Why I Rarely Write About Psychology
29 October 2026 · @Mythogen Engine
I rarely write about psychology directly. Not because it doesn't interest me, but because it is too complex. Why a person acts the way they do involves how they grew up, the wider society, their circumstances at the moment, the people around them — and any one of these can change the answer. Explaining all that properly takes more information and more time than an article can usually carry.
Explaining how psychology gets abused, by contrast, is simple. The methods of abuse are few. They come down to the same move every time: replacing a person's behaviour with a label.
So this is not an introduction to psychology. It is my own experience of it — from the first time I met a psychologist who truly understood people, to the "psychological analysis" I later ran into at school, at work and online.
First encounter: a man who asked "why"
I'm grateful that my first encounter with psychology came through Koo Hsiu-chuan (顧修全).
Anyone who grew up in Hong Kong in the 1980s and 90s probably saw him on television. He studied psychology at the University of Hong Kong, spent three years training in psychoanalysis under Professor Norman Briers, and took his first job as a clinical psychologist in the Prisons Department. In 1981 he opened his own private practice. He analysed celebrities on TV and published books and cassette tapes. I listened to his tape To Hell with Inferiority (《去你的自卑》) and read one of his early books, though I no longer remember its title.
What stayed with me most was his analysis of Stephen Chow on a show hosted by Carol Cheng. He started from Chow's childhood and worked forward to why Chow had such a strong urge to control, why he would refuse to stay beneath anyone, and why he would one day become his own boss. Years later, Chow did move behind the camera as a director and producer.
Looking back, what I remember is not whether the "prediction" came true, but how he analysed. His question was "how did this person become who he is?" — and he traced it from upbringing all the way to motive. That taught me psychology is not all nonsense. It can be a serious discipline for understanding people.
What real psychology costs
Only after reading more did I understand how much goes into that kind of analysis:
- Long-term clinical observation — not one meeting, not one block of text.
- Many factors at once: upbringing, the wider society, present circumstances.
- Repeated verification: a judgement has to survive the next observation that might overturn it.
- Attention to changes in expression and behaviour, not just to what the person says.
Even with all of this, nothing is guaranteed to be accurate. Psychology is, in a sense, an informal kind of statistics: it describes tendencies and probabilities, not laws. A theory that fits one person may not fit another, or the same person in a different situation. The real skill is knowing which lens to use when.
Psychologists have argued about this themselves. In the 1960s, Walter Mischel showed that behaviour varies a great deal across situations, and that personality types alone predict how someone will act only weakly.
I have worked in sales, sales support, marketing and back-end engineering, and each role demanded a different way of working. I don't have a fixed "type". I look at what needs doing, what the goal requires and whether I can live with it in good conscience — then I use whatever method fits best. Anyone with only one fixed way of working, who can't see the whole picture or where they stand in it, can do one role at best, and perhaps none.
So people who truly understand psychology are usually careful about drawing conclusions. When the conditions aren't met, a conclusion is only a reference point.
School: the label first, the evidence later
The trouble is that most of the "psychological analysis" I met afterwards stripped away all of these conditions and kept only the conclusion.
The most common place was school. A child doesn't listen in class. There can be many reasons: the teacher reads straight from the textbook, the material is too easy, the child reads faster alone. Children's attention spans are limited anyway. If their grades aren't affected, thinking about something else is perfectly normal.
But in a school's logic, not listening needs an explanation, and the cheapest explanation is a label: self-centred, attention-seeking, an attention problem. Once the label is in place, the problem becomes "this child is defective" instead of "why is this lesson so dull?"
Then comes punishment: copying out school rules, meetings with parents, detention. It eats the child's time and energy and keeps them from the lesson, until their grades really do fall. And the school can say: see, I told you something was wrong with him. The label came first. The evidence was manufactured afterwards.
Defending yourself is hard. Society generally doesn't believe teachers get things wrong, and most parents and onlookers are good rule-followers who see compliance as maturity. Worse, the label can't be argued with. Explain yourself, and you're making excuses. Push back, and it's an attitude problem. Stay silent, and you're avoiding it. The rebuttal itself becomes the evidence.
I'm not saying labels are useless. Some children genuinely struggle with attention, and a diagnosis is their way into support. The difference lies in what the label is for: finding support, or filing someone away, shifting blame and ending the conversation.
Work: it doesn't test your personality, it tests how fast you turn
At work, the tools change: HR psychology training, company-provided counselling, and the most popular of all, MBTI.
At dinners, plenty of people proudly announce that they're an "E". At one multinational I worked for, most of the people I dealt with said they were Es. Work there ran on connections and socialising, so even naturally introverted people had to act extroverted. I don't think MBTI is a useful psychological test. It is closer to a labelling tool.
The data doesn't support it either. The most widely cited source is Pittenger's review: roughly half of people who retake the test five weeks later get a different four-letter type. The reason is that it forces continuous traits into either-or categories. A slightly different set of answers, with no real change in personality, flips someone from I to E. To be fair, the scores on each dimension are reasonably stable. The problem is that people use the four letters.
More telling still, the publisher itself says using MBTI for hiring is unethical, that taking it must be voluntary, and that results are confidential and belong to the respondent. In other words, a company that uses it to assign, assess or judge people has gone beyond what even its maker accepts.
MBTI is self-reported. What you answer is how you see yourself, or what you think you ought to be. Taken at a company, it becomes a performance: people answer as the type the company wants. And once labelled, people slowly grow into the label.
There is a subtler game too. This season the boss likes Is, so everyone acts like an I. Next season the wind changes, and the question is who turns fastest. After a few rounds the four letters mean nothing. What is really being measured is how quickly you switch sides. Fast turners are called "adaptable". Those who stay put earn a new label: "inflexible", "resistant to change".
Whether this game is deliberate is hard for an outsider to prove. It may just be that the boss's taste changed and people below chased it. The result is the same either way. Played long enough, people can no longer tell what they actually think from what they think to match this season's wind. This is what I understand as collective gaslighting: no one tells you outright that you're wrong; you simply stop trusting your own judgement. It is hard to prove precisely because the evidence is itself psychological interpretation. There is nothing objective to put on the table.
It also looks effective because of selection. People who see through it leave, or don't have time to care. People who have been manipulated stay, and can't tell. Everyone who remains "proves" the method works; nobody counts the ones who left. It is the same trap as lying with statistics.
Online: a framework that explains everything
Online there is another version. On Facebook, a popular writer works from Jungian psychology, with many followers, students and admirers. When he introduces theory, I have no objection. The problem is that he uses Jung to explain everything.
Persona, shadow, projection, complex — these concepts are broad enough that almost any behaviour fits. If you please others, it's the persona. If you don't, it's the repressed shadow pushing back. If you save money, it's a survival complex. If you don't, it's avoidance. A framework that can explain anything in fact explains nothing, because no situation could ever prove it wrong. Readers find it "so accurate" largely because it cannot fail to be.
I have read two of his analyses. One concerned a young woman with assets of about NT$13.5 million who asked online what to do next; he read it as a "wealth complex in transition". But in Taipei, that money cannot buy a home where she could stop working, so much of her anxiety is arithmetic, not a complex. The other concerned a divorced woman's experience on dating apps; he read it as the "good woman persona". Her own numbers — roughly 1,500 swipes for a single partner — contradicted her claim of having plenty of choice. He didn't deal with that, keeping only the psychological layer. The parts that could be checked disappeared; what remained was interpretation that can't be refuted.
His pinned comment states the model plainly: public posts describe phenomena, while the subscription tier walks with you "inward", mapping your projections, shadows and complexes. I'm not saying he does this on purpose. But someone who earns subscriptions by going "inward" will naturally read everything as an inward problem. Structural problems have no course to sell. Psychological ones do.
Jung himself was a clinician who spent a long time with the same patient, over a long series of dreams, before reaching a judgement. Using his vocabulary is not the same as using his method.
Play is fine, verdicts are not
I play personality quizzes myself, and I know roughly how they work. As a game, they're harmless, even fun. The worst kind spins every result in a flattering direction, which is really just a form of self-gratification.
Cold reading works the same way. Like face reading, it sounds fluent, the listener feels it's accurate, yet most of it is vague. Astrology too: people who believe in it slowly grow closer to their sign's description, a self-fulfilling prophecy, while those who don't believe find it doesn't fit at all.
Story analysis is no different. When I wrote my piece comparing Detective Conan and Steins;Gate, I explained why I'd choose Ai Haibara over Ran Mouri. Many people love to analyse what a character's choice reveals about psychological projection. But a story gives so little information that no outsider can do real psychological analysis. My piece declared a position; it was not a diagnosis.
So my line is simple. Introducing theory: fine. Playing quizzes: fine. Using it to pass a verdict on a real person: not fine.
Why this is a social problem
Psychology itself is not dangerous. What is dangerous is turning it from understanding a person into passing a verdict on one.
This isn't a problem with a few individuals. It is a social problem, for three reasons:
- The labelled lose the right to answer back. A label can't be refuted, and the rebuttal itself becomes evidence.
- Structural problems never get fixed. Dull lessons, broken company structures, badly designed systems — all of it gets converted into "this person has a problem". Once the fault lands on the individual, the structure never has to change, and it can be passed down from one generation to the next.
- It looks effective because of selection. Those who see through it leave; those who are manipulated stay, and can't tell.
Real psychology asks: how did this person become who they are? Abused psychology asks: which category does this person belong to? The first needs time, information and repeated verification. The second needs only a label.
That is why I rarely write about psychology directly. To write about it well, I usually need more information than I have. But to write about how it gets abused, what I have is already enough.