An open notebook with handwriting and a pen lying on a tree stump in a golden meadow
◎☽✦◈◇⊕
1951 to 2014 · Cleveland, Miami, Montreal, Mannheim

What People Really Dream About: The Counting, from Hall to the Typical Dreams

Everyone has an opinion about what dreams are like. A few people counted. Calvin Hall put ten thousand dreams on index cards in the 1940s, built a coding system with Robert Van de Castle in 1966 that is still in use, and the questionnaires that followed asked tens of thousands of people which dreams they had ever had. The answers are more ordinary, more shared and more unpleasant than the folklore, and they say nothing about what any one dream means.

In essence

Most dreams are about ordinary life, in familiar places among people the dreamer knows, with more misfortune and unpleasant feeling than the reverse, which is what Calvin Hall found in 10,000 dreams in 1951 and what the coding system he published with Robert Van de Castle in 1966 has kept finding since. Asked which dreams they have ever had, people in Canada, Germany, Japan and the United States put being chased, falling, school and sex at the top, and about six recent dreams in ten contain such a theme. The counting shows how shared dreams are, not what a shared dream means.

At a glance

  • Hall's 1951 article in Scientific American tabulated 10,000 dreams by setting, characters, plot, emotion and color, and found politics, economics and daily routines nearly absent; his collection grew to more than 50,000 reports and is searchable today at DreamBank.net.
  • The 1966 norms, 500 dreams of men and 500 of women, show women's dreams with more characters, more named emotions and more indoor settings, men's with more physical aggression, sex and work; later samples find the same direction, smaller, with far more overlap than gap.
  • Among 1,181 Canadian students, being chased, falling, school and sex had each been dreamed by more than 60 percent of both sexes (Nielsen and colleagues, 2003); in 2,853 most recent dreams, about 59 percent held at least one typical theme (Mathes, Schredl and Göritz, 2014).

Where the science stands

What is established: that dream content can be counted reliably, that it is patterned and mostly mundane, that the same themes recur across countries and decades, and that men and women differ on average in the ways the norms describe. What the counting cannot do is read a dream. A theme that most people share says that most people share the concern behind it, and nothing about why it came to you last night; Hall himself took the regularities as evidence against symbol dictionaries, not for them. Two honest limits on the numbers: the norms come from American college students of the 1940s and 1950s, so the averages describe a time and a place as much as a species, and a questionnaire that asks whether you have ever dreamed of falling measures memory and self-image as much as dreaming. The lens this site takes from the research is the modest one: your dream is probably more ordinary, and more widely shared, than it feels, and its meaning is still yours to find.

Read the research free

The data this page describes are unusually open:

  • DreamBankDomhoff and Schneider, University of California, Santa Cruz. More than 20,000 dream reports, Hall's collections among them, searchable by word and by series.
  • The Quantitative Study of DreamsThe Hall and Van de Castle coding rules, the norms, and Domhoff's papers in full text.
  • The Neurocognitive Theory of DreamingDomhoff, MIT Press, 2022, open access. Fifty years of the counting pulled into one account.

The books and papers this page draws on are listed in Sources and References.

1951

Ten thousand dreams on index cards

Calvin Hall was a psychologist at Western Reserve University in Cleveland who, in the 1940s, began asking students to write down their dreams and filing the reports on cards. By 1951 he had ten thousand, and in May that year Scientific American printed his account of them under the title What People Dream About. He had tabulated each dream by its setting, its cast of characters, its plot, its emotions and its coloring, and the picture that came out was not the one dream books painted. Dreams happened in houses, cars and streets more often than in strange places, among people the dreamer knew, and about the things the dreamer cared about. Politics, economics and the routines that take up most of a day were almost never there.

Hall kept collecting for the rest of his life, and by the time he died in 1985 the archive held more than 50,000 reports from children and adults, from the 1940s onward. That archive is why the question in this page's title can be answered at all: it is the one body of dreams large enough and old enough to count, and it is now public, at DreamBank, where anyone can search it.

1966

The coding system, and what the norms say

The counting became a method in 1966, when Hall and Robert Van de Castle, his collaborator at the Institute of Dream Research in Miami, published The Content Analysis of Dreams. The book is a manual: how to score a dream report for characters, social interactions, emotions, misfortunes and good fortunes, settings, objects and activities, so that two coders working apart reach the same numbers. With it came the norms, drawn from 500 dreams of men and 500 of women, five each from a hundred college students of each sex. Those norms are still the reference point against which new samples are compared, and G. William Domhoff, who inherited the work, gathered the rules, the tables and forty years of findings in Finding Meaning in Dreams in 1996.

What the norms say has held up well, and it is worth saying plainly. Most dreams contain the dreamer and two or three other people, more often known than strangers. Misfortunes outnumber good fortunes, and unpleasant emotions outnumber pleasant ones, so the ordinary dream is a little worse than the ordinary day. The men's and women's columns differ on average: women's dreams had more characters, more named emotions and more indoor settings, men's more physical aggression, more sex and more work. Later studies with the same coding system find the same direction of difference, smaller as the decades pass, and in every study the overlap between the sexes dwarfs the gap. Hall drew one conclusion from all of it that this site shares. Dream content is regular enough to count, and the regularity is the mark of continuity with waking concerns, not of a code.

The typical dreams

Which dreams nearly everyone has had

A second line of counting asks a different question. Instead of what a dream contained last night, it asks which dreams a person has ever had, from a list: being chased, falling, flying, being unable to find a toilet, teeth falling out, being late, failing an exam. The list is older than it looks. In 1958 Richard Griffith, Otoya Miyagi and Akira Tago gave it to Japanese and American students and published the result in American Anthropologist under the word universality, because the two lists came out so alike. The questionnaire in use today was built by Tore Nielsen and Antonio Zadra in Montreal, and in 2003 they gave it to 1,181 first-year students at three Canadian universities. Four themes had been dreamed by more than 60 percent of both men and women: being chased or pursued, falling, school and studying, and sexual experiences. Arriving late, failing an exam and seeing a living person dead were close behind.

Michael Schredl and colleagues in Mannheim found the same ranking in German students in 2004, stable across samples, with gender differences in a handful of themes. The weakness of the method is obvious, a questionnaire measures what people remember and are willing to tick, so in 2014 Jonas Mathes, Schredl and Anja Göritz asked 2,853 people online to describe only their most recent dream and coded the reports for the same themes. About 59 percent of those single dreams contained at least one typical theme, and the common ones were the familiar ones, being chased, flying, trying something again and again, and sex. The two methods agree on the shape of the list, which is the best reason to believe it.

Reading the numbers

What the counting can and cannot tell you

The counting settles three things. Dreams are mostly ordinary, which cuts against the idea that a dream is a message in cipher. Dreams are shared, so a dream of being chased or of losing your teeth puts you with most of the people who have ever been asked, and the research on recurring dreams finds the same few themes returning. And dreams track waking life, which is the finding the continuity hypothesis is built on and this page's numbers support.

It leaves one thing exactly where it was. A theme that most people share says that most people share the concern behind it, that being pursued, losing footing, being tested and wanting someone are the common stuff of a human life. It does not say what the chase meant to you, and the researchers who did the counting are the last people to claim it does. Two further limits are worth keeping in view. The norms describe American students of the 1940s and 1950s, so some of what they show is a time and a place; later samples from other countries match the broad pattern and differ in the details. And a questionnaire about dreams you have ever had is also a questionnaire about what you remember and how you see yourself, which is why the single most recent dream study matters: it agrees.

Our lens

What we take from the counting

The research on this page is the reason this site refuses to be a dream dictionary. If being chased is in most people's dreams, then a table that gives it one meaning is giving most of humanity one meaning, and Hall's own reading of his cards was that the regularity speaks for continuity with waking concerns, not for a code. So the reading offered here begins with your associations, treats the shared themes as common ground and not as a key, and says plainly when science has no answer.

The counting does have one use for a single dreamer, and it is the one Hall would have recognised. A dream journal kept over months is a content analysis of one person, and what returns in it, the motifs, the people, the mood, is the pattern worth attending to. Companion members see those motifs counted across their own dreams, which is the Hall and Van de Castle method applied to a sample of one, and the only sample in which the pattern can mean something to the person it belongs to.

Questions

Frequently Asked Questions

What do people dream about most?

Ordinary life, with the volume turned up. When Calvin Hall tabulated 10,000 dreams in 1951 and then built a coding system with Robert Van de Castle in 1966, most dreams turned out to happen in familiar settings with people the dreamer knew, and misfortunes and unpleasant emotions outnumbered good fortune and pleasant ones. Politics, economics and the routines that fill a day were nearly absent. When people are instead asked which themes they have ever dreamed, being chased, falling, school and studying, and sex come out on top in almost every country asked.

What are the most common dream themes?

In the largest questionnaire study, 1,181 first-year students at three Canadian universities (Nielsen and colleagues, 2003), four themes had been dreamed by more than 60 percent of both men and women: being chased or pursued, falling, school and studying, and sexual experiences. Arriving late, failing an exam and a person now alive being dead were close behind. Japanese and American students gave much the same list in 1958, and German samples give it today. When 2,853 people were asked only about their most recent dream, about 59 percent of those dreams contained at least one typical theme.

Do men and women dream differently?

On average, and only on average. In the Hall and Van de Castle norms from the 1950s and 1960s, women's dreams had more characters, more named emotions and more indoor settings, and men's had more physical aggression, more sex and more work. Later studies with the same coding system find the same direction of difference, smaller in recent decades, and the overlap between the sexes is far larger than the gap. None of it predicts what any one woman or man will dream tonight.

Who were Hall and Van de Castle?

Calvin S. Hall was an American psychologist who began collecting dream reports from students in the 1940s, wrote What People Dream About for Scientific American in 1951 from 10,000 of them, and kept collecting for the rest of his life. Robert Van de Castle was his collaborator at the Institute of Dream Research in Miami. The Content Analysis of Dreams (1966) is their coding manual, with norms from 500 dreams of men and 500 of women, and it is still the standard way of counting what dreams contain. G. William Domhoff carried the work on and put Hall's collections online as DreamBank.

Does a common dream have a common meaning?

No, and the counting is the best argument against it. Content analysis shows that dream themes are widely shared and statistically regular, which is exactly what you would expect if dreams drew on the concerns most people have, being pursued, losing footing, being tested, wanting someone. It does not show that being chased means the same thing to two dreamers. Hall himself read the regularities as evidence against fixed symbol dictionaries, and the dreamer's own associations remain the only way to the meaning of a particular dream.

Where can I see the dream research data for free?

DreamBank.net, built by G. William Domhoff and Adam Schneider, holds more than 20,000 dream reports from Hall's collections onward and can be searched by anyone. The Hall and Van de Castle norms and coding rules are published at dreams.ucsc.edu, the University of California, Santa Cruz site that also carries Domhoff's papers. Domhoff's 2022 book The Neurocognitive Theory of Dreaming, which gathers fifty years of this work, is open access from MIT Press.

Recommended reading

Go Deeper

The Content Analysis of DreamsCalvin S. Hall and Robert L. Van de Castle (1966)

Appleton-Century-Crofts. The coding manual and the norms, 500 dreams of men and 500 of women, that every count of dream content since has been compared with.

View in Sources ↗
Finding Meaning in Dreams: A Quantitative ApproachG. William Domhoff (1996)

Plenum. The coding rules, the normative tables and forty years of findings in one volume, by the man who inherited Hall's archive.

View in Sources ↗
The Typical Dreams of Canadian University StudentsNielsen, Zadra and colleagues (2003)

Dreaming. Eleven hundred and eighty-one students and the questionnaire that became the standard: being chased, falling, school and sex above 60 percent in both sexes.

View in Sources ↗
Frequency of Typical Dream Themes in Most Recent Dreams: An Online StudyMathes, Schredl and Göritz (2014)

Dreaming. The check on the questionnaires: 2,853 single most recent dreams, about 59 percent of them with at least one typical theme.

View in Sources ↗

One dream is a night.

Kept together, they begin to show a thread. The Dream Journal letter arrives on the 1st and the 15th: one symbol read closely, one honest idea from the science of sleep.

Free · Twice a month max · One-click unsubscribe

And if you want to keep the dreams themselves: the Companion plan keeps what you cannot see alone.