Psychological Shibboleths
I’ve always loved the word “shibboleth”. It sounds like one of H.P. Lovecraft’s Great Old Ones, a horrific creature living “not in the spaces we know, but between them”. In fact, I was rather disappointed when as a child I found out that it wasn’t a monster; it was just that Hebrew makes everything sound bad-ass. While not monstrous, it did turn out to have a gruesome origin: Judge Jephtha (a kind of Biblical Judge Dredd) used the word to test fleeing Ephraimites who were trying to pass themselves off as Gileadites (Jephtha’s own clan). Apparently, Ephraimites couldn’t pronounce the “sh” sound. “Shibboleth” came to mean any word that could identify a foreigner by their pronunciation; for example, if you were Welsh you could use “Llangollen” (or indeed almost any word in Welsh) as a shibboleth. Later the sense expanded to include any behaviour that identified you as a member of an in-group, and then, in turn, the values and taboos of an in-group that could not be questioned.
In psychology, a number of events have become shibboleths, in that familiarity with them demonstrates that you’ve taken PSY 101. For those who don’t critically examine the content of their textbooks, they also become shibboleths in the sense of unquestionable beliefs. Now I’m not a psychologist, but as a linguist and educator, I need to know a fair bit of psychology, plus it overlaps with other areas I’m interested in and write about, like happiness studies and game studies. Over the years, I’ve found that several of these shibboleths were at best overstated, and at worst complete nonsense.
Kitty Genovese and the Bystander Effect
Bystander apathy is a real, and well-documented phenomenon. It is also exaggerated, and so is the event that thrust it into the limelight and social psychology textbooks: the murder of Kitty Genovese. Here is how the event was described by the New York Times in 1964.
For more than half an hour 38 respectable, law‐abiding citizens in Queens watched a killer stalk and stab a woman in three separate attacks in Kew Gardens.
Twice the sound of their voices and the sudden glow of their bedroom Iights interrupted him and frightened him off. Each time he returned, sought her out and stabbed her again. Not one person telephoned the police during the assault; one witness called after the woman was dead.
“How could 38 people stand by and watch a woman getting murdered?” asked an outraged public. Psychologists rushed to dub this the “bystander effect” whereby people who would normally rush to someone’s aid will shrug their shoulders and walk on by if they think somebody else will do the job for them. The explanation was accepted uncritically because it tied in with a pervasive mood of alienation and the feeling that urban communities were disintegrating. There is, however, a simpler explanation, which is that 38 people did not stand by and watch a woman being murdered. On closer inspection, some things did not add up.
There were not 38 witnesses; the DA said there were only “about half a dozen that saw what was going on.” “Saw” is also a generous term, given that in those days the street was dimly lit.
None of those who saw the attack saw all of it. There were two (not three) attacks; the first could have been seen by some of the residents, but the second took place in a stairwell. Sure, people heard screams, but many told police that they just assumed it was a couple having a screaming row — not unreasonable given that there was a bar on the street (closed that night because of the number of complaints about drunken altercations).
Finally, it appears that not only did more than one witness call the police, one of them yelled out at the attacker to leave her alone, and one rushed out to help her, holding the dying woman while waiting for the police to arrive. (It’s even been suggested that the police promoted the bystander apathy story to excuse their late arrival and the fact that they ignored the first witness that called them.)
In a final irony, the murderer was later apprehended by the opposite of bystander apathy: while he was trying to steal a TV from a home, neighbours confronted him and disabled his car for good measure.
This doesn’t mean that there is no such thing as the bystander effect. We’ve all been in situations where we felt we ought to do something but believed (or hoped) that someone else would do it, but this normally applies to things like doing the washing up or talking to that homeless guy who’s curled up in a corner muttering to himself. The idea that 38 people would sit and watch someone being stabbed to death because they couldn’t be bothered to pick up the phone is absurd, yet this urban legend is still in circulation.
Milgram’s Obedient Torturers
In the early 1960s, Stanley Milgram performed a famous experiment that seems designed to show that deep down we’re all Nazis. Subjects were required to ask a series of test questions, and to give the testee an electric shock whenever he got it wrong. As the testee gave more incorrect answers, the shock increased. Unknown to the subjects, they were asking the questions to a professional actor, and no electricity was involved. As the shocks increased, the actor pretended to feel pain, and later pleaded with the experimenter to stop; at the highest “voltages” the actor even feigned heart failure while subjects kept jolting the seemingly lifeless body. Although subjects were at first reluctant to press the shock button, the experimenter would give reassurances that the testee was actually fine, and when resistance got greater, ordered them outright. 65% of subjects went all the way up to the maximum shock, an absurd 330 volts.
“Surely humans couldn’t be so blindly obedient and sadistic!” exclaimed a shocked public. “Sure they can, you hippie wusses!” said Stanley (or something like that, anyway). At the time, Nazi Germany was a recent memory and the Cold War was in full swing; understandably, the fear that obedience to authority could make ordinary people do terrible things was enough to make Milgram a psychology superstar. The obedience experiment has appeared in every social psychology textbook ever since, sending a clear message to undergraduates that 65% of people will torture their fellow humans just because a man in a white coat told them to. We’re not just sheeple, we’re nasty sheeple.
Of course you could look at it from the other perspective and celebrate the fact that 35% of people will tell an authority figure to go screw themselves if they think there’s something fishy going on. We could also question the data. In an article for the journal Teaching of Psychology, Richard Griggs lists a slew of problems with Milgram’s experimental method, most notably “Milgram’s own unpublished analysis that shows that the majority of the participants disobeyed when they thought that the learner was actually being shocked.” Wait a second and let that sink in. All Milgram proved, in that case, is that most people are prepared to pretend to cause pain to someone who is pretending to feel it. It’s like a BDSM dungeon with paper whips.
Another interesting slip-up is Milgram’s failure to report the fact that the strongest “prod” for the subject to continue — the outright order — was never obeyed. There were some people who could be cajoled into shocking a bit more then they would, and some Ramsey Bolton types who would happily shock away until the victim was seemingly unconscious, but nobody could be bullied into shocking beyond what they thought was acceptable.
Attempts to replicate Milgram’s findings have also led to their reinterpretation. It’s certainly true that people are prepared to shock each other in the name of science, but the fact that outright orders don’t work weakens the obedience aspect; it looks like the subjects shock the testees because they think it was the right thing to do, which has led some researchers to talk about “active collaboration” rather than “obedience” (and also ties in with what we know about Nazis). An interesting quirk comes from a discrepancy between Milgram’s own experiments. In most runs, the testee complained to the experimenter — the same mad scientist who was ordering the subjects to carry on shocking. In one case, though, the testee complained directly to the subject, and there the results were reversed; rather than 65% obedience, there was only 15%, or in other words, 85% told mad scientist where he could put his electrodes. So much for obedience.
These are just some of the anomalies making up Grigg’s case that Milgram’s experiment is a “contentious classic” that “introductory social psychology textbooks present … as an uncontentious classic.” Other than the fact that some people are cool with electric shocks and few people have any idea how much electricity 330V actually is, Milgram’s experiment tells us next to nothing.
The Stanford Prison Experiment
If there’s one thing psychologists love more than giving electric shocks, it’s having people give each other electric shocks. Another researcher who started like this was Philip Zimbardo. In his own words:
We took women students at New York University and made them anonymous. We put them in hoods, put them in the dark, took away their names, gave them numbers, and put them in small groups. And sure enough, within half an hour those sweet women were giving painful electric shocks to other women within an experimental setting.
While this might have led to a promising career in one of the darker niches of the porn industry, Zimbardo went on to create the infamous Stanford Prison Experiment — the other shibboleth that says it only takes a man in a white coat to bring out everyone’s inner Nazi. Student volunteers were divided into “prisoners” and “guards” and were to be incarcerated in a basement area on campus for twelve days. As it happened, Zimbardo called a halt after only six days because of the extreme behaviour of the guards, some of whom took sadistic glee in tormenting the prisoners, who in turn started to show signs — even physical signs — of psychological trauma.
The experiment went down in history, appearing in countless textbooks, invoked as evidence during the Abu Ghraib scandal, and even inspiring two films: Das Experiment (2001) and The Stanford Prison Experiment (2015). Yes, as late as 2015, people still think that the Stanford prison experiment is good science, or at least good film material.
The problems with Zimbardo’s work are as great as with Milgram’s. Skeptoid’s Brian Dunning has a good debunking of his methodology; here are the main points.
Like its fictional predecessor Lord of the Flies, the Stanford Prison Experiment used a sample — 24 male college students — that was neither large nor representative of the general population. More importantly, subjects were self-selected and the experiment was advertised as being about prison life. As Dunning says, “a large segment of the general population would be repulsed by such a concept, and you’ve got to have questions about anyone attracted to that idea.” Remember that this was 1971, before acting out fantasies about guards and prisoners was all over the Internet.
Zimbardo took part in the experiment himself as the prison superintendent, and apparently egged on the guards. This is not like the director of a film taking a cameo role; it’s like someone who’s researching LSD taking the stuff themselves and tripping out with the subjects, yelling “Whoa, do you see that purple dragon?!”
Zimbardo downplays the fact that most of the guards were not at all sadistic, and in fact some of them went out of their way to make the prisoners’ lives comfortable. It’s odd that Zimbardo didn’t find this noteworthy, given that this range of types from sadistic bastard through bored functionary to guardian angel is exactly what you see in real prisons.
One further consideration is that the subjects were aware that they were playing roles and this was not a real prison. One of the “sadistic” guards later explained that he was trying to act like Strother Martin’s character in Cool Hand Luke. Giving the guards uniforms and reflecting sunglasses not only increased anonymity, which was the intent, but also contributed to the sense of play-acting, which was not. Could the notorious Stanford Prison Experiment have simply been a LARP that got a bit out of hand?
Some evidence to support that idea comes from the BBC Prison Experiment, a re-run of Stanford filmed like a reality TV show and broadcast in 2002. It turned out nothing like the original. I watched with amusement as first the guards tried to institute a more humane system, then the prisoners had a revolution and turned the prison into a commune, and finally some of the revolutionary leaders decided to stage a coup. What was very clear was that (a) they knew they were in a role-playing game, and (b) they were having fun with the psychologists as well as each other. The psychologists fell for it, terminating the experiment early, just as Zimbardo had done. Having people put on costumes and take on roles is a great way to study role-playing, and a pretty useless way to study most other things because, contrary to what psychologists like to tell us, most adults are actually pretty good at distinguishing between fantasy and reality.
Bandura’s Bobo Booboo
But what about children? It’s a psychological shibboleth that children cannot distinguish between fantasy and reality. Every time I do my course “The Philosophy and Psychology of Games” a student comes up with the story about a child who jumped out of a window because he thought he was Pikachu and could fly. (This occurs so predictably and memorably that I use “Pikachu fallacy” as a convenient shorthand for the fallacy of converse accident.) There’s a common idea that children cannot distinguish between fantasy and reality, which is why computer games are dangerous. When I was a kid, there was a similar story about a kid who jumped out of a window because he thought he was Superman, which shows that comic books are dangerous. And of course violent TV programmes must be dangerous for the same reason.
Psychologists have come up with a body of research showing that violent television makes people violent — most of it very bad research. One of the most famous studies comes from Albert Bandura, a guy who crops up in psychology textbooks almost as much as B.F. Skinner.
Bandura’s experiment took place in Stanford (yes, Stanford again) in 1961. He and two colleagues exposed toddlers to various scenarios with a Bobo doll. This is a large inflatable doll with a low centre of gravity, so that you can hit it and it will bounce back upright.
The first group of toddlers saw an adult assaulting the Bobo doll while exclaiming gleefully “Pow!” “Kick him!” “Sock him on the nose!” and so forth, as adults are wont to do when confronted by Bobo dolls. The second group watched a film of the same events, while a third group watched a cartoon cat doing the same thing to poor old Bobo. There was also a control group who saw nothing. Children in all four groups were deliberately frustrated by being shown toys they weren’t allowed to play with, which anyone who has looked after toddlers will inform you is quite enough to send many four-year-olds into fits of hysterical rage, then placed in a room with a bunch of toys including a Bobo doll.
Unsurprisingly the control group were the least likely to “exhibit aggressive behaviour”, which was what the experimenters had decided to call hitting the Bobo doll. The interesting thing is that both the film group and the cartoon group were more likely to hit the Bobo doll than the group that had seen someone actually hitting the Bobo doll. Now this could have opened up some fascinating lines of research into play, but this is not what Bandura concluded; rather, he stated that “The results of the present study provide strong evidence that exposure to filmed aggression heightens aggressive reactions in children.”
Come again? We are talking about children imitating an adult (or a cartoon cat) who is playing with a doll that is designed to be hit. The Pikachu fallacy aside, most children aren’t that bad at distinguishing between fantasy and reality. They may have some pretty weird ideas about what reality is (Santa Claus, fairies, monsters under the bed etc.) but they know the difference between a game and real life, and they sure as hell know the difference between hitting a Bobo doll and hitting a person, not least because most of them have tried hitting people and know it has painful consequences.
It is no coincidence that these shibboleths come from the USA during a certain brief time in its history. American society was changing, people were understandably anxious about the changes (not all of which were good), and psychologists were in the grip of a model called social learning theory. In a world where we were promised utopia and threatened with dystopia, they wanted something that would make sense of bad behaviour and came up with the idea that it was learned not individually, like rats in Skinner boxes, but collectively. As theories of behaviour go, it’s not a bad one, though it ignores factors that would not become intellectually acceptable for a few decades, like genetic influences, and a factor that psychologists tend to ignore as much as economists exaggerate: human beings possess reason, and tend to do what makes sense under the circumstances. The worrying thing is not social learning theory but the fact that several famous psychologists used it to justify some really bad research, and that this research in turn became canonical because it fitted the currently popular view of human nature.