The 1948 classroom
Bertram Forer gave thirty-nine of his students a personality test, took the papers away, and came back with what each of them was told was an individual profile. He asked them to rate how well it described them, from 0 to 5.
The average was 4.26. More than a third scored it a perfect 5.
Every student had been handed the same thirteen sentences, which Forer had assembled the night before out of a newsstand astrology book. He published it the following year as “The fallacy of personal validation: a classroom demonstration of gullibility”, and the finding has been reproduced continuously ever since. Paul Meehl later attached the showman's name to it, and it has been the Barnum effect since.
You have a great need for other people to like and admire you. You have a tendency to be critical of yourself. At times you are extroverted, affable and sociable, while at other times you are introverted, wary and reserved.
Those are three of Forer's originals. Read them again and notice what they are doing.
The four moves
A statement that almost everyone accepts as personal is built, not stumbled on. There are four recurring devices and once you can see them you cannot unsee them.
- Double-headedness. “At times extroverted, at other times reserved” cannot be wrong, because it has already claimed both. Any evidence you produce confirms it.
- The flattering deficit. “You have a great deal of unused capacity which you have not turned to your advantage.” It names a flaw that is really a compliment, so accepting it costs nothing.
- The universal secret. “You have found it unwise to be too frank in revealing yourself to others.” True of everyone, and it feels like an intimate observation because most people have never heard it said aloud.
- Texture instead of fact. A good line describes the feel of a situation and never a checkable detail. “There is a conversation you have both been circling” is unfalsifiable. “You argued on a Tuesday” is not, and would be wrong four fifths of the time.
What makes acceptance go up
Replications since Forer have pinned down what strengthens the effect, and none of it is about the accuracy of the text.
- Believing it was made for you. The identical paragraph rates higher when the reader thinks it was generated from their own birth date, handwriting or test answers than when they are told it is a general description.
- The authority of the source. A profile attributed to an expert or an instrument beats the same words attributed to nobody.
- Favourability. Positive statements are accepted far more readily than negative ones. A reading that says you are secretly generous will land; one that says you are secretly petty will be argued with.
Notice that a birth-date tool satisfies all three by construction. That is worth knowing before you decide how much weight to give one.
Cold reading is this, performed
Ray Hyman's 1977 essay on cold reading described the working method: open wide, watch the person, and narrow. A statement is offered vaguely, the listener's face or correction supplies the specifics, and the reader folds those specifics back in as though they had been there all along. The listener remembers the hit and forgets that they provided it.
The technique needs no dishonesty to work. A sincere reader using a symbolic system does the same thing without meaning to, because a conversation naturally rewards the parts that landed and drops the parts that did not.
What this site does about it
It says so, here, and it makes the design choices that follow.
The two-person reading gives every pairing an ease and an edge, because a reading composed only of flattering deficits is a Barnum machine. It refuses to issue a verdict. It names a recurring situation rather than a date, and it says out loud that a low score is not a worse relationship.
It is also stable: the same two dates always return the same reading, and swapping which birthday went first changes nothing but the names. That matters here. A reading that changes on refresh cannot be checked, and anything that cannot be checked can only be believed.
None of that makes the reading true. It makes it honest about what kind of object it is: a set of prompts drawn from a two-thousand-year-old symbolic vocabulary, which is useful in exactly the way a good question is useful.
The test you can run
Take any reading this site gives you and hand it to somebody it was not calculated for. Ask them how well it fits. If it fits them too, you have learned something real — about the reading, and about how to read the next one.