Using Content Analysis to Study Dreams:
Applications and Implications for the Humanities

G. William Domhoff

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. (2002). Using content analysis to study dreams: applications and implications for the humanities. In K. Bulkeley (Ed.), Dreams: A Reader on the Religious, Cultural, and Psychological Dimensions of Dreaming (pp. 307-319). New York: Palgrave.


Content analysis -- the use of carefully defined categories and quantitative techniques to find meaningful regularities in texts -- is very different from the usual approaches to dreams. It does not make use of free associations, amplifications, autobiographical statements, or any other information from outside the dream reports themselves. It has led to numerous findings on the relationship of dream content to gender, age, mental health, culture, and individual differences. It can be used in conjunction with any theory, or for descriptive empirical studies. It is most convincing when the content analyst knows nothing about the dreamer.

The application of content analysis to dream reports holds out considerable potential for questions of interest to scholars in the humanities and religious studies, where this method has been used to study literary texts, Shakespeare's plays, fairy tales, folklore, and much else. For example, it often leads to the discovery of patterns in a series of dreams from one person that were not detected by other methods.

Content analysis is very simple in principle, but difficult to carry out in practice. It consists of four basic steps: (1) creating relevant categories that can be understood and applied by any investigator; (2) tabulating frequencies for the categories; (3) using percentages, ratios, or other statistics to transform raw frequencies into meaningful data; and (4) making comparisons with normative samples or control groups. The usefulness of content analysis for the study of dreams has been increased exponentially by the constant improvements in personal computers. They make it possible to store and manipulate the large databases that are necessary to develop normative findings. They provide spreadsheets and other tools that make it easier and more accurate to analyze the codings from complex content analyses of hundreds of dream reports.

Although content analysis is simple in principle, and has been used in many different disciplines, there is no general system that is useful for texts ranging from literature to psychotherapy records to dream reports. Instead, specific systems have to emerge from a familiarity with the kind of text being analyzed. Furthermore, it is difficult to develop rules that always lead to similar codings by individual coders (reliability) and also relate to other cognitive states or behaviors (validity). Once a good system has been developed for a specific kind of text, it makes little or no sense for new investigators to try to develop their own systems, even if they think their questions are unique. It is far better to adapt established systems for their purposes.

This chapter focuses on the findings, implications, and future applications of two proven approaches to the content analysis of dreams, both of which are ready for immediate use and available via the Internet. The first method is a detailed coding system that provides over two dozen content indicators, along with normative findings for those indicators. It is modular and adaptable. The second is a computer-assisted system for making searches for phrases or strings of words within a database of over 10,000 dream reports, which adds up to a fast content analysis. When the two methodologies are used together, it may even be possible to build on the work of cognitive scientists to search for figurative meaning in dream reports.

These two new resources for the study of dream content may be especially appealing because of increasing doubts about other methods. Clinical studies of dreams are now compromised by studies showing that the social psychology of the psychotherapy relationship provides numerous opportunities for therapists to unwittingly shape the seemingly accurate interpretations of dreams that derive from free associations and amplifications (Loftus & Ketcham, 1994; Ofshe & Watters, 1994). Through the process of suggestion and persuasion, the therapist and dreamer may be establishing a common belief system, rather than discovering the "meaning" of the dream. This possibility is supported by experimental studies showing that many participants can be convinced through dream interpretations that they had experiences from before age 3 that they don't remember (Mazzoni, Loftus, Seitz, & Lynn, 1999; Tsai, Loftus, & Polage, 2000).

Similarly, the hypnotic investigation of dreams, as summarized by Moss (1967), has been called into question by studies revealing the manipulative and capricious nature of the relationship between hypnotist and participant (Spanos, 1996). As for symbolic dream interpretation, the continuing wide differences among Freudians, Jungians, and neo-Freudians show there has been no progress toward a common understanding of dream symbolism on the basis of systematic research. In fact, it can be argued that the rival schools of symbolic interpretation are particular applications of common metaphoric understandings to aspects of dreams. Even the methods used by phenomenological and existential theorists, such as Boss (1958; 1977) and Perls (Downing & Marmorstein, 1973), who claim to reject symbolic interpretations, are actually metaphoric glosses of dreams, as can be seen by reading through their case examples.

The Hall/Van De Castle coding system

The system of content analysis first outlined in the 1950s by psychologist Calvin S. Hall, and then finalized with the help of another psychologist, Robert Van de Castle, in the 1960s, is the most comprehensive and widely used coding system yet developed for the study of dreams (Hall, 1951; Hall, 1953c; Hall & Van de Castle, 1966) . It consists of 10 general categories that cover everything from characters to social interactions to such descriptive elements as intensity, size, and temperature. Most categories contain two or more subcategories. Characters, for example, consist of humans, animals, and mythical creatures, and the category for human characters is further subdivided by such features as gender, relation to the dreamer, and age. There are 5 subcategories for emotions, 8 for activities, and 12 for objects. Not all categories need to be used in each study. The most frequently used categories are characters, social interactions, misfortune and good fortune, success and failure, and emotions.

The system rests on the nominal level of measurement, due to problems of reliability and psychological validity with rating scales (Domhoff, 1999b; Hall, 1969; Van de Castle, 1969). The raw frequencies are analyzed using percentages and ratios to correct for the varying length of dream reports from sample to sample and other problems that defeat many other systems for coding dream content. The system includes normative findings on men and women that have been replicated several times. Studies based on subsamples drawn from larger samples show that it takes 100-125 dream reports to achieve results that can be repeated in future studies. The large number is due to the fact that most dream elements appear in less than half of all dream reports (Domhoff, 1996).

The entire coding system can be found on the Internet at www.dreamresearch.net, along with already-coded dreams to aid new researchers (Schneider & Domhoff, 1995). This Web site also includes a spreadsheet that computes the results for the content indicators once the codings are entered, as well as significance levels, effect sizes, and confidence intervals. The results are displayed in tables and in graphs. In addition, dreamresearch.net contains information on how to do research projects, and a library of past papers by Hall, Domhoff, and their students.

The type of findings produced by the system can be seen through a brief look at the male/female percent, the animal percent, and several different aggression indicators. The male/female percent is derived by dividing the total number of male characters by the total number of male plus female characters. In virtually every sample from every culture that has been studied, the male/female percent is about 67/33 for men and 48/52 for women, which means that women's scores are not the mirror image of men's (Hall, 1984; Hall, 1963). The same difference also appears in the dreams and stories of preschoolers and in dreams collected in the sleep laboratory (Domhoff, 1996; Foulkes, 1982). This is an unexpected regularity that is not anticipated by any dream theory. It is also the largest gender difference in dreams.

The animal percent (total animals divided by the total number of characters of any kind) provides another useful example because it has great variation by age and culture. The animal percent is as high as 30 to 40 in young children, but only 6 for men and 4 for women in American society. It is as high as 30 in some hunting and gathering societies, and exceeds the American level in all of the dream samples collected by anthropologists in many different parts of the world (Domhoff, 1996).

Aggression, defined as a desire, intention, or action to annoy or harm some other character, is the most frequent type of social interaction in dreams, occurring at least once in 47 percent of men's dreams and 44 percent of women's. It is the element that varies the most by age, gender, culture, and personality. Measures of aggression also show the largest difference between dreams written down at home and dreams collected in the sleep laboratory (Domhoff & Schneider, 1999). The A/C ratio (total aggressions divided by total characters) yields a rate of aggression per character that is useful because it controls for both word length and the total number of characters in the sample. Physical aggression percent (physical aggressions divided by physical plus nonphysical aggressions) reveals the second-largest gender difference and declines from the teenage years into adulthood, according to both cross-sectional and longitudinal studies. It is one of the few content indicators that changes after young adulthood is reached, which suggests there is a general consistency in dream content that is not expected by any dream theory (Domhoff, 1996, Chapters 5 and 7).

One of the few differences between the dreams of mental patients and normal individuals is revealed by indicators concerning friendliness. First, patients have a lower friends percent (total known characters divided by all human characters); their dreams are disproportionately populated by family members and strangers. Second, they have a lower "F/C ratio" (friendly interactions divided by total characters), which is not the same thing as the "friends percent" because there may or may not be friendly interactions with friends, and there can be friendly interactions with family members and strangers (Domhoff, 1996; Maharaj, 1997).

Anthropologist Barbara Tedlock (1991), in a paper reprinted in this book, asserts that content analysis is not useful in anthropological studies, because the dream reports cannot be taken at face value. Tedlock is right that there are many obstacles to collecting adequate samples in small traditional societies, but the many good samples gathered by anthropologists nonetheless produce interesting results in unpublished studies by Hall that are reported by Domhoff (1996,Chapter 6). In addition to similarities on the male/female percent, these samples are similar to the American norms in that there are more familiar than unfamiliar characters, more aggression than friendliness, more misfortune than good fortune, and more negative than positive emotions. At the same time, there are also content indicators that show differences from the American norms and variation from culture to culture. For example, there are wide variations in aggression, especially in the physical aggression percent, as shown in Table 1. This table also shows that Dutch and Swiss samples have lower physical aggression percents than the American norms. The findings in Thomas Gregor's (1981) detailed content analysis of Mehinaku dreams -- which appears as a chapter in this book -- also stand as a refutation to Tedlock's claims.

Table 1. Physical Aggression Percents
Society    Males    Females  
Yir Yiront92n/a
United States5034
French Canadiansn/a31

Tedlock's comments are puzzling because the article and book she cites as support for her claim (Hall, 1951; Hall, 1953c) present no cross-cultural findings. These two publications appeared before Hall had done any cross-cultural content analyses. This suggests that Tedlock's strong emphasis on cultural variation makes it difficult for her to contemplate that a content analysis system devised to study dream reports provided by Americans could prove useful cross-culturally. She therefore presents an unnecessary choice between studying dream content or cultural beliefs relating to dreams. She also prejudges a hypothesis offered by Gregor (1981, p. 389) on the basis of his empirical work with the Mehinaku: "with additional cross-cultural data it may be possible to show that the dream experience is less variant than other aspects of culture."

As surprising and interesting as many of these gender, age, and cross-cultural findings are, they are perhaps surpassed by the findings with lengthy dream journals from individuals. Although the early studies of dream journals pioneered by Wilhelm Stekel, Havelock Ellis, and Carl Jung produced many interesting findings, the great consistency in dream content over years and decades went undetected until journals were studied with the Hall/Van de Castle system. Since the results are similar in each case, despite the many different reasons why the dreamers kept personal journals, they are considered to be robust findings by those who advocate the study of archival records and other "unobtrusive" sources of data (Webb, Campbell, Schwartz, Sechrest, & Grove, 1981). Furthermore, most dream content is found to be continuous with waking thought in those cases where personal information can be obtained about the dreamer (Bell & Hall, 1971; Domhoff, 1996; Hall & Lind, 1970; Hall & Nordby, 1972).

Both of these points can be demonstrated with a natural scientist's dream journal from the summer of 1939. It was purchased from an antiquarian book dealer by psychiatrist J. Allan Hobson (1988) and used by him to illustrate the alleged bizarreness of the form and structure in dreams. Even with only 187 dreams that are over 50 words or more in length, the content is strikingly consistent over a three-month period, as shown in Table 2. In addition, a graphic profile of how he compares to the male norms is provided in Figure 1. This bar graph is based on the h statistic, which is roughly twice the size of a percentage difference. It shows that he is unaggressive, less likely to be involved in friendly interactions, more likely to be in familiar settings, and high on animal percent. These deviations from the norms fit with facts about his life found in an obituary in a scientific publication (Domhoff, 1996, pp. 147-150). These dreams can be found on DreamBank.net under the pseudonym "Natural Scientist."

Table 2. The consistency of the dreams of "The Natural Scientist."
       Four subsets       Three subsets       Two subsets       TOTAL  
     (each n=46)     (each n=62)     (each n=93)     (n=187)  
 Male/Female Percent 65%76%66%74% 66%74%71% 70%70% 70%
 Familiar Characters 42%34%41%34% 39%38%36% 38%37% 38%
 Animal Percent 15%13%08%09% 14%10%10% 14%09% 11%
 Group Percent 21%29%14%23% 26%19%21% 25%19% 22%
 Characters per dream 3.062.763.32 3.083.02 3.08
 D-involved Aggression  57%45%67%65% 55%48%67% 51%66% 59%
 Victimization Percent 58%67%60%72% 53%64%71% 62%68% 65%
 Physical Aggression 33%50%38%49% 34%52%45% 42%45% 45%
 A/C index . .15.12.25 .15.21 .18
 Aggressions per dream . .47.34.82 .46.62 .55
 D-involved Friendliness 71%65%70%73% 63%64%79% 68%72% 70%
 Befriender Percent 50%40%38%68% 53%29%63% 44%57% 52%
 F/C index . .13.16.18 .14.18 .16
 Friendliness per dream . .39.45.61 .43.54 .49
 Indoor Settings 53%37%40%42% 52%33%44% 45%41% 43%
 Familiar Settings 75%76%68%71% 73%76%71% 76%70% 73%
 Negative Emotions 89%77%70%73% 86%71%76% 84%72% 79%
 D-involved Emotions 83%46%80%60% 76%64%62% 68%68% 68%
 Emotions per dream . .34.23.34 .33.27 .30

Figure 1. h-profile for "The Natural Scientist."
Note: h is a statistic that corrects for the fact that standard deviations cannot be calculated with percentage data. h is roughly twice as large as a percentage difference. An h of 0-.20 is small, an h of .21-.40 is moderate, and an h over .40 is large.

A comparison of the extant dreams of Freud and Jung led to interesting patterns that went unnoticed despite the many commentaries on them. Although there are only 28 dream reports from Freud and 31 from Jung, which is far too view to be conclusive, earlier studies of dozens of dream journals by Hall and his students show that major themes can be discovered with as few as 20 to 25 dreams (Hall, 1947). The most striking differences concern the nature of their friendly and aggressive interactions. Freud's friendly interactions are far more frequent with males, whereas Jung has friendly interactions with both males and females. In every instance where Jung is involved in a friendly interaction, he is the initiator, whereas Freud is the recipient of friendship in 8 of 11 friendly interactions. In the case of aggressions, Jung is a typical male in that his aggressions are with males, not females, whereas Freud has an aggression with 1 in every 4 female characters and no aggressions with males. There is evidence that the waking preferences of the two men are compatible with these findings, and that the differences in the patterns of friendly and aggressive interactions might shed light on why they could not sustain their close working relationship (Domhoff, 1996; Hall & Domhoff, 1968).

While the results of studies with dream journals are encouraging, there is far more that can be done with them. The ideal study would include a large number of dreams over a considerable span of years. Inferences from blind analyses would be tested by comparing them with information provided by both the dreamer and close friends of the dreamer. If inferences are denied by the dreamer, as sometimes has happened in past studies (Domhoff, 1996, Chapter 8), then the answers given by friends provide a way to resolve the disagreement. If the friends agree with the content analyst, then it can be inferred that the dreamer has a "blind spot" on the issue. If the friends agree with the dreamer, then the content analyst is probably wrong, an outcome that can be very valuable in improving a theory.

A dream journal that is useful for studies of this kind is now available at www.DreamBank.net. It contains over 3,300 dreams over a 22-year period by a woman, Barbara Sanders, who is now in her 50s. She has been interviewed over a two-day period about both her life and her dreams, and four of her female friends have been interviewed about those aspects of her life that appear in her dreams. The unpublished research on this series that has been completed to date demonstrates once again the astonishing consistency in dreams over long periods of time, which in this case is captured by a comparison of 125 dreams from the first half of the series with 125 from the second half.

Other unpublished work on the Barbara Sanders series shows that her dreams are continuous with her waking concerns, such as her interest in cats, of which she has several. Her dreams also portray her feelings about relatives and friends with striking accuracy. However, inferences based on a continuity principle have been disconfirmed on several topics, which opens up new avenues for adding depth and complexity to the understanding of dreams. For example, Barbara Sanders' cats are often underfed, lost, or deformed in her dreams, but not in waking life, where they are treated quite well. She rides horses fearlessly in her dreams, but is not a good rider and fears them in waking life. Her dreams give the impression that she probably learned to shoot guns when she was growing up, and likes them, but neither inference is correct.

Most studies using the Hall/Van de Castle system have focused on everyday dream reports. However, the system can be adapted to study the occasional highly memorable or impactful dream that has carryover effects into waking life and is of great interest to clinicians and humanists. Then, too, the Hall/Van de Castle norms provide a basis for commenting on the rarity of such dreams, as do results from studies of lengthy dream journals. Many of the elements in highly memorable dreams can be classified as "good fortunes," which are defined as good things that happen to the dream character as a result of fate, as opposed to intentional striving by the character or beneficial actions by other dream characters. Sudden recovery from injuries or illness, the possession of extraordinary abilities, the unexpected discovery of valuable objects, and arriving in a very bountiful environment are all examples of good fortunes.

Highly memorable dreams also seem to be characterized by particular patterns of Hall/Van de Castle indicators. This can be seen by looking at the patterns in the three types of "impactful" dreams -- transcendent, existential, and anxiety -- identified in studies using one-month diaries from college students; these three types were then compared to everyday dreams (Busink & Kuiken, 1996; Kuiken & Sikora, 1993). Although the phenomenologists who developed this typology do not think a standardized coding system can be useful in studying such dreams, their examples show that these dreams fit the following coding patterns: Transcendent dreams are characterized by unusual or distorted settings, famous or mythical characters, friendly interactions, good fortunes, success, and positive emotions. Existential dreams are more likely to involve deceased characters, major misfortune, thinking instead of physical activity, and intense feelings of sadness or confusion. Anxiety dreams contain unfamiliar settings, unknown characters or animals, physical aggression directed at the dreamer, physical activities instead of thinking, and intense feelings of apprehension. Ordinary dreams are far more likely to have familiar settings, familiar characters, nonphysical aggressions, minor misfortunes, and no emotions.

Fast content analysis via DreamBank.net

DreamBank.net is a resource for dream researchers that contains over 10,000 dreams from individuals and groups. Most of the dreams are in a few lengthy dream journals contributed by older adult women, but there are also sets of dreams from children, teenagers, college students, and blind people (Schneider & Domhoff, 1999). The search engine on the site makes it possible to do very fast content analyses of one or more of these sets of dreams, and to create new subsets that contain a specific element, such as "bridges," or a specific activity, such as "weddings," that might be of interest for making metaphoric studies of dream content. This Web site also provides private space where researchers can use the DreamBank search engine to analyze dream reports that are already in a digital format. A private space also can be reserved to make daily entries into a dream journal for those who want to use DreamBank to study their own dreams.

The phrases or strings of words that are entered into the DreamBank search engine obviously fulfill the first step in making a content analysis -- the creation of clearly defined categories that lead to the same results each time. The program also carries out the second step of a content analysis, which is to provide a frequency count. The frequencies can be turned into percentages and ratios in ways that will be explained shortly. Finally, normative findings can be obtained, because the dream reports that were used to create the Hall/Van de Castle norms are part of the DreamBank, and they can be searched at the same time the sample of interest is being searched.

These points can be demonstrated through a simple study that asks whether people's concern with cats and dogs can be predicted on the basis of their dreams. The program is put in "OR" mode so that it will find any instance of a cat OR a dog in a dream. Then two search expressions are entered: ^(cat|kitten|kitty|kittie)s?^ and ^(dog|doggy|doggie|puppy|puppie)s?^. Next the dreams of two cat lovers, Alta and Barbara Sanders, are selected for searching, along with those of a young girl, Melissa, who loves all animals, and one older woman, Arlie, who has little or no interest in animals. In addition, the 500 dreams from 100 college women that were used to create the female norms for the Hall/Van de Castle system are selected. The search produces the results that are presented in Table 3.

Table 3. Search results for "cat" and "dog" in the DreamBank
  ^(cat|kitten|kitty|kittie)s?^  ^(dog|doggy|doggie|puppy|puppie)s?^   ANY term  
Alta: a detailed dreamer (n=422)55 (13.0%)21 (5.0%)67 (15.9%)
Arlie: a middle-aged woman (n=212)1 (0.5%)3 (1.4%)4 (1.9%)
Barb Sanders (n=3116)176 (5.6%)105 (3.4%)257 (8.2%)
Female Norms (n=491)9 (1.8%)11 (2.2%)19 (3.9%)
Melissa: a young girl (n=67)9 (13.4%)11 (16.4%)18 (26.9%)
TOTAL (n=4308)250 (5.8%)151 (3.5%)365 (8.5%)

As can be seen in the table, the results reveal the number and percentage of dreams that contain at least one dog (left-hand column), at least one cat (center column), and at least one dog or cat (right-hand column). The right-hand column provides an indication of a general interest in pet animals. As expected, Arlie is below the female normative sample, and the other three are above it. The determination of the degree to which a person is concerned with cats as compared to dogs is accomplished by dividing the total number of dreams with at least one cat by the total number of dreams with at least one cat plus the total number with at least one dog. Alta and Barbara Sanders' high "cat percents" (72 and 63, respectively) reveal the fact that they much prefer cats to dogs, whereas young Melissa's cat percent of 45 rightly suggests that she is interested in both cats and dogs. The cat percent for the normative sample (45) shows that Alta and Barbara Sanders are well above the average woman on cat percent.

DreamSearch can be used in conjunction with Hall/Van de Castle codings in studying lengthy dream journals. The ideal starting point for such studies is the Barbara Sanders series, because 675 of the 3000 dreams already have been coded, and there is ample information available on her life due to the interviews with her and her friends. By entering the name "Howard," for example, all the dreams of her ex-husband can be studied for possible themes or metaphors in the context of the Hall/Van de Castle codings for those dreams. The name "Derek" can be entered to study her dreams about a man in whom she had a romantic interest when she was in her late 40s. The trajectory of her feelings can be compared with the codings for friendly, sexual, and aggressive interactions throughout the Derek subset.

Implications and future directions

The methods and findings presented in this chapter have two major implications for scholars in the humanities and related fields who want to contribute to the attempt to understand dreams, or to use ideas and methods concerning dreams as part of other projects. First, the best-known dream theories cannot be used with confidence to understand dream content. There are too many findings from content analysis that are not consistent with their claims. For example, dreams are more patterned and consistent over time than Hobson and his colleagues (1988; Hobson, Pace-Schott, & Stickgold, 2000; Hobson, Stickgold, & Pace-Schott, 1998) would expect, and more similar across cultures than neo-Freudians (Fromm, 1951) and most anthropologists might think. They are more consistent in late adulthood and more continuous with waking life than Jung (1974) would predict.

This skepticism about the best-known theories is reinforced by other types of systematic dream studies that are not discussed in this chapter. The developmental findings with children studied in sleep laboratories contradict Freud, Jung, and Hobson on key points (Foulkes, 1982; Foulkes, 1999; Foulkes, Hollifield, Sullivan, Bradley, & Terry, 1990). So do many other laboratory and experimental studies (Domhoff, 1999a; Domhoff, 2001; Foulkes, 1996). In the case of Freud, there is not a single hypothesis put forth by him on any topic related to dreams that receives empirical support in the dream literature (Domhoff, 2000b; Domhoff, 2000c; Fisher & Greenberg, 1977; Fisher & Greenberg, 1996). Freud and the other clinical theorists made an important contribution by suggesting there is psychological meaning in dreams, and they put forth many ideas that stimulated future research, but continuing adherence to their theories is not likely to advance the understanding of dreams.

Second, the findings presented in this chapter give pause about the use of traditional methods of dream analysis. The findings from content analysis add to the numerous studies casting doubt on claims based on either psychotherapy sessions or hypnosis, because they are often very different from what those methods lead us to expect (Domhoff, 2000a; Foulkes, 1999). Although symbolic interpretations might be useful someday if they are viewed as the product of figurative thought (Gibbs, 1994; Gibbs, 1999; Grady, Oakley, & Coulson, 1999), there is still much work to be done before it can be certain that waking figurative thinking is also present in dreaming.

What, then, are the future directions that theorizing about dreams might take? First, the developmental, content analysis, and laboratory findings are most consistent with an open-ended and loosely structured cognitive approach. Within that context, there is a wide range of opinion concerning the degree to which dream content is the product of coherent thought. Foulkes (1999), who argues that dreaming is the form that consciousness takes during sleep, concludes that dreaming, like consciousness, is a cognitive achievement that only gradually becomes more or less coherent. In his view, much of dream content is the product of seat-of-the-pants, one-thing-after-another type of thinking, rather than deeply motivated metaphoric thought, and some of it is not completely sensible. On the other hand, Lakoff (1993; 1997), one of the founders of cognitive linguistics, argues that dreams are often structured by conceptual metaphors.

Within a cognitive framework, future studies of dream content might start with the idea that dreams reflect concerns and interests, and that they are generally consistent over time and continuous with waking thoughts. Such a theory can encompass wish fulfillment dreams as well as the realistic post-traumatic stress disorder nightmares that often occur after deadly accidents, shocking assaults, or the witnessing of murders. Then the deviations and discrepancies from consistency and continuity could be used to nuance and modify the theory. Why, for example, would Barbara Sanders' dreams so accurately depict her feelings and concerns in regard to the key people in her life, but so inaccurately portray the health of her cats, her feelings about horses, and her proficiency with guns? Why do some of those who are successful in ending their dependency on tobacco or alcohol nonetheless have upsetting dreams in which they find themselves back to their old habits and feeling terrible about it? More generally, why do some people dream of former lovers they are no longer attracted to in waking life?

Finally, it seems critical to build on the insights of cognitive linguists to see if there is sense in those elements in dreams that are not captured by the content indicators, or that do not seem to relate to the concerns, interests, and emotional preoccupations of the dreamer. Once again, it is the Barbara Sanders series that provides a good starting point in a search for conceptual metaphors and other forms of figurative thought in dreams. It is through such detailed studies that it will be possible to see if all aspects of dreams "make sense," as Freud (1900), Jung (1974), and Hall (1953a; 1953b) would maintain, while at the same time disagreeing among themselves about what that sense is. If such a search fails, then it may be that some aspects of dream content make no sense, as Foulkes (1985; 1999) and Hobson (1988; Hobson et al., 2000) would argue, even while they disagree about the origins of dreams and the point where sense fades into nonsense.


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

Boss, M. (1958). The Analysis of Dreams. New York: Philosophical Library.

Boss, M. (1977). I Dreamt Last Night. New York: Gardner Press.

Busink, R., & Kuiken, D. (1996). Identifying types of impactful dreams: A replication. Dreaming, 6, 97-119.

Domhoff, G. W. (1996). Finding Meaning in Dreams: A Quantitative Approach. New York: Plenum Publishing Co.

Domhoff, G. W. (1999a). Drawing theoretical implications from descriptive empirical findings on dream content. Dreaming, 9, 201-210.

Domhoff, G. W. (1999b). New directions in the study of dream content using the Hall and Van de Castle coding system. Dreaming, 9, 115-137.

Domhoff, G. W. (2000a). Methods and measures for the study of dream content. In M. Kryger, T. Roth, & W. Dement (Eds.), Principles and Practices of Sleep Medicine(Third ed., pp. 463-471). Philadelphia: W. B. Saunders.

Domhoff, G. W. (2000b). The misinterpretation of dreams. American Scientist, 88(March-April), 175-178.

Domhoff, G. W. (2000c). Moving dream theory beyond Freud and Jung. Paper presented at the Conference on "Beyond Freud and Jung? The Interpretation of Dreams, Religion, and Culture." September 23, Graduate Theological Union, Berkeley, CA.

Domhoff, G. W. (2001). A new neurocognitive theory of dreams. Dreaming, 11, 13-33.

Domhoff, G. W., & Schneider, A. (1999). Much ado about very little: The small effect sizes when home and laboratory collected dreams are compared. Dreaming, 9, 139-151.

Downing, J., & Marmorstein, R. (1973). Dreams and nightmares: A book of Gestalt therapy sessions. New York: Harper & Row.

Fisher, S., & Greenberg, R. (1977). The Scientific Credibility of Freud's Theories and Therapy. New York: Basic Books.

Fisher, S., & Greenberg, R. (1996). Freud Scientifically Appraised. New York: John Wiley.

Foulkes, D. (1982). Children's Dreams. New York: Wiley.

Foulkes, D. (1985). Dreaming: A Cognitive-Psychological Analysis. Hillsdale, NJ: Lawrence Erlbaum.

Foulkes, D. (1996). Dream research: 1953-1993. Sleep, 19, 609-624.

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

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.

Freud, S. (1900). The Interpretation of Dreams(J. Crick, Trans.). London: Oxford University Press.

Fromm, E. (1951). The Forgotten Language. New York: Grove Press.

Gibbs, R. (1994). The poetics of mind: Figurative thought, language, and understanding. New York: Cambridge University Press.

Gibbs, R. (1999). Speaking and thinking with metonymy. In K. Panther & G. Radden (Eds.), Metonymy in Language and Thought(pp. 61-75). Philadelphia: John Benjamins.

Grady, J., Oakley, T., & Coulson, S. (1999). Blending and metaphor. In R. Gibbs & G. Steen (Eds.), Metaphor In Cognitive Linguistics(pp. 101-124). Philadelphia: John Benjamins.

Gregor, T. (1981). A content analysis of Mehinaku dreams. Ethos, 9, 353-390.

Hall, C. (1947). Diagnosing personality by the analysis of dreams. Journal of Abnormal and Social Psychology, 42, 68-79.

Hall, C. (1951). What people dream about. Scientific American, 184, 60-63.

Hall, C. (1953a). A cognitive theory of dream symbols. Journal of General Psychology, 48, 169-186.

Hall, C. (1953b). A cognitive theory of dreams. Journal of General Psychology, 49, 273-282.

Hall, C. (1953c). The meaning of dreams. New York: McGraw-Hill.

Hall, C. (1969). Content analysis of dreams: Categories, units, and norms. In G. Gerbner (Ed.), The Analysis of Communication Content(pp. 147-158). New York: Wiley.

Hall, C. (1984). A ubiquitous sex difference in dreams, revisited. Journal of Personality and Social Psychology, 46, 1109-1117.

Hall, C., & Domhoff, G.W. (1963). A ubiquitous sex difference in dreams. Journal of Abnormal and Social Psychology, 66, 278-280.

Hall, C., & Domhoff, G. W. (1968). The dreams of Freud and Jung. Psychology Today, 2, 42-45.

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

Hall, C., & Nordby, V. (1972). The individual and his dreams. New York: New American Library.

Hall, C., & Van de Castle, R. (1966). The Content Analysis of Dreams. New York: Appleton-Century-Crofts.

Hobson, J. (1988). The Dreaming Brain. New York: Basic Books.

Hobson, J., Pace-Schott, E., & Stickgold, R. (2000). Dreaming and the brain: Towards a cognitive neuroscience of conscious states. Behavioral and Brain Sciences, 23(6), in press.

Hobson, J., Stickgold, R., & Pace-Schott, E. (1998). The neuropsychology of REM sleep dreaming. NeuroReport, 9, R1-R14.

Jung, C. (1974). Dreams. Princeton, NJ: Princeton University Press.

Kuiken, D., & Sikora, S. (1993). The impact of dreams on waking thoughts and feelings. In E. Alan Moffitt, E. Milton Kramer, & et al. (Eds.), The functions of dreaming.(pp. x, 610). Albany, NY, USA: State University of New York Press.

Lakoff, G. (1993). How metaphor structures dreams. Dreaming, 3, 77-98.

Lakoff, G. (1997). How unconscious metaphorical thought shapes dreams. In D. Stein (Ed.), Cognitive Science and the Unconscious(pp. 89-120). Washington: American Psychiatric Press.

Loftus, E., & Ketcham, K. (1994). The myth of repressed memory. New York: St. Martin's Press.

Maharaj, N. (1997). An Investigation into the Content and Structure of Schizophrenic Dreams. Doctoral dissertation, Leiden University, Leiden, The Netherlands.

Mazzoni, G., Loftus, E., Seitz, A., & Lynn, S. (1999). Changing beliefs and memories through dream interpretation. Applied Cognitive Psychology, 13(2), 125-144.

Moss, C. S. (Ed.). (1967). The Hypnotic Investigation of Dreams. New York: Wiley & Sons.

Ofshe, R., & Watters, E. (1994). Making monsters: False memories, psychotherapy, and sexual hysteria. New York: Charles Scribner's Sons.

Schneider, A., & Domhoff, G. W. (1995, updated 2000). The Quantitative Study of Dreams[Web site]. Available: http://www.dreamresearch.net/.

Schneider, A., & Domhoff, G. W. (1999, updated 2000). DreamBank[Web site]. Available: http://www.DreamBank.net/.

Spanos, N. (1996). Multiple Identities and False Memories: A Socio-cognitive Approach. Washington: American Psychological Association.

Tedlock, B. (1991). The New Anthropology of Dreaming. Dreaming, 1, 161-178.

Tsai, A., Loftus, E., & Polage, D. (2000). Current directions in false-memory research. In E. David F. Bjorklund & et al. (Eds.), False-memory creation in children and adults: Theory, research, and implications.(pp. x, 254). Mahwah, NJ, US: Lawrence Erlbaum Associates, Inc., Publishers.

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.

Webb, E., Campbell, D., Schwartz, R., Sechrest, L., & Grove, J. (1981). Nonreactive Measures in the Social Sciences(2nd ed.). Chicago: Rand McNally.

Go back to the Dream Library index.
dreamresearch.net home page dreamresearch.net contact info