main topics Gladwell discusses in the book: default to truth, transparency, and coupling. Then I summarise the more detailed examples in the book and explain what points Gladwell was trying to illustrate with them.

1. Default to Truth

Psychologist Tim Levine conducted studies where a person encourages a student to cheat. They then ask the students, on camera, whether they cheated. Subjects then have to guess based on the video interviews whether the students actually cheated.

It seems we’re much better at detecting when someone is telling the truth than detecting when they’re lying. For example, if someone is telling the truth, we’ll usually correctly assume that they’re being truthful. But if someone is telling a lie, we’re worse than chance at detecting it.

People tend to default to truth

We tend to assume people are being truthful. We only move away from this assumption if there’s a “trigger”. But slight doubt will not be enough of a trigger, as you can often explain away those doubts. We’ll only stop defaulting to truth if the evidence against truth is “definitive”. (This is consistent with Annie Duke’s claim that we default to believing things we read or hear, based on work done by a different psychologist.)

Gladwell refers to the Milgram experiment as an example of default to truth. The guy supposedly getting the electric shocks was, by Milgram’s own description, a terrible actor. And his “cries of pain” came from the loudspeaker in the corner, rather than from the door to the room where he was. But over 50% of the subjects fully believed they were shocking the learner. A further 30% had doubts but still thought that the learner was probably getting shocked, or were just unsure.

But default to truth is generally a good thing

Levine reminds us that in real life, most people do tell the truth. So it doesn’t usually matter if we default to truth. It’ll usually lead us in the right direction. [I agree. I think the default to truth is a sensible default. We don’t want to live in a society where everyone assumes a lie based on nothing more than a suspicion.]

We need some people in our society who don’t default to truth – they perform a valuable function. But it would be disastrous if nobody defaulted to truth. Levine points out that defaulting to truth is efficient. We get deceived once in a while but that’s just the cost of doing business. If we didn’t default to truth and scrutinised everything, we would just spend all our time being suspicious of others.

2. Transparency

Transparency is the idea that people’s behaviour and demeanour (the way they represent themselves on the outside) tells us accurately how they feel on the inside.

Jennifer Fugate is an expert in the Facial Action Coding System (FACS). FACS assigns every muscle movement in the face a number, called an “action unit”. You can then objectively score a person’s facial expression according to that system. For example, a Duchenne smile (a smile involving the eyes) is AU 12 + AU 6. The FACS manual is over 500 pages long.

Gladwell asked Fugate to score some facial expressions in scenes from the TV show, FRIENDS. Fugate found that the actors in FRIENDS show every emotion clearly on their face. You can even watch the scene with the sound turned off and understand how the characters are feeling. The characters in FRIENDS are transparent. But real life is not like FRIENDS.

Facial expressions are not universal

Sergio Jarillo and Carlos Crivelli did some studies on some isolated islands called the Trobriands near Papua New Guinea. He showed Trobrianders pictures of people with different facial expressions (e.g. smiling, scowling, pouting) and asked them to guess what the matching emotion was (e.g. happy, angry, sad). Kids at a Spanish primary school did very well on this test. Trobrianders struggled. For example, 100% of the kids correctly identified a smiling face as being happy, but only 58% of Trobrianders did. The Trobrianders found anger particularly difficult – 20% thought the angry face was a happy face.

Jarillo and Criveli did the same study in Mozambique and got similar results. Another group, led by Maria Gendron, did similar experiments in Namibia and found similar results. This suggests that facial expressions are not universal.

Even within a culture, people aren’t as expressive as they think

Achim Schutzwohl and Rainer Reisenzein conducted an experiment where they surprised subjects with an unexpected scenario. They had to rate their surprise on a scale from 1 to 10. The average was 8.14.

When asked, most subjects expected that their faces reflected their level of surprise. But they didn’t. The researchers coded people’s expressions using the FACS system and found only 5% of cases showed wide eyes, raised eyebrows and dropped jaws. Schutzwohl noted that the participants grossly overestimated their surprise expressivity.

Gladwell says that this is part of what makes it hard for us to understand strangers. We have to substitute an idea, a stereotype, for direct experience – and often that stereotype is wrong.

The illusion of asymmetric insight

Emily Pronin conducted an experiment where she gave people some letters in a word. The subjects then had to fill in the blanks. For example, if you got B _ _ K, you could write BOOK or BEAK or BANK, etc. The researchers then asked subjects what they thought their choices said about them. All subjects said nothing – their word completions were just random and didn’t reveal anything about their personality.

The researchers then gave subjects a list of words that other people had filled in. Subjects then thought the stranger’s choices revealed a lot about their personality.

Pronin suggests that we should be more patient with others and slower to judge them. Gladwell says that while we see ourselves as complex and nuanced, we jump at the chance to judge others (this links to how we perceive in-group vs out-group members, where we are our in-group and strangers are our out-group).

Some people are mismatched

We tend to judge people’s honesty based on their demeanour. But some people are mismatched. Some honest people act the way we think liars act – nervous, shifty, won’t meet your gaze. While some liars act very believable – confident, friendly, look you straight in the eye.

When Levine showed his videotapes to experienced law enforcement agents, they correctly guessed 100% of the “matched” videos. But on the mismatched videos, they only got 20%.

Gladwell argues that Madoff and Hitler were both mismatched. They were liars that acted trustworthy. Amanda Knox was the opposite. She was an innocent person who acted guilty.

3. Displacement vs Coupling

Gladwell talks about displacement and coupling, mostly in the context of suicide.

• Displacement is the idea that, if one course of action is unavailable, people will just find another course to achieve the same ends. Most people believe in displacement for suicide. They think that if a person is willing to do something as serious as suicide, they’ll be determined to find a way to do it.
• Coupling is the idea that behaviours are linked to very specific circumstances and conditions. If suicide is coupled, we shouldn’t assume that suicidal people will kill themselves even if we remove their choice of suicide method. They will only do it at a particular moment of extreme vulnerability in combination with a particular, readily available, suicide method.

Town gas and suicide rates

Sylvia Plath and Anne Sexton were both poets who committed suicide. The two were friends and both were fascinated by the idea of suicide. Plath committed suicide using an oven with town gas (town gas contains carbon monoxide). Sexton poisoned herself with carbon monoxide in her car.

Shortly after Plath’s death, the UK started to replace town gas with natural gas. Gladwell refers to a line chart showing suicide rates in England and Wales and the US from 1900 to 1990. He says that if there is displacement, then suicide rates should remain quite steady over time, fluctuating with major social events (e.g. war, economic distress). On the other hand, if suicide is coupled, the chart should look random because suicide rates should vary with the availability of particular suicide methods.

Gladwell claims that his chart looks random, like a “roller coaster”. He says suicide rates “plung[ed] down” when town gas was phased out from the late 1960s. [As I note in my Criticisms of Talking to Strangers, there are some problems with this claim.]

Golden Gate Bridge and suicide

Richard Seiden followed up on 515 people who tried to jump from the Golden Gate Bridge but had been unexpectedly restrained. Only 25 of those 515 persisted in killing themselves some other way.

[I looked up this study. The figure refers to the number of people who committed suicide in the 26 years (on average) after their attempt. But it’s likely that more than 25 of those people subsequently attempted suicide.

Also, Seiden found that 31 of that group had died of “natural” causes and another 7 of “accidental” causes during that study period. Seiden points out that some of those natural deaths were indicative of self-destructive tendencies (e.g. alcohol abuse). He also cautions that the distinction between accident and suicide often seemed arbitrary, as alcohol-poisoning and one-car accidents were categorised as “accidental”.]

Crime hotspots

David Weisburd and Larry Sherman, two criminologists, looked at crime in multiple areas. They found that, wherever they went, crime was extremely concentrated in just one or two streets (or even street segments). Even in the “bad” neighbourhood of town, most streets had no crime. In Minneapolis, 3.3% of the street segments accounted for 50% of police calls. Researchers have found similar results in Boston, New York, Seattle, even Tel Aviv.

Weisburd conducted another experiment where he assigned an extra 10 police officers to police a prostitution hot spot. The amount of prostitution in the “hotspot” area fell by two-thirds. Weisburd also found that prostitutes didn’t just move one or two streets over. Instead, most preferred to leave prostitution entirely rather than change their behaviour. They talked about how hard it would be to move, the need to re-establish themselves. One concern was the people they don’t know if they moved to a new area. They were concerned some people would call the police.

Gladwell says that, like suicide, crime is tied to very specific places and contexts.

Kansas City experiments to reduce crime

The Kansas City Police tried a variety of experiments deploying police in different ways to see what was effective. One idea was to take advantage of the broad powers available to police to stop a vehicle. Once a vehicle is stopped, an officer can search the car if they reasonably believe that the motorist might be armed or dangerous. In the experiment:

• two squad cars were assigned to the highest crime area in Kansas City.
• officers were told to watch out for suspicious drivers, use whatever pretext they could to pull them over, and search their car where possible.
• the experiment found that deploying just two squad cars that way, they were able to cut gun crimes in that high-crime district in half.
• in just 200 days, the four officers assigned to that high-crime district did more “policing” than most officers back then did in their entire career.
  Police departments across the country took notice of these stunning results. They tried to copy the Kansas City model but made a crucial mistake. They told officers to pull cars over using whatever pretext and to search them if possible. But they did not concentrate that strategy in high-crime hotspots.

Gladwell says that the Kansas City had been a coupling experiment, but people just found it hard to believe that crime doesn’t just move. [I find it hard to believe too, but I think it depends on the type of crime and on the time period. I can easily accept that in the short term, there is no displacement since it takes time for people to move. But what about the longer term? That said, I agree it makes sense to concentrate policing in high-crime areas. If there is displacement but it’s longer term, you could just deploy police to the “new” crime hotspot after it’s moved.]

Cuban spies in the US

Florentino Aspillaga was a Cuban spy who defected to the US during the Cold War. He knew a lot of sensitive information about Cuba and the Soviet Union. He told the US that dozens of their secret agents in Cuba were actually double agents working for Cuba.

When Fidel Castro learned that Aspillaga had told the US about all the spies, he rounded them up and paraded them across Cuba. He even released a long documentary, La Guerra de la CIA contra Cuba – The CIA’s War Against Cuba. The documentary consisted of footage of everything CIA agents had done in Cuba in the past 10 years, taken by Cuban spies

Gladwell uses this example of how easily “strangers” can mislead us. He questions why we aren’t able to tell when a stranger is lying to our face.

Neville Chamberlain and Hitler

In 1938, Hitler had been making statements about wanting to invade the “Sudetenland”, the German-speaking part of what was then Czechoslovakia. If he did, that would likely lead to a world war. Chamberlain was desperate to avoid this.

The UK had elected Chamberlain as Prime Minister the previous year. But his background was as a businessman, and he was more experienced with domestic policy than foreign policy. He decided to meet Hitler in Germany to see if he could avoid war. At the time, most people in Britain thought that was a good idea. Chamberlain was the only Allied leader at the time to spend any significant time with Hitler.

When Chamberlain and Hitler did meet, Hitler made is clear he was going to invade the Sudetenland. Chamberlain asked Hitler if that’s all he wanted. Hitler said yes. Chamberlain flew back to Germany two more times after that. On his final visit, Chamberlain got Hitler to sign Munich Agreement. The agreement was that Hitler would have the Sudetenland, but Germany would not make any further demands for land in Europe. Hitler signed it enthusiastically. On Chamberlain’s return to Britain, he waved the agreement to the crowd and famously declared it showed “peace in our time”.

In less than six months, Hitler broke the Munich Agreement and invaded the rest of Czechoslovakia.

Gladwell says Chamberlain’s mistake isn’t simply because he was inexperienced in foreign policy. His foreign secretary, Lord Halifax, also met Hitler and believed that Hitler didn’t want to go to war. And the British ambassador to Germany, Neville Henderson, spent more time with Hitler than any other British diplomat and similarly believed that Hitler didn’t want war. In contrast, Winston Churchill, who never met Hitler, always believed Hitler wanted war. One of Chamberlain’s cabinet ministers, Duff Cooper, also never met Hitler and was horrified by Chamberlain’s account of his meeting.

Gladwell uses this as an example of how we, like Chamberlain, believe that the information we gather from a personal interaction with a stranger is uniquely valuable. We don’t hire babysitters for our children or employees for our company without meeting them first. Yet the information Chamberlain got from his visits to Germany didn’t help him see Hitler more clearly – it did the opposite. [This seems cherry-picked. Hitler could’ve been unusually charismatic in person, but that doesn’t mean meeting someone will necessarily give you a less accurate impression of them than not meeting them. Moreover, as Gladwell himself recognises, many who never met Hitler also thought Chamberlain’s visit had been successful at averting war. There are simply far more people who never met Hitler than people who did, so it’s very easy to find examples of people who never met him and “saw the truth of him”.]

Ana Montes – Cuban Spy

Ana Montes was a Cuban expert in the Defense Intelligence Agency (DIA). She was a rising star with multiple promotions and glowing reviews. Her nickname in the intelligence community was the “Queen of Cuba”.

Montes turned out to be a Cuban spy. She’d been spying for the Cubans ever since she joined the DIA. Reg Brown, a counterintelligence analyst also at the DIA, had some suspicions about Montes when Cuba shot down two Hermanos al Rescate (Brothers to the Rescue) planes. Brown took his suspicions to Scott Carmichael, a counterintelligence officer. Carmichael did interview Montes, but he found her explanations persuasive enough. He only saw red flags in retrospect.

Even Montes’ family didn’t suspect her. Montes had a brother and sister, who were both FBI agents. Neither of them had any idea. Her boyfriend worked for the Pentagon and specialised in Latin American intelligence. He still didn’t know. [So this example shows that we don’t just misunderstand strangers. We just misunderstand people, period.]

Gladwell uses this case as an example of default to truth. He argues that the reason Montes was able to fool so many, including Carmichael, for so long was because of that tendency. They only eventually found out about Montes because the National Security Agency (NSA) were able to break some codes that Cuba had been using to communicate with their agents.

Bernie Madoff – Ponzi Scheme

Many people in Wall Street thought that something didn’t add up with Madoff. Madoff claimed that he made his profits by trading derivatives. Harry Markopolos knew that, to trade derivatives in the volumes that Madoff was doing, Madoff would have to go to one of the largest five banks. So Markopolos talked to the banks. But none of them were trading with Madoff.

Markopolos was alone in not defaulting to truth in the Madoff case. He tried multiple times to get the Securities and Exchange Commission to investigate Madoff and shut down his Ponzi scheme but they didn’t – at least not for a long time.

Markopolos was an inherently suspicious guy. To him, dishonesty and stupidity are everywhere. He thought that between 20 to 25% of public companies were cheating on their financial statements. Gladwell uses the Russian term yurodivy or “Holy Fool” to describe Markopolos. The Holy Fool tells the truth because he is an outcast. He does not have a stake in the society’s institutions, so he is free to act as a whistleblower. Holy Fools do not default to truth.

Jerry Sandusky – Boy in Shower Sexual Assault

In 2001, an assistant coach, Michael McQueary, saw Jerry Sandusky showering with a young boy in the locker room on a Friday evening. He thought he heard some sexual “slapping” sounds.

McQueary, shaken, went and told his head coach, Joe Paterno. Paterno told his boss, Tim Curley. Curley told another senior administrator at the university, Gary Schultz. Schultz told the school’s president, Graham Spanier. An investigation followed but not for another decade. Curley and Schultz were both charged with conspiracy, obstruction of justice and failure to report child abuse. Both went to jail. (Paterno died shortly after Sandusky was convicted.) Spanier was fired and eventually convicted of child endangerment.

Gladwell is quite sceptical about the Sandusky case and points out there are many things that don’t add up. For example, some dates don’t line up, and some witnesses, including the actual boy in the shower (Allan Myers), changed their evidence, etc. Gladwell also points out that, before the 2001 shower incident that McQueary witnessed, there were two complaints against Sandusky but neither led to Sandusky’s arrest. He suggests this is because of default to truth.

Overall, Gladwell thinks that Spanier (and maybe Curley and Schultz) were treated unfairly. Reports of the incident further up the chain described it as “horseplay” and “horsin’ around”, not as sodomy or sexual assault, so it was natural for Spanier, Curley and Schultz to default to truth and not press further. [I point out some issues with that reasoning in my criticisms of the book here.]

Larry Nassar – Gymnastics Doctor Sexual Assaults

Nassar was the gymnastics physician for the USA Gymnastics national women’s team. He used his position as a cover, so he could finger his patients for his own sexual pleasure. Sometimes he even did this in front of the patient’s parents! Nassar did this for years. He had hundreds of accusers, telling very similar stories. The police found a 37,000 child pornography photos (including photos of his patients) in his hard drive.

Gladwell explains that over the course of Nassar’s career, there were as many as 14 occasions in which people in positions of authority were warned about him, but nothing happened. Even the parents of Nassar’s patients defended him. In doing so, they weren’t trying to protect larger institutional or financial interests. They just defaulted to truth.

[I feel like “default to truth” is an inaccurate term here. There were conflicting truths. Nassar’s version of the truth conflicted with some gymnasts’ versions. I think it’s more accurate to say people default to believing that people aren’t criminals, or to believing that there’s been a misunderstanding instead of repeated and flagrant sexual assaults of gymnasts in front of their own parents. And frankly, in most cases a misunderstanding would be the more likely explanation.]

Amanda Knox – Italy murder case

Amanda Knox came home one day and saw blood in the bathroom. She and her boyfriend, Raffaele Sollecito, called the police. The police found Knox’s roommate, Meredith Kercher, dead in her bedroom. The police treated Knox and Sollecito as suspects. They suspected that there had been some drug- and alcohol-fueled sex game gone wrong.

Knox was not, contrary to popular belief, a brazen femme fatale. She was actually sort of a quirky misfit. People found many of her behaviours hard to understand, and the media spun it out of control. She was seen buying red underwear the day after Kercher’s death. But that was because she couldn’t get into her house, which was a crime scene. She did the splits once at a police station. This was because she had been stretching, and an officer said she was really flexible and asked what else she could do. Knox probably was a bit weird. Being weird isn’t a crime.

Gladwell uses Knox as an example of a mismatched person. He says that, in hindsight, it was completely clear that Knox and Sollecito had nothing to do with the murder. There was no physical evidence linking them to the crime. The police investigation was shockingly inept. The actual murderer, Rudy Guede, had a criminal history and had been in the Kercher’s room on the night of the murder. His DNA was all over the crime scene. He fled for Germany as soon as the authorities discovered Kercher’s body.

Brock Turner – Stanford Sexual Assault

This is easily the most controversial part of Talking to Strangers.

Brock Turner was a 19-year-old freshman at Stanford. Two Swedish graduate students found him on top of an unconscious and partially-naked woman, thrusting, near a dumpster outside a frat party. He admitted to at least fingering the woman. Talking to Strangers refers to the woman, Chanel Miller, as Emily Doe. In 2019, she gave up her anonymity and released a memoir, Know My Name.

Gladwell says that when the rules for establishing sexual consent are clear, each party can accurately infer what the other wants. But he argues that there are no rules at all. To support this claim, Gladwell refers to a 2015 Washington Post/Kaiser Family Foundation survey. The survey asked college students whether they thought certain behaviours – e.g. taking off their own clothes, getting a condom, nodding, not saying “no” – established consent for more sexual activity. Results differed. On “nodding” for example, 54% thought that this established consent, 40% thought it did not (the rest were “Depends” or “No Opinion”).

Gladwell says:

[As I explain in my criticisms of this book, there are many problems with this. I also note that just because people disagree on what they think the rules are, doesn’t mean there are no rules. If you asked students whether it is illegal for an employer to ask a job applicant if they have kids, you will get different answers. That doesn’t mean there’s no law; it just means students have an imperfect understanding of the law.]

Gladwell then goes on to talk about how alcohol works and why it makes it much harder to make sense of strangers.

How alcohol works

Dwight Heath, an anthropologist, went to Bolivia and spent some time living with the Camba people. The Camba people lived in a poor and undeveloped part of the world. They held drinking parties with 180-proof alcohol (which is like laboratory-grade stuff) every weekend from Saturday night until Monday morning. Surprisingly, Heath found the Camba people had no social pathologies. No arguments, sexual or social aggression, or even alcoholism. Many other anthropologists subsequently found the same thing.

People used to think that alcohol acted as a disinhibitor. Many people who study alcohol no longer think that. Rather, they think that it makes people more myopic (short-sighted). Claude Steele and Robert Josephs first suggested the myopia theory. Alcohol makes us focus on immediate, visible consequences, and ignore the longer-term ones.

This explains why the Camba people were able to drink without many social problems. Their drinking was very structured, with strict rules. They drank in a circle, never alone, and only on weekends. They passed around one bottle at a time.

According to Gladwell, when we drink alcohol it affects the following parts of our brain:

• The frontal lobe is the part of the brain that governs higher order thinking such as attention, motivation, and planning. Alcohol dampens activity in the frontal lobe, making us a bit dumber.
• Alcohol stimulates the reward centres of the brain.
• The amygdala is the part of our brain that tells us how to react to things. Alcohol turns down the amygdala.
• The cerebellum is involved in balance and coordination. Alcohol affects this area too, making us stumble when drunk.
• The hippocampus is the part of our brain that forms memories. At a blood alcohol level of around 0.08, the hippocampus is affected. Once that happens, there’s no particular logic to which experiences we remember and which ones we don’t. Unintuitively, it’s not like we are more likely to remember more salient experiences (such as rape). At a blood alcohol level of around 0.15, the hippocampus shuts down entirely – i.e. people blackout. But at this level, it’s entirely possible the other parts of the brain continue to function mostly normally. So people can do reasonably complicated things like buying stuff on Amazon in a “blackout” state. They just won’t remember.

Gladwell says that this means it’s really hard to tell, just by looking at someone, whether they’ve blacked out. [I mean, this may well be true in the sense that Gladwell uses the term “blackout”. But people usually use the term “blackout” to mean passing out, which is not what Gladwell means here. In the Brock Turner case, it wasn’t like Chanel Miller’s hippocampus shut down and the other parts of her brain were mostly working. Miller wasn’t buying stuff on Amazon. She was unconscious.]

Terrorist Interrogation Techniques

James Mitchell and Bruce Jessen were CIA psychologists specialising in interrogation. They met when they were psychologists for the Air Force. Mitchell and Jessen taught people what to do if they got captured. For example, they would subject officers to interrogation techniques and try to get information out of them. They also tried out interrogation techniques on themselves.

Khalid Sheikh Mohammed (KSM) was a senior al Qaeda official. The CIA managed to capture him and interrogated him. The interrogation techniques used are controversial. Some would call it torture. Others just call them “enhanced interrogation techniques” (EITs). Mitchell and Jessen interrogated KSM for 3 weeks before he stopped resisting. But it’s questionable how reliable the information he gave up was.

A psychiatrist, Charles Morgan, specialised in post-traumatic stress syndrome. He conducted some experiments testing soldiers’ ability to recall information after a tough interrogation. Morgan found that people being interrogated were much worse at recalling information accurately. For example, after being interrogated, he asked soldiers to identify out of a lineup the person that had ordered their punishments. 20 out of 52 mistakenly picked the doctor.

Gladwell uses this example to point out that the “truth” is not something that we can extract from a stranger if only we try hard enough. He says we need to accept that there are limits to how we understand strangers. We will never know the whole truth.

Sandra Bland – Suicide after getting pulled over

Talking to Strangers starts and ends with the Sandra Bland case.

In 2015, Bland was a 28-year-old African American woman. She had just moved from Chicago to start a new job in Texas, at Prairie View University. A police officer, Brian Encinia, pulled her over and starts asking her questions.

Footage of their interaction is available here. Briefly, what happened is that Encinia pulled her over for failing to signal when changing lanes. Bland had only changed lanes because Encinia had been tailing her, so she moved over to let him pass. So she’s a bit miffed and she explains why (pretty reasonably, I might add). Encinia asks if she’s done. She says she is.

Bland then takes out a cigarette and lights it to calm her nerves. Encinia asks her to put it out. Bland refuses to, since Encinia has no right to ask her to stop smoking. Encinia then asks her to step out of her car. Bland refuses. She repeatedly says that Encinia does not have the right to do that; Encinia says he does. Things get heated. Encinia calls for backup. He then threatens to physically drag her out of the car and to use his taser on her. Bland gets out and is handcuffed. She cannot believe Encinia is doing all of this over a failure to signal.

Encinia takes Bland into custody on felony assault charges. Three days later she killed herself in her cell.

Many people who saw this case saw it as a case of a rogue or racist cop bullying an African-American woman just because he could. Gladwell disagrees. He thinks that Encinia was doing exactly what he was trained to do. It’s just that the training and deployment of police that were flawed.

Aggressive police tactics in low-crime area

The first problem was that Encinia was using “aggressive” police tactics, like those in the Kansas City experiments, but in a low crime place. Gladwell points out that stopping a car on a pretext is a search for a needle in a haystack. The vast majority of the time, the cars won’t contain anything incriminating. For this strategy to work, officers cannot default to truth – they have to do the opposite. The most savvy criminals will be careful not to give anything away. So to catch those criminals, officers have to assume that every driver may be hiding something. They have to drag things out and be alert to the tiniest clues. For example, out-of-state licence plates could be a clue. Fast food packaging in a car could suggest the driver is reluctant to leave their car unattended.

Now, in a crime hotspot, this type of policing might be okay. While stopping someone on a pretext and scrutinising every word they say is still an intrusion, the community may still be comfortable with it if it means that police catch a lot more criminals. Confining this style of policing to a hotspot makes the “haystack” a little smaller. It’s a delicate balance.

But this style of policing had become commonplace in low crime areas, like Prairie View, Texas. In a low crime area, there would be too many false positives to justify such aggressive policing (i.e. the base rate for crime is low). Police would just end up harassing innocent people with nothing to show for it. Worse than that, it destroys any trust that community may have in the police.

Bland was mismatched

The second main problem was that Bland was “mismatched”. Police departments train officers to believe people are transparent and to look out for certain demeanour cues. But that training is just plain wrong.

Two-thirds of US police departments use the Reid Technique. The Reid Technique teaches police to use demeanour to judge innocence and guilt. In an extremely long footnote, Gladwell describes the Reid Technique as “nonsense”. He refers to research done by criminologist Richard R Johnson based on old episodes of Cops. For example, the Reid Technique claims that liars will look away (“gaze aversion”) while truthful people will maintain eye contact. Johnson found that innocent black people are actually less likely to look police in the eye than black suspects. White people, as a group, are far more likely to look police in the eye than black people. And this is especially true for white people suspected of a crime!

There is also enormous variability in how people act regardless of their innocence. Some innocent people never make eye contact while other innocent people make lots of eye contact. The same is true for cues like smiling.

Gladwell thinks that Encinia was genuinely afraid for his safety as Bland was acting unusual and he thought at one point she could be reaching for a gun or other weapon. But Bland felt upset. She had been stopped by the police 5 times before in her life, and those stops had left her with almost $8,000 in outstanding fines. She had also tried to commit suicide the year before, after losing a baby, and had ongoing mental health struggles. Bland had just moved to a new town, starting a new job, hopeful for a fresh start. And a police officer pulls her over on her first day for the most trivial offence. Of course she feels upset.

Gladwell says that Sandra Bland’s death is what happens when a society does not know how to talk to strangers.

Other Interesting Points

• Gladwell mentions a study of bail decisions conducted by Sendhil Mullainathan (Kahneman, Sibony and Sunstein in Noise also refer to this study). The study found that the defendants the computer suggested releasing, based only on their age and previous record, were 25% less likely to commit a crime while on bail than the defendants released by the judges. This is not because the judges had a higher risk threshold, either. The judges released almost half of defendants the computer had identified as high risk.
• Gladwell says that we need the criminal justice system, the hiring process to be “human”. But the requirement of humanity means that we have to tolerate an enormous amount of error. [Why? Why can’t we improve these systems by adopting some of the measures suggested in the book Noise? The problem with Gladwell is he doesn’t really suggest any solutions other than maybe “be less quick to judge others”.]
• In footnote 2 to Chapter Six, Gladwell explains how he spent a lot of time in his book Blink to Paul Ekman’s work. Based on that work, Gladwell wrote that whenever we experience a basic emotion, our facial muscles automatically express that emotion, even if just fleetingly. He issues a semi-retraction of that in the footnote here. Gladwell admits that Ekmans’ paper is, on reflection, a little strange and that his results haven’t replicated.

My Thoughts

I have a like-hate relationship with Gladwell. He is a very good writer, but not much of a thinker.

Gladwell tells interesting stories. In Talking to Strangers, he manages to link together a bunch of high-profile, engaging, and controversial news stories that appear to be unrelated. He does this by first picking an incredibly broad theme. “Talking to strangers” could just as easily be “talking to people” or “human interactions” or, frankly, just “life”. His examples often don’t involve “strangers” in the ordinary sense as the people often know each other (e.g. Ana Montes deceived her own family, Larry Nassar deceived long-time patients and their parents). And then he shoehorns his examples into that theme, even if they don’t fit (I comment on this more below).

But I don’t think Gladwell is much of a thinker. He’s certainly not very rigorous, as I explain in detail here. Yet he states his conclusions on difficult issues with confidence. He’ll string a bunch of cherry-picked anecdotes together that support the point he’s making, and he’ll completely ignore counterarguments (or straw-man them).

So when I read Gladwell, I take him with a grain of salt. I enjoy the stories for what they are, but I don’t place much if any weight on Gladwell’s opinions of them. (And this was the case before Talking to Strangers; I didn’t think much of The Tipping Point either.)

If, after all that, you still want to buy Talking to Strangers, you can do so at: Amazon | Kobo <– These are affiliate links, which means I may earn a small commission if you make a purchase. I’d be grateful if you considered supporting the site in this way! 🙂

Do you think I’m being too harsh on Gladwell? Share your thoughts in the comments below!

If you enjoyed this summary of Talking to Strangers, you may also want to check out:

• Criticisms of Talking to Strangers by Malcolm Gladwell
• Summary of Thinking in Bets by Annie Duke
• Summary of Noise by Daniel Kahneman, Olivier Sibony and Cass Sunstein