Are Dreams Social Simulations, or Are They Enactments of Conceptions and Personal Concerns? An Empirical and Theoretical Comparison of Two Dream Theories

G. William Domhoff & Adam Schneider

University of California, Santa Cruz

NOTE: If you use this paper in research, please use the following citation, as this on-line version is simply a reprint of the original article:
Domhoff, G. W., & Schneider, A. (2018). Are dreams social simulations? Or are they enactments of conceptions and personal concerns? An empirical and theoretical comparison of two dream theories. Dreaming, 28, 1-23. doi:10.1037/drm0000080


This article compares social simulation theory, which claims that dreaming is a rehearsal for waking social perceptions and interactions, and therefore has adaptive value (Revonsuo, Tuominen, & Valli, 2015), with a neurocognitive theory of dreaming, which claims dreaming is an intensified form of mind-wandering that makes use of embodied simulation, primarily to enact the dreamer's major conceptions and personal concerns, but has no adaptive value (Domhoff, 2011). The article presents new findings on types of embodied simulations in dream reports, 6.5% of which are not social simulations according to social simulation theorists. The article concludes with a critique of social simulation theory and 3 other social rehearsal theories based on (a) new results presented in this article, including findings on dreams with no characters other than the dreamer; (b) the replicated finding that social interactions with deceased loved ones are repeated consistently in dream series kept for years and decades, which does not fit with the idea of rehearsal for the future; (c) the cognitive-developmental evidence that dreaming does not reach the frequency of adult dreaming until ages 9-11, and is not fully like that of adults until early adolescence, which does not fit with the claim that dreaming has adaptive value for young children as a rehearsal of perceptual and interaction skills; (d) the theory's questionable underlying assumptions concerning "implicit learning" during dreaming and "transfer of learning" to waking-life situations; and (e) the wide-ranging evidence that does not fit with the claim that dreaming has adaptive value.

The purpose of this article is to compare two relatively new theories of dreaming. The first, social simulation theory, claims that dreaming is first and foremost a form of social rehearsal for waking social interactions that was conserved in human evolutionary history because it had adaptive value in the ancestral environment (Revonsuo, Tuominen, & Valli, 2015). The second, an updated version of an earlier neurocognitive theory of dreaming, claims that dreaming is an intensified form of mind-wandering, which makes use of embodied simulation to enact the dreamer's major conceptions and personal concerns in approximately 70%-80% of dream reports (Domhoff, 2018, pp. 64-67, for studies of the nonfamiliar and nonpersonal in six samples of dream reports). According to this theory, dreaming occurs when an augmented portion of the brain's default network is highly activated at the same time that external stimuli are occluded, but dreaming is a nonadaptation, a byproduct of human cognitive developments important in waking life, and has no adaptive value (Domhoff, 2018, Chapter 8).

We will begin with a statement of the main tenets of social simulation theory, along with the results from an empirical study by social simulation theorists, which provides a test of their hypothesis. We will then present an overview of the neurocognitive theory and introduce five new categories for studying embodied simulation in dreams. These categories are based on normative findings for characters, social interactions, and activities in dream reports using the Hall and Van de Castle (1966) (hereafter HVdC) coding system.

After showing that the new results with the HVdC coding system encompass all the findings related to social simulation theory with great precision and detail, and also add findings on two other types of embodied simulations as well, we will turn to a critique of social simulation theory that is based on (a) the new results presented in this article; (b) the replicated findings from numerous lengthy dream series contributed by a wide range of individuals; (c) the cognitive-developmental findings that do not fit with the claim that dreaming has an adaptive function, at least until preadolescence; (d) a discussion of two questionable assumptions upon which the theory is based; and (e) other types of evidence from dream studies based on adult samples that does not fit with the idea that dreaming has an adaptive function. We shall then conclude with the suggestion that dreaming is the accidental byproduct of relatively recent evolutionary changes in the brain.

Social Simulation Theory

According to the most general formulation of social simulation theory, dreaming is a form of social rehearsal and has an adaptive function. More specifically, dreams are selectively specialized in simulating social perceptions and social interactions in a way that enhanced the survival and reproductive success of ancestral human beings (Revonsuo et al., 2015, pp. 15-16). This theory parallels and possibly complements the threat simulation theory developed by two of the same researchers, which hypothesizes that interactions with threatening people, animals, and situations are differentially simulated in dreams and have had, at least in the distant past, an adaptive function as rehearsals for waking-life situations (Revonsuo, 2000; Revonsuo & Valli, 2000). The threat simulation theory has been analyzed separately and will be only tangentially mentioned in this article (Domhoff, 2018, pp. 258-268).

In social simulation theory, the focus is on social perception and on interactions of a positive nature, such as friendly and supportive interactions (Revonsuo et al., 2015). However, the coding system the social simulation theorists developed to study their theory perforce includes neutral interactions, such as seeing or talking with someone, as well as categories for coding negative interactions, the most salient of which are physical and nonphysical aggressions of various kinds (Tuominen, Stenberg, Revonsuo, & Valli, 2018).

In the social simulation researchers' coding system, "The minimal criterion for a dream to count as a social simulation is that the Dream Self is not alone in the dream," and the "presence of simulated human characters must be perceptually detected and registered in the dream by the dreamer" (Revonsuo et al., 2015, p. 7). Once a human character is detected and registered, there is a "more sophisticated form" of recognition "in terms of some basic perceptual and semantic categories (male/female; familiar/stranger), and then in terms of more detailed semantic and autobiographical information about the precise identity and name of the person" (Revonsuo et al., 2015, p. 7). Recognition is sometimes followed by a neutral, positive, or negative social interaction. Although the new coding scale includes negative interactions, it does not include interactions with animals, which are frequently threatening in dreams. In that regard, it does not encompass a type of interaction that was included in the threat simulation scale (Revonsuo & Valli, 2000). This is because the social simulation theory is focused on human social interactions and the threat simulation theory is focused on anything that could be considered a threat.

The social simulation theorists' data analysis using their new coding system was carried out on carefully drawn matched samples of 115 REM and 116 non-REM dream reports from a large database. Other investigators collected these reports from 16 participants living in university dormitory rooms. The dream reports were gathered by means of automated awakenings several times each night over a 2-week period. A portable electronic device that monitors eye-movement activity and head movements initiated the awakenings. The device arouses participants during REM periods when there are eye movements and during non-REM when there is an absence of eye movements and head movements (Ajilore, Stickgold, Rittenhouse, & Hobson, 1995; Stickgold, Malia, Fosse, Propper, & Hobson, 2001, p. 172). This study led to the collection of 229 REM reports and 165 non-REM reports, along with 244 sleep-onset reports, the latter of which were not analyzed by the social simulation theorists (Stickgold et al., 2001, p. 13, Table 2). In addition, the study included waking thought samples from the same participants, who were contacted at random times throughout the day by means of a pager. The social simulation researchers' analysis of the waking reports is not relevant to the purposes of this article, and will not be discussed.

The interrater agreement was very high: 92.9%-97.3% across the different categories using the percentage-of-agreement method for determining reliability. The social simulation researchers first of all found that the REM and non-REM dream reports were similar for their various categories, and that the time of night at which the report was collected did not make any difference. Overall, 83.5% of the dream reports had at least one perceptual or social interaction, and the dreamer was part of 84.8% of the social interactions (i.e., not simply an observer). Most social interactions (70.3%) were carried out with familiar characters. The over- whelming majority of the social interactions (67.2%) were "neutral" (neither positive nor negative), and positive interactions were higher than negative interactions (21. 6% vs. 11.0%; Tuominen et al., 2018, Table 3). The most frequent negative interactions were abandonments and verbal criticisms. Strikingly, only 7.6% of the social simulations were reciprocal ones in which both characters made utterances or carried out actions that were either altruistic/friendly or hostile/aggressive, or a combination of both. Another 30.7% were "one-way" interactions, and the other 61.7% were more in the realm of shared or joint interactions/activities (Tuominen et al., 2018).

Based on these findings, the authors conclude that the study provides solid evidence for the importance of social simulation in dreams, although they also note limitations and suggest that further studies are needed.

The Neurocognitive Theory

The neurocognitive theory of dreaming builds on a cognitive theory of dreaming first proposed in the early 1950s. It begins with the idea that dreams are the "embodiment of thoughts," and notes that "scenes, people, objects, and actions are experienced as though they were impressing themselves on the senses from the external world" (Hall, 1953, p. 273, our italics). ("Experienced as" and "real" were two of the several ways that the realistic feeling that accompanies dreaming was described by virtually all dream researchers before the concept of "simulation" was suggested by a laboratory dream researcher to better capture the highly imaginative and realistic nature of dreaming; Foulkes, 1982a, 1985.)

According to the early cognitive theory, the embodied thoughts called dreams express people's "conceptions" of the world and the people around them. For example, do the settings, activities, and social interactions portray the dreamer as favoring the outdoors or indoors, as aggressive or friendly, as a victim of other people's criticisms and taunts, or perhaps as without any friends? Although the theory starts abstractly by talking in terms of conceptions of the world, the primary focus is on the dreamers' personal social worlds. This is due to the results of an empirical study that discovered dreams only occasionally include mentions of worldly concerns, such as politics, economics, or the world of work (Hall, 1951). This result has been replicated many times in both sleep-and-dream lab studies and nonlab studies (Domhoff, 2018, Chapters 1-2; Dorus, Dorus, & Rechtschaffen, 1971; Foulkes, 1985; Snyder, 1970; Strauch & Meier, 1996).

As a result of detailed studies of a handful of lengthy personal dream journals, which are a form of unobtrusive archival data long used by research psychologists (Allport, 1942; Baldwin, 1942; Webb, Campbell, Schwartz, Sechrest, & Grove, 1981), the theory was updated to state that dreams express personal "concerns" as well as personal conceptions (Bell & Hall, 1971; Domhoff, 1996, Chapters 7-8; Hall, Domhoff, Blick, & Weesner, 1982, p. 193). More exactly, as many as 70%-80% of dreams dramatize personal conceptions and concerns, whether they express lingering concerns about the past (e.g., deceased loved ones), present-day concerns (e.g., conflicts with parents or siblings and the enjoyment of favorite avocations), or thoughts about the future (e.g., worries about an upcoming examination or the birth of a baby; Domhoff, 2018, Chapters 2-3).

The theory was later extended into the neurocognitive realm on the basis of three surprising discoveries (Domhoff, 2001). First, longitudinal and cross-sectional developmental studies of young children in sleep laboratories unexpectedly revealed that dreaming is a gradual cognitive achievement that is not often present or well formed until age 5, not as frequent or as complex as adults' dreams until ages 9-11, and not expressive of personal concerns or many negative emotions until ages 11-13 (Foulkes, 1982b; Foulkes, Hollifield, Sullivan, Bradley, & Terry, 1990). Second, neuroimaging studies of sleep in the mid-1990s discovered that the brain is differentially activated during REM (Maquet et al., 1996; Nofzinger, Mintun, Wiseman, Kupfer, & Moore, 1997; Nofzinger et al., 1999). Third, a lesion study of 361 consecutive new neurological patients revealed that injuries to the same brain areas activated during REM, and only to those areas, led to alterations in dreaming or the complete loss of it, which added crucial subjective, first-person testimony to the neuroimaging findings (Solms, 1997). The lesion study also called attention to earlier laboratory studies of patients whose dreaming had been lost or altered, which added further credence to the neuroimaging and neurological evidence (Jus, Jus, Gautier, et al., 1973; Jus, Jus, Villeneuve, et al., 1973; Kerr & Foulkes, 1981; Kerr, Foulkes, & Jurkovic, 1978).

Several years later, the neurocognitive theory of dreams was amended on the basis of the serendipitous discovery of a waking brain network (the default network) that supports mind-wandering (Andrews-Hanna, Reidler, Huang, & Buckner, 2010; Buckner, Andrews-Hanna, & Schacter, 2008; Mason et al., 2007; Raichle et al., 2001). This breakthrough finding soon led dream researchers to the realization that the default network is similar to the REM network (Domhoff, 2011; Hartmann, 2011; Pace-Schott, 2011). The neurocognitive theory henceforth claimed, based in good part on earlier evidence of dreaming during relaxed wakefulness, collected in a lab setting using the electroencephalogram (EEG) to monitor wakefulness, that dreaming is an intensified and enhanced form of mind-wandering (Foulkes & Fleisher, 1975; Foulkes & Scott, 1973, two pioneering studies of brief episodes of dreaming during extended mind-wandering). The step from daydreaming to dreaming is supported by an augmented portion of the default network (Domhoff, 2011, Domhoff, 2018, Chapter 5; Domhoff & Fox, 2015; Fox, Nijeboer, Solomonova, Domhoff, & Christoff, 2013).

New Simulation Categories and Content Indicators

Our approach to studying embodied simulations in the context of the neurocognitive theory builds upon the established HVdC coding system, which classifies all of the elements that appear in dream reports. This system rests on the nominal (categorical) level of measurement and contains 10 general empirical categories and numerous subcategories. It has been found to be adaptable and comprehensive as well as transportable from one research setting to another (Domhoff, 2003, Chapter 3; Van de Castle, 1969). The comprehensiveness of the system was first demonstrated by the fact that the categories for victimization, misfortune, and failure detected exactly the same type of content as did Beck and Hurvich's (1959) misnamed "masochism" scale, and did so in a more accurate fashion as well (Clark, Trinder, Kramer, Roth, & Day, 1972).

To study the topic of embodied simulations in dream reports, we created five new categories that were used to analyze the established normative findings with the HVdC coding system for characters, social interactions, and activities. The deployment of these new simulation categories has three potential advantages over any de novo system. First, the high reliability of each of the HVdC coding categories has been previously demonstrated by Hall and Van de Castle's original codings, as well as by the reports on reliability by others who have used one or more of these scales (Domhoff, 1996; Hall & Van de Castle, 1966, Chapter 13; Tonay, 1990/1991).

Second, the HVdC system has the advantage that norms for the new categories can be easily derived from existing data sets. In this case, new normative figures were calculated using the spreadsheet and search programs constructed by the second author of this article on the basis of the independent codings by Hall and Van de Castle of 500 men's and 500 women's dream reports (Hall & Van de Castle, 1966, Chapter 14). (Unfortunately, nine of the original coding cards for women were lost, as Hall and Van de Castle sent them back and forth between Santa Cruz and Charlottesville for subsequent studies after they finished their common research effort in Miami. Luckily, the normative findings with the remaining 491 reports are similar to the published findings for the complete set of 500 reports.) The original normative findings were subsequently replicated several times for selected coding categories or for all the categories that are frequently used (Dudley & Fungaroli, 1987; Dudley & Swank, 1990; Hall et al., 1982; Reichers, Kramer, & Trinder, 1970; Tonay, 1990/1991). Similar findings were reported in several other countries, with the largest differences relating to various aggression categories (Dale, Lortie- Lussier, Wong, & De Koninck, 2016; Mazandarani, Aguilar-Vafaie, & Domhoff, 2013; Prasad, 1982; Yamanaka, Morita, & Matsumoto, 1982).

There is one further possible advantage to using the HVdC categories in conjunction with the normative findings. Due to the fact that the dream reports were coded for every type of element that appears in dreams — including good fortunes, misfortunes, success, failure, and five categories of emotions — it is possible to determine if there are differences between any of the five new embodied simulation categories and the normative samples in terms of the pattern of accompanying elements.

Aggression, Friendliness, Sexuality

The first of the five new categories encompasses social interactions involving aggression , friendliness , or sexuality (A/F/S) between two or more human characters. These three types of social interactions have been used in many dozens of investigations relating to age, gender, culture, and individual differences by a wide range of researchers in at least 11 countries (as summarized in Domhoff, 1996, Chapters 4-6; 2018, Chapters 2-4). In the combined HVdC normative findings used in this study, 45.4% of the dream reports have at least one aggression interaction, 39.7% have at least one friendly interaction, and 7.6% have at least one sexual interaction. (These percentages are slightly lower than those presented in previous studies using the HVdC norms because interactions with animals have been omitted for purposes of this article, which makes its finding comparable with those from the study by social simulation researchers.) In this study, these three types of social interactions are combined to create a content indicator that expresses the percentage of dream reports that have "at least one A/F/S interaction" (involving humans).

Directed and Joint Activities

Second, social interactions often occur in the context of activities in dream reports, such as talking, walking, looking, listening, and thinking. There are eight subcategories of the overall activities category, all of which focus on "what characters do in dreams" (Hall & Van de Castle, 1966, p. 87, italics in the original). The activity subcategories include three that involve the body: physical (voluntary movements of the body or body parts while staying in about the same place), movement (self-propelled changes in location), and location change (changes in locale via a vehicle). The activities category also includes five sensory and cognitive categories that are for the most part self-explanatory on the basis of their labels: verbal , visual , auditory , expressive communication (primarily laughing and crying), and thinking/cognition (Domhoff, 1996, pp. 236-242; Hall & Van de Castle, 1966, Chapter 7). (To read the coding rules and coding examples online, visit the section on "HVdC coding rules" on the "Resources for Scientists" page on DreamResearch.net.) The reliability of these categories is demonstrated by a correlation of .92 between Hall and Van de Castle's separate codings for physical activities, of .98 for their codings of verbal activities, and of .92 for movement activities (Hall & Van de Castle, 1966, p. 154, Table 13-8). In addition, the percentage of agreement for "the type of activity class, as well as the characters involved, was 85% between the two judges" (Hall & Van de Castle, 1966, p. 155).

Overall, activities are frequent in the HVdC normative dream reports, averaging 4.9 per dream report for women and 4.7 for men; 25% of women's dream reports and 22% of men's reports contain from six to 20 activities, which means that a large amount of time is needed to code for all the activity categories (Hall & Van de Castle, 1966, pp. 182-185, for the highly detailed findings on activities). Due to the resource constraints faced by most investigators, these categories have not been used often and therefore are not well known, although they have been of value in some investigations (Domhoff, 1996, pp. 112-113, 165-166, for summaries; Hall & Lind, 1970, pp. 30-31, 45-48; Yamanaka et al., 1982). To provide an idea of the relative frequencies of various types of activities in dream reports, Table 1 displays the percentage of dream reports with at least one instance of each of the eight subcategories. But it is important to emphasize that not all of the instances counted in this table are necessarily part of our new indicator, as explained in the next paragraph.

Table 1. Percentage of Dreams With at Least One Instance of the Eight Activity Subcategories for the Combined, Male, and Female Hall and Van de Castle Normative Codings
Combined norms
Male norms
Female norms
Movement (M)66.3%65.4%67.2%
Verbal (V)61.6%58.2%65.0%
Physical (P)60.5%64.2%56.8%
Sight/visual (S)44.2%41.8%46.2%
Location (L)28.8%29.8%27.7%
Cognitive (C)17.4%13.8%21.0%
Expressive (E)11.5%9.0%14.1%
Auditory (A)6.6%6.8%6.3%

Activities are sometimes carried out in "interactions between characters" and in "conjunction with other characters" (Hall & Van de Castle, 1966, p. 87). It is the characters' interactions with each other (which we call "directed" activities) or in conjunction with each other (which we call "joint" activities) that are defined as a second type of social simulation in dream reports. Directed and joint activities may or may not involve the dreamer, and sometimes, they simply involve a group of unnamed people carrying out an activity together.

These directed and joint activities include characters talking with, or moving to another locale with, another character or characters, or two or more characters seeing or listening to someone together. They also can include such joint activities as lifting an object together or working to repair a vehicle together. There are even a few instances of characters laughing or crying together, or thinking together, as in "deciding" or "plotting" some activity together. This type of social simulation is analyzed by means of a content indicator expressed as the percentage of dream reports with "at least one directed/joint activity" involving human characters.


In addition to A/F/S interactions and directed/joint activities, there is a type of social simulation that includes only visual, auditory, or thinking activities by the dreamer in which another character may be seen, heard, or thought about (e.g., "I saw/heard Joe across the room" or "There were hundreds of people along a fence and I thought of a woman I know"). This category also includes dream reports in which one or more human characters are reported to be in the same setting with the dreamer but are not interacting with the dreamer or each other (e.g., "I was standing in a railroad station where there was a great crowd of people"). These dream reports are labeled "perception/cognition" simulations because they only include the dreamer's perceptions and/or thoughts in relation to some other character or characters. By definition, perception/cognition dream reports do not include any type of social interaction or any directed or joint activities. In that sense, they are a "residual" category. Dream reports that fit within the perception/cognition category are analyzed by means of a content indicator expressed as the percentage of dream reports with "at least one perception/cognition activity" involving human characters.


A fourth type of simulation in dreams is one in which only the dreamer and one or more animals or creatures are present. (It should be noted that "creatures" — the stuff of science fiction, fairy tales, and cartoons — are rare in the normative samples and appear only twice in the combined norms.) The dreamer may or may not be involved in an A/F/S social interaction with an animal or creature, and there may or may not be an activity coding. In fact, however, dreamers regularly have aggressive and friendly social interactions with animals, such as the dreamer being chased by a wild animal or playing with a pet. On the other hand, few dream reports contain a directed or joint activity with an animal-but it is at least plausible to imagine a movement coding (e.g., running along with a dog) or a location change (e.g., taking a cat to the veterinarian). Still, the defining feature of this category is the absence of human characters other than the dreamer, so results are expressed with a content indicator named the "dreamer-plus-animal-or-creature percent."


Fifth, and finally, there is a type of simulation in dreams in which the dreamer carries out one or more activities but, by definition, can have no social interactions, directed activities, joint activities, or social/perception activities because there are no other characters in the dream report. Nor is there even any thinking of other characters in what are labeled as "dreamer-only" dream reports. In these dreams, the person is carrying out some activity alone, such as enjoying a new electronics gadget, shaving, driving, or trying to fix an automobile, or else walking alone on a street or in a forest. The findings in this category are expressed as the "dreamer-only percent."

The dreamer-plus-animal-or-creature and dreamer-only categories are substantively interesting as potential differentiating points between the social simulation and embodied simulation theories because they are not social simulations according to the coding rules for the scales created by the social simulation theorists. More exactly, such dreams may provide support for the idea that dreams differentially enact and dramatize conceptions and personal concerns in general, not simply in the context of social simulations (as social simulation researchers define that term). These two categories include singular types of dreams to which not enough theoretical attention has been paid in the past.


The results with the new simulation categories are derived from analyses of the HVdC men's norms, women's norms, and combined men and women's norms. They are reported in Tables 2 and 3, which are based on the content indicators that present the percentage of dream reports with at least one instance of a category. Table 2 displays the percentage of dream reports with at least one of the following: (a) an A/F/S interaction or (b) a directed/joint activity. Column 4 of Table 2 displays the cumulative results when A/F/S interactions are combined with directed/joint activities to determine the percentage of dream reports with at least one of either type of social simulation. The table shows, among several things, that (a) 67.2% of the 991 men's and women's dreams combined have at least one A/F/S interaction, (b) 77.1% have at least one directed or joint activity, and (c) 86.9% have at least one A/F/S interaction or at least one directed/joint activity.

Table 2. Percentage of Dream Reports With at Least One Aggression, Friendliness, or Sexuality (A/F/S) Social Interaction, at Least One Directed or Joint Activity, and at Least One of Either of These Two Types of Social Simulations
A/F/S interactionDirected/joint activityA/F/S interaction or directed/joint activity
Combined norms (n=991)
Female norms (n=491)
Male norms (n=500)

In addition, Table 2 provides the results for the women's and men's dream reports separately, which reveals that women have a slightly lower percentage of dreams with at least one A/F/S interaction (65.6% for women and 68.8% for men) and a slightly higher percentage of dreams with a directed or joint activity (79.2% for women and 75.0% for men). However, both women and men end up with approximately the same percentage of dream reports with at least one of these two indicators combined (87.4% for women and 86.4% for men). Table 2 also presents the percentages for dreamer-involved simulations and for simulations in which the dreamer is not involved. The findings demonstrate that a large majority of these simulations involve the dreamer and that the differences between women and men on dreamer involvement are within a few percentage points of each other.

Table 3. Percentage of Dream Reports With at Least One Aggression, Friendliness, or Sexuality (A/F/S) Interaction or Directed/Joint Activity, With at Least One Perception/Cognition, With Only Dreamer and Animals/Creatures, and with Dreamer Only
 A/F/S + directed/joint activitiesPerception/cognition of human charactersDreamer and animals/creaturesDreamer only
Combined norms (n=991)86.9%6.7%2.2%4.3%
Female norms (n=491)87.4%7.5%2.2%3.1%
Male norms (n=500)86.4%5.9%2.2%5.6%

Turning to Table 3, it first of all reproduces the summary findings on the combined female and male norms that appeared in the rightmost column of Table 2 (A/F/S or directed/joint activities) for the sake of easy comparisons with the next three columns. Table 3 then shows that the perception/cognition percent is 6.7% in the norms for the 991 men's and women's dream reports combined. This means that 93.6% of dream reports (86.9% + 6.7%) include a social simulation, as social simulation theorists define that concept. The remaining dream reports are dreamer-plus-animal-or-creature reports (2.2%) and dreamer-only dreams (4.3%).

The two categories listed in the last two columns in Table 3 (dreamer-plusanimal- or-creature and dreamer-only categories), which account for 6.5% of the dream reports in the combined normative sample, are distinctive in terms of the degree to which the dream content within each of them does or does not contain aggressions, apprehension (fear), or misfortunes, as compared with the HVdC norms for men and women combined.

The dreamer-plus-animal-or-creature dream reports are very high on aggressive interactions, due to the dreamers being chased or physically attacked by an animal or dangerous insect; the physical aggression percent is 73.7%, as compared with 42.9% for the combined male and female norms (h=.64, p<.001). In addition, the percentage of dream reports with at least one apprehension emotion is very high (62.8% vs. 21.8%, h=.99, p<.01). A slight majority of these apprehensions (eight of 15) occur in dreams without an aggression by an animal, which suggests that dreamers tend to be apprehensive in the mere presence of animals for various reasons. The dreamer-only dreams, on the other hand, show another distinctive pattern: They are higher than the normative baselines for the presence of at least one misfortune (62.8% vs. 34.8%; h=.57, p<.001). However, the auxiliary findings (more frequent aggression and apprehension in dreams with only animals as the other characters, and more frequent misfortunes in the case of dreamer-only dreams) have to be treated with caution. Even though the p values are very low, and the effect sizes are large, for these auxiliary findings, they need to be replicated with larger sample sizes because there were relatively few instances of all three elements. Although some researchers may be convinced by large statistical differences, we remain proponents of both large sample sizes and replications (Domhoff, 1996, Appendix D; Domhoff & Schneider, 2008, Domhoff & Schneider, 2015).


This article first of all demonstrates that it is possible for the HVdC coding system to account for social interactions, directed and joint activities, and perceptions and cognitions at least as well as the social simulation theorists' scales do, by making use of some of HVdC's original categories to create several new simulation categories. Specifically, the new indicators that relate to human social interactions account for 93.6% of the dream reports in the combined male and female norms, whereas the scale created by the social simulation theorists accounts for 83.5% of the reports in the sample that they studied. (As will be suggested when the two samples are compared shortly, this difference is likely due to differences in the samples.) In addition, the two coding systems agree that dreams with only perceptions/cognitions or social perceptions are a small percentage of all dreams (6.7% for the HVdC categories and 12.6% for the social simulation researchers' categories). The two coding systems further concur that the dreamer is involved in the great majority of all interactions. Finally, the two coding systems agree, based on HVdC normative findings not previously mentioned in this article, that there are few reciprocal social interactions in dreams. These results are in many ways those that might be expected because there is considerable overlap in how our HVdC social simulation categories and those of the social simulation theorists are defined in terms of characters, social perceptions, social interactions, and social activities involving two or more characters.

However, there is also one disagreement. The social simulation theorists found that the overwhelming majority of the interactions were neutral and that positive interactions predominated over negative ones (21.6% vs. 11.0%). In most HVdC studies, including those based on dream reports collected by anthropologists in small-scale societies, aggressions usually outnumber friendly interactions, and dreamers are more often victims than aggressors in aggressive interactions (Domhoff, 1996, Chapters 4, 6, 7). In a study of the HVdC normative codings in which "negative elements" were defined in terms of aggressions, misfortunes, failures, and negative emotions, 80% of men's dreams and 77% of women's dreams had at least one of these negative elements. In sharp contrast, only 53% of dreams for both men and women had at least one of several positive elements, including friendly interactions, good fortune, success, and positive emotions (Domhoff, 2007, p. 13).

There are two possible reasons for the differences in findings on some issues. First, the social simulation categories in the two coding systems may be slightly different even though they are for the most part similar. Second, and far more likely to account for most of the differences, there are fewer codeable elements in general and fewer social interactions in particular in the REM and non-REM reports in the database used by the social simulation theorists than in the nonlab reports that comprise the HVdC norms. This fact was discovered by an earlier published comparison of the HVdC normative samples with two carefully drawn samples from the same dormitory-collected database that was used by the social simulation theorists. The samples from the dorm-collected database used in that earlier comparison consisted of 100 REM and 100 non-REM reports of approximately the same length, and they were coded with HVdC categories for social interactions (McNamara, McLaren, & Durso, 2007; McNamara, McLaren, Smith, Brown, & Stickgold, 2005). (There was no control for time of night in these two studies, but this turned out not to matter according to the results of the social simulation study discussed in this article.)

The relevant and crucial point is that the 24% of REM reports with at least one aggression in both of those earlier studies is well below the figure of 45.4% for the HVdC normative findings when the codings for the 500 men's dream reports and 491 women's dream reports are combined. Then, too, our own unpublished HVdC coding of these dorm-collected dream reports for a large number of categories, which was carried out in an exploratory manner on all the dream reports, with no controls for length, supports the point that there were many less codeable elements in these dream reports than in either REM-collected dreams in a sleep lab or in the HVdC norms. Moreover, the shortest reports in the HVdC sample used in the study in this article contained at least 50 words, whereas the shortest reports in the social simulation theorists' sample contained at least 30 words. For several reasons, then, the differences from the HVdC norms in the social simulation theorists' study are likely due to differences between the two samples.

It therefore seems likely that the differences in findings are not primarily due to differences in the two coding systems. However, the only way to know with certainty that the social simulation theorists' new coding scale is as comprehensive as the social simulation scales based on HVdC coding categories would be to apply it to the same dream reports from men and women upon which our findings are based. Those dream reports are available on DreamBank.net under the name "Hall/Van de Castle norms" (Schneider & Domhoff, 1999).

Assuming that the differences between the dorm-based sample and the HVdC normative sample account for most of the differences between our results and those of the social simulation theorists in regard to the simulation of human social interactions, the most important substantive result of the comparison of the two simulation scales is the discovery that 6.5% of the dream reports in the HVdC normative sample (those with the dreamer and only animals or creatures in them, and those with only the dreamer in them) are not social simulations as defined by social simulation theorists. Instead, we think they serve as ideal examples of dreams as embodied simulations that differentially dramatize personal conceptions and concerns. Although 6.5% is a relatively small difference for most general descriptive comparisons, it is theoretically important when two somewhat similar theories are being closely compared. This is because, all other things equal, more comprehensive theories are usually judged as more adequate than those that are less comprehensive. In addition, the content of these dreams is different than what would be expected if dreams have the adaptive function of preparing people for waking social interactions.

Specifically, the two embodied simulation categories without human social interactions often contain negative events not in keeping with the idea of positive social rehearsals. Instead, they support the more general idea that dreams enact a range of conceptions and concerns. For example, the 2.2% of dream reports that include only the dreamer and an animal (or, on very rare occasions, a creature) enact a widely shared personal concern. In this instance, the main conception and personal concern is that animals can be a great danger to the dreamer. This point is exemplified by the fact that there is a higher physical aggression percent in these dream reports than in the overall HVdC normative findings, and the animals are far more often the aggressors than the dreamers. Moreover, the percentage of these dreams in which apprehension is experienced is almost three times higher than the normative figure (62.8% vs. 21.8%), sometimes at such a high level that the dreamers report that they awakened with a lingering sense of fear.

In the case of the 4.3% of dream reports that include only the dreamer, a wide variety of mostly negative conceptions are embodied in exaggerated forms. For example, people worry about being lost while in isolated areas or experiencing car failure when driving alone on a desolate road. This point is underscored by the large percentage of dreamer-only dreams with at least one misfortune, which is almost twice as high as the normative expectation (62.8% vs. 34.8%).

Thus, the overall findings with the HVdC simulation categories suggest that they may be more useful than the new categories devised by the social simulation theorists. From a methodological perspective, dreams containing social simulations are just as easily detected by extending the time-tested and reliable HVdC coding system as with a purpose-built social simulation scale, rendering the latter superfluous, at least until the comprehensiveness of the social simulation theorists' new scale is ascertained by applying it to the extant HVdC normative dream reports for men and women available on DreamBank.net. The HVdC simulation categories also detect two other types of dreams as well, those with only animals and the dreamer and those with only the dreamer.

The deployment of the five new HVdC embodied simulation categories also seems more useful because any new findings with them might be better understood if the replicated normative findings, based on large sample sizes, revealed other differences as well. The need for replication with larger sample sizes notwithstanding, the present study provides tentative evidence that the HVdC normative findings for misfortunes and types of emotions yield useful information for understanding the content within two of the new categories for embodied simulations.

The methodological advantages of using the HVdC coding system aside, the small but significant percentage of dreams uncovered with the new HVdC simulation categories that feature only the dreamer, or only the dreamer and animals or creatures, have theoretical implications as well. They strongly suggest that dreams are first of all embodied simulations of conceptions and personal concerns, not social simulations, which are only one (albeit frequent) type of embodied simulation in dreams. Then, too, these are strong findings because the sample sizes are large and the two additional types of embodied simulations are found for both men and women.

Moreover, it would be necessary to combine the social simulation theory and the threat simulation theory into a more general social rehearsal theory to encompass the full range of embodied simulations in dreams. On the one hand, the social simulation theory downplays the amount of physical aggression in dreams and leaves out dreamer-only dreams as well as those dreams that contain only the dreamer and animals or creatures. On the other hand, the threat simulation theory leaves out the 25%-35% of dreams that contain no "threats" (which are broadly defined in a way that turns out to almost completely overlap with what are called physical aggressions nonphysical aggressions, misfortunes, and failures in the HVdC coding system; Domhoff, 2018, pp. 261-262; Revonsuo & Valli, 2000; Valli & Revonsuo, 2009). Thus, to repeat, only a combination of the two can come close to matching the results from the new HVdC categories for the types of embodied simulations and for the ability of the neurocognitive theory to encompass the full range of dream content.

At that point, there might appear to be a stalemate as far as choosing between the two theories. That is, the social and threat simulation theories, taken together, would encompass interactions with animals, although they still would be unable to explain the 4.3% of dreamer-only dreams. However, all social rehearsal theories, including two not discussed in this article, face four other major problems that eliminate any apparent stalemate (Brereton, 2000; Franklin & Zyphur, 2005, for two rehearsal theories that have not been studied empirically since they were put forward). Any one of these four unexplained difficulties undercuts all four of the social rehearsal theories.

First, due to their orientation toward the dreamer's present circumstances and future social interactions, no social rehearsal theory can easily explain the findings on the frequent appearance of past failures in competitive situations and in social relationships over the time span of years and decades in lengthy dream series, and in particular those that concern failed relationships and deceased loved ones. Although it might at first glance seem plausible to argue that the significant percentage of dreams about past issues may be attempts to resolve those issues so that more of dream life can be devoted to social rehearsals related to future events, what is striking about the several examples that follow is that the issues are resolved very slowly, if ever. If social rehearsals are an important function of dreaming, they are a function that is severely limited by dreaming about events that are not social rehearsals.

For example, a well-educated professional woman (pseudonym "Dorothea") wrote down 904 of her dreams over a period of 53 years, which she did not offer to investigators until she was in her early 70s, a few years before she died at age 78 (Domhoff, 1996, pp. 142-145, 206). Aside from the consistencies on various HVdC indicators in a systematic quantitative study of the first 600 dreams, she dreamed about her parents with the same frequency throughout the 53 years of recorded dreams, although her father died when she was very young and her mother died when she was 61.

On a somewhat more qualitative note, there were also six themes that appeared regularly in her dream reports, which accounted for at least part of the content in just over 70% of her dreams (Domhoff, 1993, p. 302). First, and most strikingly, she was eating, preparing to eat, preparing a meal, buying or seeing food, watching someone eat, or mentioning she is hungry in just over 20% of her dream reports. In many of these dreams, she is at the family dinner table with her parents and some of her several siblings, and is often being treated unfairly as to portions; her last recorded dream, 4 days before she died, concerned this theme:

Mother had dished out too liberally to the younger children so asked E [one of the dreamer's brothers] to give her some of his. I still had nothing. Then we saw a potato on the floor by the door and it was divided with me. (Domhoff, 1996, p. 145)

In about 16% of Dorothea's dreams, she lost an object, which was most often her purse, and in about 10% of her dreams, she was in a small or disorderly room, or people were barging into her room. Another 10% involved the dreamer and her mother doing something together. The dreamer was going to the toilet in about 8% of the dreams, and she was late or worried about being late, or missing a bus or train, in 6% (Domhoff, 1993, p. 302). This consistency in dreaming about her deceased parents, and the repetition of several themes, some of which have nothing to do with social interactions, does not accord well with the idea of social rehearsal for future social interactions.

A more dramatic and poignant example of the repetitiveness in dream series concerns the 315 dream reports written down over a period of several years from the end of the 1990s to the early 2000s by an artist ("Merri") in her late 30s. Her dream series is mentioned briefly by the social simulation theorists in a discussion that primarily concerns the relationship between the characters in dream series and how frequently they are part of social interactions with the dreamer in waking life. They rightly note "the most frequently occurring character is her sister, who was no longer alive at the time when the dream reports were collected" (Revonsuo et al., 2015, p. 21). However, they do not mention that this older sister was no longer alive because she was killed 3 years before the dream series began by a hit-and-run driver while she was jogging. She was in a hospital on life support for 5 days, with the dreamer and her brother by their sister's side when she died (Bulkeley, 2009, p. 98).

Contrary to what a forward-looking social rehearsal theory might expect, Merri's deceased sister appeared in 34.3% of the dream reports in this series. These dreams express the dreamer's major personal concern: her preoccupation with the deceased sister she deeply admired for her many accomplishments. If these dreams represent social rehearsals that will allow her to move forward in waking life, they were very unsuccessful during the time span during which she recorded her dreams. In addition, at least one member of her nuclear family (her mother, father, brother, and deceased sister) appeared in 60.0% of her dream reports, even though she had moved far away from the state in which she grew up and seldom saw her three closest living relatives. (These and other findings on the Merri series are derived from word searches on DreamBank.net. They are presented in the "Information" and "Further Information" pages on Merri on that website. Additional information on Merri, along with a discussion of her family, upbringing, and spiritual life, based on blind inferences and the dreamer's responses to them, is also available; Bulkeley, 2009.)

Overall, the main findings with the Merri series fit better with the claim that dreams dramatize the same conceptions and personal concerns that animate waking thought than with the idea that dreams are social rehearsals with adaptive value. These kinds of results also call into question the social simulation theorists' use of the concept of "predictive codes" — a concept coined to replace "expectancies" as the way to talk about what the brain tries to create in waking life in the context of incoming sensory information — in theorizing about dreams (Revonsuo et al., 2015, p. 6).

Among many other examples of differential dreaming about people and events from the past that could be provided, the results from a study of the dreams of "Ed" may be the most difficult for any social rehearsal theory to accommodate. This unique series consists of 143 dream reports that were written down over the space of 22 years by a widower because they were about his deceased wife (Domhoff, 2015). The pleasure of seeing and experiencing her presence in his dreams (including dreams shortly after her death, in which she assured him she was comfortable, and even expressed approval of him remarrying) led him to write down these dreams and to comment upon them in a journal that he did not think about sharing with dream researchers until many years after he began documenting his dreams.

Inferences were drawn from a blind HVdC analysis of emotions, misfortunes, good fortunes, and A/F/S interactions in the dream series; these inferences were consistent with the widower's written reflections on his marriage and his answers to questions sent to him by the researcher, which is strong support for the neurocognitive theory of dreaming as embodied simulation of conceptions and personal concerns.

Rather than being rehearsals for waking life, these dreams reenact Ed's concern for his deceased wife's health in the last few years of her life and dramatize his conflicted relationship with her:

In his dreams he wishes his wife were still alive, sometimes wants her help or assurance, can't quite believe she is really dead, on occasion enacts fond memories of the good times they had together, often puzzles over the sexual tensions in their marriage, and relives the horrible events of the terminal illness that ended her life. (Domhoff, 2015, p. 20)

Not only were these dreams far from being social rehearsals, but also they did not change much over the space of 22 years. In one of the last dreams he wrote down about her, he saw her coming back home because an operation had proven successful after all. Although there is no doubt that there were numerous varying topics in the dreams that Ed recalled but did not write down, the point is that the portion of his dream life that concerned his wife is akin to our previous examples in that it cannot be explained by any of the social rehearsal theories (Brereton, 2000; Franklin & Zyphur, 2005; Revonsuo, 2000; Revonsuo et al., 2015).

Social rehearsal theories, with their emphasis on the adaptive value of dreaming, are also called into question, at least until preadolescence, by descriptive empirical findings from longitudinal and cross-sectional studies in sleep labs. Those studies reveal the rarity and meager content of the dreams of children of ages 3-5, and their relative infrequency and only slightly greater complexity up to age 9, a period when children are acquiring enormous social perception, social interaction, and threat avoidance skills (Foulkes, 1982b, 1999; Foulkes et al., 1990). In the case of threat simulation theory, its originator argued that traumatic events lead to a more rapid development of the capacity to dream in children (Revonsuo, 2000, p. 889, Note 5). This argument is implausible in terms of what is known about the gradual sequential unfolding of cognitive and neural development, but it cannot be put forward in the case of a social simulation theory because there is no claim that everyday social interactions are somehow traumatic (Domhoff, 2018, Chapters 4-5; Fair et al., 2008, 2009; Nelson, 2005).

Nor is there anything about the content of children's dream reports collected between the ages of 5 and 9 in the longitudinal and cross-sectional studies that fits with either the social simulation or the threat simulation version of a social rehearsal theory (Foulkes, 1982b, 1999; Foulkes et al., 1990). To begin with, there are few of the strangers that are hypothesized to be helpful, according to social simulation theory, in learning to distinguish strangers and people that are known (Revonsuo et al., 2015, pp. 15-17). There are few friendly interactions in which to practice social simulations, nor even many aggressive interactions (at least until preadolescence and early adolescence) from which to learn to deal with threats (Domhoff, 1996, pp. 93-94, for HVdC findings with Foulkes's cross-sectional dream reports). Dreams could not possibly play the roles assigned to them by either social simulation or threat simulation theory before ages 9-11, and more likely 11-13, when more aggressions and more indications of personal concerns gradually begin to appear. In the face of these findings, perhaps it could be argued that the adaptive social rehearsals in dreams require the complex social cognition that develops between ages 9 and 11, but no adaptive theories focused on teenagers and adults have been proposed as yet.

In addition to the empirical findings on the dream content of both adults and children that do not fit with social rehearsal theories, there is a theoretical problem as well: These theories are based on two tacit assumptions about learning and memory during sleep for which there is no evidence. To begin with, these theories assume that there is "implicit learning" during sleep, that is, learning that is not intentional and happens without awareness. There is no one accepted definition of implicit learning, it can be of many types, and its scope and importance are still being debated. Most experimental psychologists assign minor importance at best to implicit learning, due to its limited scope, its occurrence in only brief durations, and its relevance for only a few types of information, as conceded in an article that attempts to reopen the issue from a neurocognitive vantage point (Ryals & Voss, 2015, pp. 44-45). As one of the many doubters put it after discussing the need for attention and awareness for the occurrence of implicit sequence learning, which is the most relevant type of implicit learning in terms of the sequential, quasinarrative nature of most dreams:

It is not clear that either of the claims about implicit sequence learning described at the outset of this chapter has been established. If the goal is to demonstrate the existence of a form of learning which is both functionally and neurally separate from explicit learning then I would argue that such a goal has not yet been achieved. (Shanks, 2003, p. 38)

The apparently limited nature of implicit learning to one side, it is also the case that there is no direct evidence for implicit learning during sleep or dreaming. Any form of adaptive rehearsal theory of dreams that invokes implicit learning therefore has to rely on arguments extrapolated from studies of memory consolidation of waking learning during sleep (Revonsuo, 2000, p. 1081). However, new learning and memory consolidation are two different processes (Domhoff, 2018, pp. 268-275, for the lack of evidence for any connection between memory consolidation and dreaming).

Beyond the assumption of implicit learning, there is the issue of "transfer of learning," or "transfer of training," which concerns the degree to which learning based on one task can be used in the mastery of a new task. Transfer of learning, like implicit learning, seems to be limited in its scope according to most researchers, with little or no evidence for improvement on other types of tasks that are outside of tightly related, overlapping domains (Sala & Gobet, 2017; Sala, Tatlidil, & Gobet, 2017, for meta-analyses and overviews of large literatures). Still, one pair of related studies suggest that there may be at least some degree of transfer of learning if the focus is on learning a general principle (Garner, Lynch, & Dux, 2016). However, another study carried out by some of the same authors provides evidence that the necessary capacities for this sophisticated form of transfer, if such studies can be replicated, involve the regions in the brain that support the central executive network — especially the dorsolateral prefrontal cortex (Verghese, Garner, Mattingley, & Dux, 2016). This focus on the dorsolateral prefrontal cortex presents a major stumbling block for social rehearsal theories, because most of the dorsolateral prefrontal cortex, as well as other prefrontal regions supporting executive functions, are deactivated in both REM and non-REM sleep (Dang-Vu, Schabus, Desseilles, Schwartz, & Maquet, 2007; Domhoff, 2018, Chapter 5; Fox et al., 2013).

Although advocates of the threat and social simulation theories have continued to pursue their search for an adaptive function for dreaming, they do note that dreaming could be "non-functional," which they in effect say would be a sensible conclusion if their adaptive theories prove to be wrong (Revonsuo et al., 2015, pp. 1, 3). However, their argument is given over to lengthy discussions on "possible functions " and on "why it would be useful to simulate" in dreams, to take only two of many possible examples (Revonsuo et al., 2015, pp. 2, 3, italics in the original). They therefore do not discuss the final major difficulty facing all social rehearsal theories: There are several long-known empirical reasons why it is very unlikely that dreams have any adaptive value, as first explained by cognitive dream researchers decades ago (Antrobus, 1993; Blagrove, 1992, 2000; Foulkes, 1985, pp. 196-204; Foulkes, 1993).

These reasons begin with the finding that dreaming occurs at sleep onset, in both REM and non-REM sleep, and even in brief episodes during relaxed waking thought; this makes dreaming a very diffuse state, not limited to REM sleep, or even to sleep (Domhoff, 2018, Chapter 1; Foulkes, 1985, Chapter 2, for detailed overviews of the evidence). As far as REM sleep itself, it is by now a commonplace to note that any function that is finally shown for REM sleep, which has yet to be discovered, does not imply a function for dreaming, especially when it is added that every land animal has widely varying amounts of REM sleep, but only humans of age 5 and older show any evidence of dreaming (Feinberg, 2004; Foulkes, 1983, 2017; Siegel, 2005, 2017). Then there is the surprising post-1953 fact of the very infrequent recall of the several dreams of the night (likely less than 5%-7% of all dreams for most people, if it is conservatively assumed that most people have at least six or seven dreams per night). This seems to be a much lower level of recall than would be expected if dream content contains problem-solving ideas that people need to recall. Beyond low recall, there is also the almost complete absence of systematic evidence that dreams ever contain solutions to problems of any kind, personal or intellectual (Domhoff, 2003, pp. 157-167, for a discussion of the problems with a range of problem-solving theories, including the questionable nature of famous anecdotal examples sometimes put forth as evidence). These general conclusions, based for the most part on the first 20 years of dream research after the discovery of REM sleep in 1953, had led to the abandonment of most pre-REM theories of dream function by the early 1970s (Dallett, 1973). Thus, lab studies served to reduce the plausible functions of dreaming to those effects that allegedly happen without any recall of dreams, which for the most part so far means the various social rehearsal theories.

Finally, there is evidence that dreaming can be largely or completely absent in a tiny few adolescents (Foulkes, 1982b, pp. 181, 225-226) and in adults without any obvious ill effects (Pagel, 2003). In addition, the temporary or permanent loss of dreaming in adults due to neurological defects or lesions does not lead to any striking or obvious changes in consciousnesses or behavior in the space of days, weeks, or months (Bischof & Bassetti, 2004; Domhoff, 2003, pp. 12-15; Jus, Jus, Villeneuve, et al., 1973; Kerr & Foulkes, 1981; Kerr et al., 1978; Poza & Martí Massó, 2006; Solms, 1997).

For a variety of reasons then, any one of which raises serious questions about any adaptive theory of dreaming, including the various social rehearsal theories, it seems increasingly unlikely that dreaming has any adaptive function. It seems more likely that dreaming is a byproduct of waking cognitive capacities that turned out to have great adaptive value (Domhoff, 2018, pp. 275-276; Foulkes, 1999, pp. 137-142; Foulkes, 2017). Dreaming has psychological and cultural "uses" that were invented by human beings in the course of history, such as in religious ceremonies and healing functions, but uses created due to the imaginative capacities of the human mind are not adaptive functions based on natural selection in the course of human evolution (Domhoff, 2018, pp. 276-278).


A neurocognitive theory that includes the idea that dreams are embodied simulations is more viable than any version of any social rehearsal theory, including the social simulation theory focused on in this article. It encompasses the widest range of adult dream content and does not conflict with what is known about cognitive development in children. Moreover, the neurocognitive theory does not necessitate unsubstantiated assumptions about implicit learning and/or transfer of training, nor insist upon an adaptive function for dreaming in the face of strong empirical evidence to the contrary.

If theories are meant to explain the widest range of findings about phenomena with hypotheses based on solidly grounded past findings, and at the same time generate ideas for further studies that potentially could add nuance to the theory or refute it, then the neurocognitive theory of dreams defended in this and other venues far outdistances social rehearsal theories at this time (Domhoff, 2011, 2015, 2018; Domhoff & Fox, 2015).


Ajilore, O., Stickgold, R., Rittenhouse, C. D., & Hobson, J. A. (1995). Nightcap: Laboratory and home-based evaluation of a portable sleep monitor. Psychophysiology, 32, 92-98. http://dx.doi.org/10.1111/j.1469-8986.1995.tb03410.x

Allport, G. (1942). The use of personal documents in psychological science New York, NY: Social Science Research Council.

Andrews-Hanna, J. R., Reidler, J. S., Huang, C.,&Buckner, R. L. (2010). Evidence for the default network's role in spontaneous cognition. Journal of Neurophysiology, 104, 322-335. http://dx.doi.org/10.1152/jn.00830.2009

Antrobus, J. (1993). Dreaming: Could we do without it? In A. Moffitt, M. Kramer, & R. Hoffmann (Eds.), The functions of dreaming (pp. 549-558). Albany, NY: SUNY Press.

Baldwin, A. (1942). Personal structure analysis: A statistical method for investigating the single personality. The Journal of Abnormal and Social Psychology, 37, 163-183. http://dx.doi.org/10.1037/h0061697

Beck, A. T., & Hurvich, M. S. (1959). Psychological correlates of depression. 1. Frequency of masochistic dream content in a private practice sample. Psychosomatic Medicine, 21, 50-55. http://dx.doi.org/10.1097/00006842-195901000-00007

Bell, A., & Hall, C. (1971). The personality of a child molester: An analysis of dreams Chicago, IL: Aldine.

Bischof, M., & Bassetti, C. L. (2004). Total dream loss: A distinct neuropsychological dysfunction after bilateral PCA stroke. Annals of Neurology, 56, 583-586. http://dx.doi.org/10.1002/ana.20246

Blagrove, M. (1992). Dreams as a reflection of our waking concerns and abilities: A critique of the problem-solving paradigm in dream research. Dreaming, 2, 205-220. http://dx.doi.org/10.1037/h0094361

Blagrove, M. (2000). Dreams have meaning but no function. Behavioral and Brain Sciences, 23, 910-911. http://dx.doi.org/10.1017/S0140525X00244026

Brereton, D. (2000). Dreaming, adaptation, and consciousness: The social mapping hypothesis. Ethos, 28, 379-409. http://dx.doi.org/10.1525/eth.2000.28.3.379

Buckner, R. L., Andrews-Hanna, J. R., & Schacter, D. L. (2008). The brain's default network: Anatomy, function, and relevance to disease. Annals of the New York Academy of Sciences, 1124, 1-38. http://dx.doi.org/10.1196/annals.1440.011

Bulkeley, K. (2009). The religious content of dreams: A new scientific foundation. Pastoral Psychology, 58, 93-106. http://dx.doi.org/10.1007/s11089-008-0180-8

Clark, J., Trinder, J., Kramer, M., Roth, T., & Day, N. (1972). An approach to the content analysis of dream scales. In M. Chase, W. Stern, & P. Walter (Eds.), Sleep research (Vol. 1, p. 118). Los Angeles, CA: Brain Research Institute, UCLA.

Dale, A., Lortie-Lussier, M., Wong, C., & De Koninck, J. (2016). Dreams of Canadian students: Norms, gender differences, and comparison with Americans. Journal of Cross-Cultural Psychology, 47, 941-955. http://dx.doi.org/10.1177/0022022116655788

Dallett, J. (1973). Theories of dream function. Psychological Bulletin, 79, 408-416. http://dx.doi.org/10.1037/h0034485

Dang-Vu, T., Schabus, M., Desseilles, M., Schwartz, S., & Maquet, P. (2007). Neuroimaging of REM sleep and dreaming. In D. Barrett & P. McNamara (Eds.), The new science of dreaming: Biological aspects (Vol. 1, pp. 95-113). Westport, CT: Praeger/Greenwood.

Domhoff, G. W. (1993). The repetition of dreams and dream elements: A possible clue to a function of dreams. In A. Moffitt, M. Kramer, & R. Hoffmann (Eds.), The Functions of Dreams (pp. 293-320). Albany, NY: SUNY Press.

Domhoff, G. W. (1996). Finding meaning in dreams: A quantitative approach New York, NY: Plenum Press. http://dx.doi.org/10.1007/978-1-4899-0298-6

Domhoff, G. W. (2001). A new neurocognitive theory of dreams. Dreaming, 11, 13-33. http://dx.doi.org/10.1023/A:1009464416649

Domhoff, G. W. (2003). The scientific study of dreams: Neural networks, cognitive development, and content analysis Washington, DC: American Psychological Association. http://dx.doi.org/10.1037/10463-000

Domhoff, G. W. (2007). Realistic simulation and bizarreness in dream content: Past findings and suggestions for future research. In D. Barrett & P. McNamara (Eds.), The new science of dreaming: Content, recall, and personality correlates (Vol. 2, pp. 1-27). Westport, CT: Praeger.

Domhoff, G. W. (2011). The neural substrate for dreaming: Is it a subsystem of the default network? Consciousness and Cognition, 20, 1163-1174. http://dx.doi.org/10.1016/j.concog.2011.03.001

Domhoff, G. W. (2015). Dreaming as embodied simulation: A widower dreams of his deceased wife. Dreaming, 25, 232-256. http://dx.doi.org/10.1037/a0039291

Domhoff, G. W. (2018). The emergence of dreaming: Mind-wandering, embodied simulation, and the default network New York, NY: Oxford University Press.

Domhoff, G. W., & Fox, K. C. (2015). Dreaming and the default network: A review, synthesis, and counterintuitive research proposal. Consciousness and Cognition: An International Journal, 33, 342-353. http://dx.doi.org/10.1016/j.concog.2015.01.019

Domhoff, G. W., & Schneider, A. (2008). Similarities and differences in dream content at the cross-cultural, gender, and individual levels. Consciousness and Cognition, 17, 1257-1265. http://dx.doi.org/10.1016/j.concog.2008.08.005

Domhoff, G. W., & Schneider, A. (2015). Correcting for multiple comparisons in studies of dream content: A statistical addition to the Hall/Van de Castle coding system. Dreaming, 25, 59-69. http://dx.doi.org/10.1037/a0038790

Dorus, E., Dorus, W., & Rechtschaffen, A. (1971). The incidence of novelty in dreams. Archives of General Psychiatry, 25, 364-368. http://dx.doi.org/10.1001/archpsyc.1971.01750160076014

Dudley, L., & Fungaroli, J. (1987). The dreams of students in a women's college: Are they different? ASD Newsletter, 4, 6-7.

Dudley, L., & Swank, M. (1990). A comparison of the dreams of college women in 1950 and 1990. ASD Newsletter, 7, 3.

Fair, D. A., Cohen, A. L., Dosenbach, N. U., Church, J. A., Miezin, F. M., Barch, D. M., . . . Schlaggar, B. L. (2008). The maturing architecture of the brain's default network. Proceedings of the National Academy of Sciences of the United States of America, 105, 4028-4032. http://dx.doi.org/10.1073/pnas.0800376105

Fair, D. A., Cohen, A. L., Power, J. D., Dosenbach, N. U., Church, J. A., Miezin, F. M., . . . Petersen, S. E. (2009). Functional brain networks develop from a "local to distributed" organization. PLoS Computational Biology, 5, e1000381. http://dx.doi.org/10.1371/journal.pcbi.1000381

Feinberg, I. (2004). Hobson on the neuroscience of sleep: A case of persistent REM tunnel vision. The American Journal of Psychology, 117, 621-627. http://dx.doi.org/10.2307/4148996

Foulkes, D. (1982a). A cognitive-psychological model of REM dream production. Sleep, 5, 169-187. http://dx.doi.org/10.1093/sleep/5.2.169

Foulkes, D. (1982b). Children's dreams: Longitudinal studies. New York, NY: Wiley.

Foulkes, D. (1983). Cognitive processes during sleep: Evolutionary aspects. In A. Mayes (Ed.), Sleep mechanisms and functions in humans and animals: An evolutionary perspective (pp. 313-337). Workington, United Kingdom: Van Nostrand Reinhold.

Foulkes, D. (1985). Dreaming: A cognitive-psychological analysis Hillsdale, NJ: Erlbaum.

Foulkes, D. (1993). Data constraints on theorizing about dream function. In A. Moffitt, M. Kramer, & R. Hoffmann (Eds.), The functions of dreaming (pp. 11-20). Albany, NY: SUNY Press.

Foulkes, D. (1999). Children's dreaming and the development of consciousness Cambridge, MA: Harvard University Press.

Foulkes, D. (2017). Dreaming, reflective consciousness, and feelings in the preschool child. Dreaming, 27, 1-13. http://dx.doi.org/10.1037/drm0000040

Foulkes, D., & Fleisher, S. (1975). Mental activity in relaxed wakefulness. Journal of Abnormal Psychology, 84, 66-75. http://dx.doi.org/10.1037/h0076164

Foulkes, D., Hollifield, M., Sullivan, B., Bradley, L., & Terry, R. (1990). REM dreaming and cognitive skills at ages 5-8: A cross-sectional study. International Journal of Behavioral Development, 13, 447-465. http://dx.doi.org/10.1177/016502549001300404

Foulkes, D., & Scott, E. (1973). An above-zero baseline for the incidence of momentarily hallucinatory mentation. Sleep Research, 2, 108.

Fox, K., Nijeboer, S., Solomonova, E., Domhoff, G. W., & Christoff, K. (2013). Dreaming as mind wandering: Evidence from functional neuroimaging and first-person content reports. Frontiers in Human Neuroscience, 7, 1-18. http://dx.doi.org/10.3389/fnhum.2013.00412.eCollection02013

Franklin, M., & Zyphur, M. (2005). The role of dreams in the evolution of the human mind. Evolutionary Psychology, 3, 59-78. http://dx.doi.org/10.1177/147470490500300106

Garner, K. G., Lynch, C. R., & Dux, P. E. (2016). Transfer of training benefits requires rules we cannot see (or hear). Journal of Experimental Psychology: Human Perception and Performance, 42, 1148-1157. http://dx.doi.org/10.1037/xhp0000215

Hall, C. (1951). What people dream about. Scientific American, 184, 60-63. http://dx.doi.org/10.1038/scientificamerican0551-60

Hall, C. (1953). A cognitive theory of dreams. Journal of General Psychology, 49, 273-282. http://dx.doi.org/10.1080/00221309.1953.9710091

Hall, C. S., Domhoff, G. W., Blick, K. A., & Weesner, K. E. (1982). The dreams of college men and women in 1950 and 1980: A comparison of dream contents and sex differences. Sleep, 5, 188-194. http://dx.doi.org/10.1093/sleep/5.2.188

Hall, C., & Lind, R. (1970). Dreams, life and literature: A study of Franz Kafka Chapel Hill, NC: University of North Carolina Press.

Hall, C., & Van de Castle, R. (1966). The content analysis of dreams New York, NY: Appleton-Century- Crofts.

Hartmann, E. (2011). The nature and functions of dreaming New York, NY: Oxford University Press.

Jus, A., Jus, K., Gautier, J., Villeneuve, A., Pires, P., Lachance, R., . . .. (1973). Dream reports after reserpine in chronic lobotomized schizophrenic patients [The recall of dreams after administration of reserpine in chronic schizophrenics treated with prefrontal lobotomy]. La Vie Medicale au Canada Francais, 2, 843-848.

Jus, A., Jus, K., Villeneuve, A., Pires, A., Lachance, R., Fortier, J., & Villeneuve, R. (1973). Studies on dream recall in chronic schizophrenic patients after prefrontal lobotomy. Biological Psychiatry, 6, 275-293.

Kerr, N. H., & Foulkes, D. (1981). Right hemispheric mediation of dream visualization: A case study. Cortex, 17, 603-609. http://dx.doi.org/10.1016/S0010-9452(81)80066-4

Kerr, N., Foulkes, D., & Jurkovic, G. (1978). Reported absence of visual dream imagery in a normally sighted subject with Turner's syndrome. Journal of Mental Imagery, 2, 247-264.

Maquet, P., Péters, J., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A., & Franck, G. (1996). Functional neuroanatomy of human rapid-eye-movement sleep and dreaming. Nature, 383, 163-166. http://dx.doi.org/10.1038/383163a0

Mason, M. F., Norton, M. I., Van Horn, J. D., Wegner, D. M., Grafton, S. T., & Macrae, C. N. (2007). Wandering minds: The default network and stimulus-independent thought. Science, 315, 393-395. http://dx.doi.org/10.1126/science.1131295

Mazandarani, A. A., Aguilar-Vafaie, M. E., & Domhoff, G. W. (2013). Content analysis of Iranian college students' dreams: Comparison with American data. Dreaming, 23, 163-174. http://dx.doi.org/10.1037/a0032352

McNamara, P., McLaren, D., & Durso, K. (2007). Representation of the self in REM and NREM dreams. Dreaming, 17, 113-126. http://dx.doi.org/10.1037/1053-0797.17.2.113

McNamara, P., McLaren, D., Smith, D., Brown, A., & Stickgold, R. (2005). A "Jekyll and Hyde" within: Aggressive versus friendly interactions in REM and non-REM dreams. Psychological Science, 16, 130-136. http://dx.doi.org/10.1111/j.0956-7976.2005.00793.x

Nelson, K. (2005). Emerging levels of consciousness in early human development. In H. S. Terrace & J. Metcalfe (Eds.), The missing link in cognition: Origins of self-reflective consciousness (pp. 116-141). New York, NY: Oxford University Press. http://dx.doi.org/10.1093/acprof:oso/9780195161564.003.0004

Nofzinger, E. A., Mintun, M. A., Wiseman, M., Kupfer, D. J., & Moore, R. Y. (1997). Forebrain activation in REM sleep: An FDG PET study. Brain Research, 770, 192-201. http://dx.doi.org/10.1016/S0006-8993(97)00807-X

Nofzinger, E. A., Nichols, T. E., Meltzer, C. C., Price, J., Steppe, D. A., Miewald, J. M., Moore, R. Y. (1999). Changes in forebrain function from waking to REM sleep in depression: Preliminary analyses of [18F]FDG PET studies. Psychiatry Research: Neuroimaging, 91, 59-78. http://dx.doi.org/10.1016/S0925-4927(99)00025-6

Pace-Schott, E. (2011). The neurobiology of dreaming. In M. Kryger, T. Roth, & W. Dement (Eds.), Principles and practices of sleep medicine (5th ed., pp. 563-575). Philadelphia, PA: Elsevier. http://dx.doi.org/10.1016/B978-1-4160-6645-3.00048-7

Pagel, J. F. (2003). Non-dreamers. Sleep Medicine, 4, 235-241. http://dx.doi.org/10.1016/S1389-9457(02)00255-1

Poza, J. J., & Martí Massó, J. F. (2006). Total dream loss secondary to left temporo-occipital brain injury. Neurologia, 21, 152-154.

Prasad, B. (1982). Content analysis of dreams of Indian and American college students: A cultural comparison. Journal of Indian Psychology, 4, 54-64.

Raichle, M. E., MacLeod, A. M., Snyder, A. Z., Powers, W. J., Gusnard, D. A., & Shulman, G. L. (2001). A default mode of brain function. Proceedings of the National Academy of Sciences of the United States of America, 98, 676-682. http://dx.doi.org/10.1073/pnas.98.2.676

Reichers, M., Kramer, M., & Trinder, J. (1970). A replication of the Hall-Van de Castle Character Scale norms. Psychophysiology, 7, 238.

Revonsuo, A. (2000). The reinterpretation of dreams: An evolutionary hypothesis of the function of dreaming. Behavioral and Brain Sciences, 23, 877-901. http://dx.doi.org/10.1017/S0140525X00004015

Revonsuo, A., Tuominen, J., & Valli, K. (2015). The avatars in the machine: Dreaming as a simulation of social reality. In T. Metzinger & J. Windt (Eds.), Open MIND: 32(T) (pp. 1-28). Frankfurt am Main, Germany: MIND Group.

Revonsuo, A., & Valli, K. (2000). Dreaming and consciousness: Testing the threat simulation theory of the function of dreaming. Psyche: An Interdisciplinary Journal of Research on Consciousness, 6, 1-31.

Ryals, A., & Voss, J. (2015). The outer limits of implicit memory. In D. R. Addis, M. Barense & Duarte (Eds.), The Wiley handbook on the cognitive neuroscience of memory (pp. 44-59). New York, NY: Wiley-Blackwell. http://dx.doi.org/10.1002/9781118332634.ch3

Sala, G., & Gobet, F. (2017). Does far transfer exist? Negative evidence from chess, music, and working memory training. Current Directions in Psychological Science, 26, 515-520. http://dx.doi.org/10.1177/0963721417712760

Sala, G., Tatlidil, K. S., & Gobet, F. (2017). Video game training does not enhance cognitive ability: A comprehensive meta-analytic investigation. Psychological Bulletin, 144, 111-139. http://dx.doi.org/10.1037/bul0000139

Schneider, A., & Domhoff, G. W. (1999). DreamBank.net. Retrieved from www.dreambank.net

Shanks, D. R. (2003). Attention and awareness in "implicit" sequence learning. In L. Jimenez (Ed.), Attention and implicit learning (pp. 11-42). Philadelphia, PA: John Benjamins. http://dx.doi.org/10.1075/aicr.48.05sha

Siegel, J. M. (2005). Clues to the functions of mammalian sleep. Nature, 437, 1264-1271. http://dx.doi.org/10.1038/nature04285

Siegel, J. (2017). Sleep in animals: A state of adaptive inactivity. In M. Kryger, T. Roth, & W. Dement (Eds.), Principles and practices of sleep medicine (5th ed., pp. 103-114). Philadelphia, PA: Elsevier. http://dx.doi.org/10.1016/B978-0-323-24288-2.00010-6

Snyder, F. (1970). The phenomenology of dreaming. In L. Madow & L. Snow (Eds.), The psychodynamic implications of the physiological studies on dreams (pp. 124-151). Springfield, IL: Thomas.

Solms, M. (1997). The neuropsychology of dreams: A clinico-anatomical study Hillsdale, NJ: Erlbaum.

Stickgold, R., Malia, A., Fosse, R., Propper, R., & Hobson, J. A. (2001). Brain-mind states: I. Longitudinal field study of sleep/wake factors influencing mentation report length. Sleep: Journal of Sleep and Sleep Disorders Research, 24, 171-179. http://dx.doi.org/10.1093/sleep/24.2.171

Strauch, I., & Meier, B. (1996). In search of dreams: Results of experimental dream research Albany, NY: SUNY Press.

Tonay, V. (1990/1991). California women and their dreams: A historical and sub-cultural comparison of dream content. Imagination, Cognition and Personality, 10, 85-97. http://dx.doi.org/10.2190/M29JQQTB-NMYD-QP1F

Tuominen, J., Stenberg, T., Revonsuo, A., & Valli, K. (2018). Social contents in dreams: An empirical test of the social simulation theory. Manuscript submitted for publication.

Valli, K., & Revonsuo, A. (2009). The threat simulation theory in light of recent empirical evidence: A review. The American Journal of Psychology, 122, 17-38. Van de Castle, R. (1969). Problems in applying methodology of content analysis. In M. Kramer (Ed.), Dream psychology and the new biology of dreaming (pp. 185-197). Springfield, IL: Charles C Thomas.

Verghese, A., Garner, K. G., Mattingley, J. B., & Dux, P. E. (2016). Prefrontal cortex structure predicts training-induced improvements in multitasking performance. The Journal of Neuroscience, 36, 2638-2645. http://dx.doi.org/10.1523/JNEUROSCI.3410-15.2016

Webb, E., Campbell, D., Schwartz, R., Sechrest, L., & Grove, J. (1981). Nonreactive measures in the social sciences Boston, MA: Houghton Mifflin.

Yamanaka, T., Morita, Y., & Matsumoto, J. (1982). Analysis of the dream contents in Japanese college students by REMP-awakening technique. Folia Psychiatrica et Neurologica Japonica, 36, 33-52.


Our thanks to Kelly Bulkeley and David Foulkes for comments on earlier versions of this article, and to Jarno Tuominen and Katja Valli for generously sharing their results with their social simulation scale and for helpful suggestions to improve our discussion of the social simulation theory.

Go back to the Dream Library index.

dreamresearch.net home page dreamresearch.net contact info