Why Freud Really Cannot Represent Psychology
Freud, the Unfalsifiable
“Why do people dream?” This may be the most broadly and deeply misunderstood question in all of psychology. And we have Sigmund Freud’s The Interpretation of Dreams to thank for that.
What I mean, of course, is that the ideas in The Interpretation of Dreams are unreliable. This is not just my personal opinion; it is the consensus among contemporary scientists who study sleep and dreams.
In 2000, sleep scientist William Domhoff published a review analyzing a century of dream research after Freud. His conclusion was: “There is no reason to believe any of Freud’s interpretations of the purpose or symbolism of dreams.”
In plain English: Freud’s interpretation of dreams is completely unreliable.
And yet The Interpretation of Dreams occupies two utterly different positions in academia and among the general public. Even today, many people who become interested in psychology still start with The Interpretation of Dreams; countless people still treat the book’s ideas as gospel.
So since the misunderstanding runs so deep, let’s clear away the confusion before setting things straight. In this episode, we’ll talk about The Interpretation of Dreams and Freud’s psychoanalysis. In the next episode, we’ll discuss the contemporary theories of dreams in Why We Sleep.
But what I’m going to do next is not pull out the claims in The Interpretation of Dreams one by one and refute them. Rather than simply bash The Interpretation of Dreams, I’m using it as a way to talk about what science is and what mysticism is. This is obviously a huge question—an entire semester might not be enough to explain it completely—but through one or two entry points in this series, I think you can get a feel for it.
Let’s get to the point. The core theory Freud proposed in The Interpretation of Dreams is this: dreams are expressions of repressed desires after those desires have been disguised. Freud’s psychoanalytic theory holds that people have many instinctive desires, which are repressed into the unconscious because they fail to meet various social and moral standards.
Those desires want release but cannot be released directly, so they have to disguise themselves and emerge in roundabout ways. That is what dreaming is: in dreams, consciousness related to social norms and morality grows weaker, allowing desires that cannot be satisfied in reality to be packaged, concealed, and expressed.
Sounds pretty reasonable, right? So why is it unreliable?
Actually, this is not the first time I’ve said on the show that Freud and his school of psychoanalysis are unreliable. Whenever I bring it up, listeners leave comments asking why I have such a strong opinion of psychoanalysis—of Freud, Jung, and Wu Zhihong. Do I have some kind of bias? Am I secretly part of another school with an axe to grind?
Most of the replies I give to that question in the comments use a fairly simple reason: if you want to prove that an idea you’ve proposed is correct, you need to use data to prove it.
That’s how all modern science works. What school am I from? I don’t have a school. I only look at data. More than 90% of psychologists in the world don’t belong to a school either. As long as you have solid data, you can persuade us.
Traditionally, there have been roughly two ways to obtain data.
First, data obtained by conducting experiments. This is the strongest kind of data, because experiments can precisely control variables. The results can often tell you clearly that a change in variable a caused a change in variable b. Data like this is especially informative.
Second, somewhat weaker data obtained through large-scale surveys. If the survey is large enough and the sample broad enough, this data is fairly reliable too.
Besides these two traditional methods, we now have big-data technology, which gives us a third kind of data: “all-data.” If Alibaba wants to study shopping habits on Taobao, for example, it doesn’t need to run an experiment or draw a sample; it can simply analyze the entire Taobao database.
But the ideas of Freud, Jung, and Wu Zhihong are mazes leading from concept to concept and theory to theory. At most, they contain a few isolated cases—but they have no systematic data.
Making a theory sound internally coherent, arguing from thin air to thin air, is actually easy. Using a few isolated cases rather than systematic data to support it is easy too. But none of that is reliable.
For example, Wu Zhihong used several cases to argue that China is a nation of giant babies. But go ahead and look: you can find similar cases in the United States. Using exactly the same logic, you could conclude that the United States is a nation of giant babies!
Likewise, The Interpretation of Dreams contains only scattered cases and absolutely no systematic data. Wait—surely you’re going to ask next: Freud lived a hundred years ago, after all. Experiments, surveys, and big data hadn’t developed yet. Can’t later researchers use data to determine whether dreams really are “expressions of disguised desires”?
Sadly, no.
Why not?
Because most of Freud’s theories are unfalsifiable. In other words, no matter what evidence you produce, it can only prove that the theory is right. Let me repeat that: it’s not that you can’t produce evidence showing the theory is wrong; it’s that no matter what evidence you produce, it can only prove the theory right.
That is the deeper—and real—reason Freud’s theories are unreliable.
For example, Freud often said that “sex” is the most common kind of repressed desire. You see an attractive woman on the street and want to do something with her, but the law and morality won’t allow it, so your sexual desire is repressed and often appears in dreams in disguised form.
Freud also said that sexually repressed women would dream of all kinds of elongated objects—guns, skyscrapers, sticks. These objects were disguises for sexual desire because, supposedly, they were really male genitals.
If we think this claim sounds unreliable, is there any way to prove it wrong? Sorry, but according to Freud, no!
Suppose you surveyed the dreams of one hundred sexually repressed women and found virtually no elongated objects in them.
You take the data back through time to Vienna in 1900 and show it to Freud, triumphantly saying, “See? Your theory doesn’t hold up, does it?”
How would Freud respond?
He would probably do exactly what he did in The Interpretation of Dreams: analyze the dreams one by one. He might say, “There’s no elongated object in this dream, but in the dream she said a word to a man, and that word is actually a homophone of the male genitalia. These desires are repressed so deeply that you didn’t look carefully enough. With analysis as careful as mine, you can find them.”
Young man, too young too simple.
Freud got to decide which things in a dream counted as disguises for a given desire. Whatever dream you brought him, he could use it to prove his own theory right.
There are countless examples like this in The Interpretation of Dreams. Freud would confidently insist that a homophone, an object of some shape, or an ambiguous gesture in a dream was a disguise for sexual desire. Could you refute him? You couldn’t. If he said it was, then it was—even when it wasn’t.
Scientists have in fact tried to determine, with evidence, whether dreams are used to express disguised desires. The story I just told isn’t entirely made up.
One scientist really did recruit a group of male participants, have them abstain for two weeks—no masturbation and no sex—and record their dreams. The result was that their dreams contained no more sex-related themes than those of a non-abstinent control group.
In theory, this evidence should have refuted Freud’s view.
But in practice, what good did it do to show the evidence to Freud’s followers? They could casually find a whole pile of “sexual hints” in those abstinence dreams.
Freud’s ideas about dreams, along with the vast majority of his other ideas, cannot be falsified with data; there is no possibility of data disproving them. A theory like that is not science. It’s mysticism.
Do you love your mother?
Yes, I love her.
Good. Then you have an Oedipus complex.
Do you love your mother?
No, I hate her.
Oh, then you are repressing your nature. That means you love your mother in your unconscious, so you have an Oedipus complex.
—See? The most amusing thing about mysticism is that it is self-consistent: mutually contradictory data all prove the same idea correct. That is precisely why it can seem so fascinating.
Part of the reason Freud’s theories became so popular is right here: they appeared to explain everything. Every corner of human nature seemed to have been seen through by Freud. The moment The Interpretation of Dreams was finished, psychology had already reached its end—it was over.
This kind of mysticism looks infinitely wonderful, doesn’t it? But it has one enormous, black-hole-sized flaw: it is u-s-e-l-e-s-s.
What does “useless” mean?
Think about it: if any data whatsoever can produce the same theory, that also means the theory has no discriminating power when it comes to data.
And hidden right here is one weapon for exposing the unreliable truth about mysticism.
The General-Purpose Weapon for Cracking Mysticism
Now I’m going to teach you the general-purpose weapon that can crack all mysticism.
Let’s use the example from before. This time, we find one hundred sexually repressed women, then recruit another one hundred women matched with them in every way but sexually quite satisfied. We record the dreams of all two hundred over a period of time.
Then the next step can go one of two ways.
The Sommelier Teasing Method
The first one I call the “Sommelier Teasing Method.”
Someone once conducted an experiment in which the labels on $99 bottles of wine and $9 bottles of wine were swapped, then gave them to sommeliers to taste. The sommeliers ended up praising the wine labeled $99—which actually cost only $9—to the heavens.
Borrowing this method, we swap the labels on the two groups of women. We put the dreams of the sexually repressed group in a folder labeled “sexually satisfied,” and the dreams of the sexually satisfied group in a folder labeled “sexually repressed.”
Then we give them to psychoanalysts and ask them to analyze the differences between the two groups of dreams. I’m confident that nine out of ten psychoanalysts would find more sexual symbols in the dreams labeled “sexually repressed.”
This method isn’t something I just made up on the spot either. Someone really did it.
Psychoanalysts now generally believe that they can uncover repressed memories through dreams—memories often connected to catastrophic events experienced during childhood.
So a group of scientists decided to test how reliable this claim really was. They first recruited more than 100 participants and asked whether they had ever gotten lost in a public place or been abandoned by their parents as children. These are two childhood experiences that psychoanalysts often claim to interpret from dreams.
Then the researchers pulled a little trick. They deliberately selected participants who had never experienced either of these childhood traumas and sent them to a “dream interpretation” conference held by psychoanalysts. At the conference, the participants told the analysts about their dreams.
The result? No matter what the participants had actually dreamed about, the psychoanalysts said their dreams hinted at repressed memories of a difficult childhood—for example, getting lost in public or being abandoned by their parents.
Those psychoanalysts had, essentially, been publicly humiliated by the evidence.
The Instrument Connoisseur Tease
Now let’s look at the second variation, which I call the “Instrument Connoisseur Tease.”
There were once plenty of connoisseurs with deep expertise in musical instruments who swore that certain handmade violins passed down from antiquity had such beautiful tone that no violin made with modern technology could possibly compare.
Then someone ran an experiment: they blindfolded these instrument connoisseurs and had them listen to pieces played on the two kinds of violin. The violins they thought sounded better turned out, ironically, to be the ones made with modern technology.
If we borrow this “blind listening” method, we could shuffle and mix up the dreams of those 200 women, then hand them to psychoanalysts. We would tell them: “In these 200 cases, half are dreams dreamed by sexually repressed women, and the other half by sexually satisfied women. Please use the dreams to sort them into the two groups.”
My guess is that the results wouldn’t be much better.
The “Wine Taster Tease” and the “Instrument Connoisseur Tease” are two weapons that amount, essentially, to a method of misdirection and a method of blind testing. That’s all we need to say about them. If a theory is really pseudoscience, stuffed with all kinds of forced explanations, it is bound to be exposed when faced with these two kinds of teasing.
That’s enough on the critical examination of pseudoscience. Today’s material is actually a little deep. If anything I said gave you thoughts or raised questions, feel free to leave a comment and tell me.
Because of the limited space, I focused this episode on criticizing pseudoscience and didn’t spend much time directly discussing the science that stands opposite it. That part is covered in two episodes of our show called “The Thinking Methods Psychology Taught Me.” You can look them up; those two episodes are really a direct explanation of what science actually is.
That’s all for this episode. If you’d like to hear a contemporary scientific version of “dream interpretation,” you can listen to my paid program, “Zhichao Wei: New Developments in Psychology.” Sweet dreams tonight. See you next time!
A review of the “Freud’s dream interpretation theory is unreliable” research mentioned in the article:
Domhoff, G.W. (2000). ‘Moving Dream Theory Beyond Freud and Jung’. Paper presented to thesymposium Beyond Freud and Jung?, Graduate Theological Union, Berkeley, CA, 23September 2000.
The research mentioned in the article on “teasing” psychoanalysts:
Mazzoni, G.A., Lombardo, P., Malvagia, S., & Loftus, E. F. (1999). Dreaminterpretation and false beliefs. Professional Psychology: Research andPractice, 30(1), 45.
Leave a Reply