This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Dreaming, Vol. 1, No.4, 1991

The Effects of Dream Length on the Relationship
Between Primary Process in Dreams and Creativity
Glenn Livingston l and Ross Levin l ,2

The effects of dream length on the relationship between primaty process in dreams and a measure of creativity unconfounded by IQ was investigated in a sample of 93 graduate students using the Auld, Goldenberg, & Weiss (1968) Scale of Primary Process Thought
(SPPT) and a modified Wallach-Kogan (1965) creativity batte/yo Consistent with previous research, total and mean primary process were found to correlate significantly with creativity (r = .28, P < .01 and r = .23, P < .05, respectively). Both significant relationships disappeared, however, once the effects of dream length were partialled out, confirming Wood, Sebba, & Domino's (1989-90) contention that this relationship may be artifactual. It is suggested that dream length as an individual difference in and of itself may thus be a more fruitful variable to examine in future research investigating the relationship between creativity and dreams.
KEY WORDS: dreaming; dream length; primary process; creativity.

The contention that creativity and dreaming may reflect similar psychological processes has long been maintained by both the lay public and philosophers alike.
In support of this, both the anecdotal and empirical psychological literature contain numerous references suggesting both a direct and analogical relationship between dreams and creativity (Adelson, 1960; Dave, 1978; Domino, 1976, 1982; Sladeczek
& Domino, 1985; Dreistadt, 1971; Giovacchini, 1966; Krippner, 1981; Lewis, 1984;
Palombo, 1983, 1984; Sylvia, Clark, & Monroe, 1978; Wood, Sebba, & Domino,
1989-90). Some anecdotal examples of creative achievements ascribed directly to dream material include the prelude to one of Wagner's famous operas (Dreistadt,
1971), inventions such as flaked cereals (Sylvia et aI., 1978) and the sewing machine needle (Domino, 1982), and musical works of Beethoven, Tartini, Mozart, and
Schumann (Dreistadt, 1971; Sylvia et aI., 1978; Dave, 1978). Furthermore, such poets and writers as Henry James, Heinrich Heine, William Blake, Lev Tolstoy,
Nathaniel Hawthorne, Mary Shelley, Rimbaud, Edgar Allen Poe, Robert Louis
IFerkauf Graduate School of Psychology, Yeshiva University, Bronx, New York.
2Correspolldence should be directed to Ross Levin, Ph.D., Ferkauf Graduate School of Psychology,
Albert Einstein College of Medicine Campus, 1300 Morris Park Avenue, Bronx, New York 10461.
301
1053·0797/91/1200·0301$06.50/1 © 1991 Association lor the Study of Dreams

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

302

Livingston and Levin

Stevenson and Samuel Coleridge have all been noted to have drawn extensively from their dreams as raw material for their creative output (Dreistadt, 1971; Fiss,
1986; Hock, 1960; Krippner, 1981; MacKenzie, 1965).
Conceptual linkage between creativity and dreaming also exist. Psychoanalytic theories, in particular (Arieti, 1976; Bellak, 1967; Erikson, 1954; Freud, 1910, 1924;
Kris, 1952; Kubie, 1958) have long posited a continuity between dream imagery and creativity. According to these writers, both processes invoke similar underlying psychic mechanisms, namely heightened access to what Freud (1900) first termed primary process mentation. For instance, lung (1966) felt that dreams triggered archetypal images which could be enhanced and refined towards creative ends (Dallet, 1973; Woodman, 1981). Adler (1929) viewed the dream as representative of the dreamer's orientation to everyday life. More recently, Ullman (1962) hypothesized that dreams motivated the dreamer toward the creative solution of real life problems. In his paper, The Unconscious, Freud (1915) posited that, in artistic persons, unconscious thoughts (organized to a large degree via primary process thinking) were more acceptable to the ego (Woodman, 1981). Creative individuals were also thought to enjoy an increased propensity for sublimation. Although he believed these creative processes occurred unconsciously through primary process mechanisms, Freud was prevented from postulating a direct link between creativity and dreams because of his conceptualization of the primary process as a primitive and static mode of cognition which was "illogical, undersocialized, and affect-predominant" (Levin, Galin, & Zywiak, 1990, p. 6). Freud reasoned that primary process mentation operated in accordance with the pleasure principle, thus seeking immediate discharge of pent-up tension without regard for object-specificity.
Later revisions in the understanding of the primary process (Noy, 1969;
Fosshage, 1983, 1986; Suler, 1980) have challenged many of Freud's (1900) original notions. Rather than conceptualizing primary process as an infantile mode of thought, these theorists view the primary process as a dynamic and progressive cognitive system which serves an important developmental function, distinct but not subordinate to the secondary process. Thus, Suler (1980) notes that primary process is metaphoric thinking and underlies all cognitive output which eschew rationality
(e.g., dreams, fantasy, psychotic symptomatology). These views also fit well with recently formulated models of creativity (Rothenberg, 1990) and dream function
(Breger, 1967; Fosshage, 1983). For example, both Breger and Greenberg (1987) have written extensively about creative problem solving in dreams from an information processing model.
Given this, a relationship between access to primary process in dream reports and measures of creativity would be expected. However, despite the conceptual linkage, there has been little direct empirical investigation on the relationship between creativity and dreaming in general, and only a handful of studies exist linking creativity to the degree of primary process organization of dream reports in particular. A brief review of this research follows.
Utilizing a scale of primary process thinking in dream reports developed by
Auld, Goldenberg, & Weiss (SPPT) (1968), Domino (1976) examined the association between creative achievement and primary process thinking in the dreams

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Primary Process, Dreams, and Creativity

303

of 76 high school students using a contrasted groups design. Creativity groups were formed via teacher nominations (based upon creative production) and psychometric criteria (Remote Associates Test and Alternative Uses Test). Groups were matched for pertinent demographic variables. Domino found significantly more primary process thinking in creative students' dreams than in the reports of the matched controls.
In a similar study, Siadeczek and Domino (1985) demonstrated that the degree to which dream reports deviate from normal waking experience was significantly related to subjects' creativity scores. High-creative subjects (selected psychometrically through performance on a combination of convergent and divergent creativity tasks) had dreams with significantly more primary process as measured by several scales: dream distortion, regressive dream content, and visual mentation. Both of the above studies used measures of creativity which have been shown to share significant amounts of variance with conventional intelligence (Wallach,
1970; Livingston, 1989). Noting this possible confound, Livingston examined the relationship between primary process in dream reports and creativity in a group of
55 subjects drawn largely from a university population, using purely ideational fluency measures which have been documented to be largely independent from intelligence (Wallach, 1970). Primary process in dreams was found to be significantly correlated with creativity (r = .53, p < .02), but only for a small sample (n = 14) which reported significantly higher dream recall.
In an important recent paper, Wood et aI., (1989-90) contend that the hitherto established relationship between creativity and dream-like mentation may be confounded by such extraneous factors as verbal intelligence and dream report length.
They based this hypothesis on two lines of converging evidence: first, the documented association between intelligence and many traditional creativity tests, most notably those that measure convergent thinking (Wallach, 1970), and; Antrobus'
(1983) data suggesting a highly positive correlation between dream bizarreness and distortion with dream length. To test this presumption, Wood et al. administered a battery of four creativity tests and a measure of verbal intelligence to 126 college students who were also asked to keep a dream diary for 14 nights. Dreams were scored for primary process using the Auld et al. (1968) scale and dream word length measures were tabulated. The results supported the authors' contention that the association between waking creativity and dream mentation may be due to a failure to control for verbal intelligence and report length. Thus, initially significant relationships between dream primary process and creativity were eliminated when both dream length and intelligence were controlled by partial correlations. Significantly, dream report length correlated quite highly (r = .54) with the Auld et al. scale.
The authors conclude that "although there is a modest relation between scores on certain creativity tests and bizarre dream reports, the relation does not prove what earlier studies have assumed . . . the higher dream bizarreness scores of verbally intelligent people may simply reflect the fact that such people write longer dream reports" (p. 11). Significantly, however, it should be noted that the authors again used creativity measures which may be confounded with intelligence.

304

Livingston and Levin

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Thus, the purpose of the present study is to replicate Wood et al.'s (1989-90) basic findings utilizing an unconfounded measure of creativity, namely ideational fluency. It is predicted that although primary process will correlate with total dream length, the magnitude of this relationship will be attenuated given the nature of the creativity measure.

METHOD

Subjects
106 persons (49 male, 57 female) from several different universities located in
New York City and New Jersey served as subjects for the larger study. Age ranged from 22 to 65 with a mean of 26.5 and race was predominantly (but not exclusively)
White non-Hispanic. All subjects were enrolled in a graduate program with a major other than psychology, and none had any previous exposure to psychological tests.
Subjects were solicited via local advertisements, and through the use of a data base of persons who had previously indicated their interest in participating in research at
. a facility in Fort Lee, N.J. All were asked to indicate their average dream recall frequency on an initial self-report screening questionnaire, and those specifying less than one per week were excluded from the study. 13 subjects failed to complete the creativity measures leaving a final pool of 93 (51 female, 42 male) for this investigation. Subjects were paid thirty dollars for their participation in the research.
Materials and Procedure
Two meetings with each subject were required to complete the measures. The first meeting was conducted in groups of 10 to 20 by the first author during which introductory and screening measures were administered as part of a larger study investigating dreams and creativity. At this meeting, subjects were instructed to keep dream diaries for one week before returning for a second appointment, also with the first author, which was conducted individually. At that time, subjects were placed in a small room with a desk, a VCR, and a television. Subjects were told that the instructions for this part of the study would be on VCR tape, and that it would take approximately 90 minutes. The tape contained the instructions for a modified Wallach-Kogan (1965) creativity battery. The Wallach-Kogan battery consist of verbal and visual stimuli to which several of Guilford's (1959) original divergent tests are applied (e.g., Alternate Uses, Consequences, Plot Titles, etc.). It is traditionally scored only ideational fluency and construct validity and reliability for this measure are well documented (Wallach, 1970). For this investigation, the battery was modified in the following manner: 1) in the interest of time, only four verbal items and no visual items were utilized (newspaper, cork, shoe, and chair), and; 2) items were presented three-dimensionally given recent research which suggests that creativity is best enhanced by this mode of presentation (Fu, Kelso, &
Moran, 1984). Alternate uses was the task requirement for all objects.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Primary Process, Dreams, ami Creativity

305

Dreams were scored blindly by the first author and a research assistant trained in the use of the SPPT (Auld et aI., 1968). The SPPT is a seven-point rating scale ranging from extremely reality oriented dream reports to highly bizarre and logically implausible dream reports. For example, a dream depicting logical, reality-based events (e.g., "I was eating in a restaurant") would receive a 1; those dreams which include illogical or implausible events yet retain a semblance of a coherent dream narrative would receive a 4; while those dreams predominated by highly bizarre and logically implausible mentation would receive a 7. Previous research has demonstrated adequate reliability and construct validity (Domino, 1976; Levin & Livingston, 1991). Inter-rater reliability for the SPPT in the present study was .82 for a random sample of 100 dreams.

RESULTS
Age and Sex Effects

Between group block t-tests were conducted on the main variables to check for sex differences. Additionally, correlations were computed between age and the above variables. As no significant relationships were found, all data were pooled for subsequent analyses.
SPPT ami Creativity
The modified Wallach-Kogan (1965) battery correlated significantly with both the mean (r = .28,p < .01) and total (r = .23,p < .05) amount of primary process manifestations in the dream protocols.
Given Wood et a1.'s (1989-90) findings suggesting that these relationships may be an artifact of uncontrolled dream length, correlations between this variable (average and total) and both primary process and creativity were computed. Mean dream word length in this investigation was 74.9, with a range of 10 to 366 and a
SD of 50.1. Consistent with Wood et a!., dream length was highly correlated with primary process in dreams and creativity (see Table 1).
Thus, separate partial correlation coefficients were computed, correcting for the effects of both mean and total dream lengths, respectively. As a result, all significant relationships between primary process in dreams and creativity disappeared once the effects of dream length were controlled for (r = .08, mean primary process and r == .16 total primary process for mean dream length; r = .12, mean primary process and r := .04, total primary process for total dream length).

DISCUSSION
Consistent with both previous research (Domino, 1976; Siadezcek & Domino,
1985; Livingston, 1989; Wood et a!., 1989-90) as well as theoretical formulations

306

Livingston ami Ll!vin
Table L Simple correlations between creativity measures, primary process measures, lolal dream length, and average dream length TP

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

AL
TL

MP

w

.53**
.86**

.59"
.38*"

.24'
.29*

*p < .05, ** p < .01, two-Iailed, W
Modified Wallach-Kogan (1965) creativity battery, AL = Average dream length, TL '" Total dream length, MP '" Mean primary process in dream protocol, TP =
Tolal primary process in dream protocol. =

(e.g., Freud, 1910; Jung, 1966), primary process manifestations in dream reports were significantly associated with creativity, confirming our experimental hypothesis.
Furthermore, the relationship between primary process in dreams and the modified
Wallach-Kogan (1965) battery would appear to suggest that individuals who recall bizarre dreams have greater quantities of ideational material available to them for creative production. This finding is of particular interest given this measure's independence from verbal intelligence.
These findings are mitigated by several factors. First, only a relatively small percentage of variance was accounted for (no more than 9%) by the relationship supported herein. In this sense, it should be noted that the magnitude of the correlation was very similar to those reported by Wood et al. (1989-90). Relationships of such magnitude, while useful for indicating possible paths for future research, are less clinically meaningful.
The second and perhaps most important factor mitigating the findings of this study are the result of the partial correlation analyses correcting for dream length.
The finding that all significant relationships between creativity and primary process in dreams disappeared when word length was controlled is consistent with Wood et a1.'s (1989-90) contention that previously reported relationships between creativity and primary process in dream reports may be an artifact of dream productivity.
An alternative possibility to account for the contrasting results of this study
(as well as Wood et aI., 1989-90) with earlier investigations (Domino, 1976; Sladeczek & Domino, 1985) may lie in the nature of the creativity measure utilized. It has been noted that the dimensionality of creativity has largely been ignored (Hocevar, 1981), with most experimenters operating on the assumption that creativity is a unitary entity which can be measured equally well by any of a number of tests.
It is quite likely that the term "creativity" may need to be replaced with more specific construct labels such as ideational fluency (Guilord, 1959), adaptive regression (Holt, 1977; Kris, 1952) and homospatial thinking (Rothenberg, 1984), before the confusion about the various relationships between creativity and other constructs can be resolved.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Primary Process, Dreams, ami Creaaivity

307

However, the lack of a significant relationship between primary process in dreams and creativity once the effects of dream length are removed should not discourage the empirical pursuit of a relationship between dreams and creativity.
On the contrary, the consistent finding that dream length related strongly to creative ability raises considerable questions regarding the causal direction of this relationship. Rather than presuming that word length confounds dream bizarreness, it is equally plausible that the elucidation of bizarre events necessitate, by their very nature, greater narrative complexity and quantity. In this sense, the act of describing any unusual or idiosyncratic action, be it in the form of dream reports, literary prose, or journalism, would require lengthier narratives than descriptions of common, every day events. One way to investigate this hypothesis empirically might be to compare systematically descriptive waking reports of both mundane and bizarre pictures or videos in order to determine whether novelty impacts significantly on word length.
Dream length is an individual difference which has received only a minimum of attention in the literature (Antrobus, 1983) and certainly deserves more focus in future research. Perhaps it is the sum total of material processed out of awareness but available for conscious use (in other words the length of the recalled dream) which is the valuable element of dreams for creativity, and not merely dream material processed via the primary process. This explanation would be consistent with the information-processing position on dream function (e.g., Breger, 1967). Several other writers have emphasized unconscious processing as essential for creative achievement (e.g., Adler, 1929; Freud, 1915, 1924) without directly specifying that the material be processed by the primary process. Future research is warranted to determine exactly what it is about dream length that accounts for individual differences in creativity. In any case, by replicating previous research, the results of the present study underscore the importance of controlling for dream report length in the investigation of creativity and dream mentation.

ACKNOWLEDGMENTS
This research was completed in partial fulfillment of the first author's requirements for the doctoral degree in clinical psychology at Yeshiva University.

REFERENCES
Adelson, J. (1960). Creativity and the dream. Men'ill Palmer Quarterly, 6, 92-104.
Adler, A. (1929). Problems of neurosis. London: Kegan Paul.
Antrobus, J. S. (1983). REM and NREM sleep reports: Comparison of word frequencies by cognitive classes. Psychophysiology, 20, 562-568.
Arieti, S. (1976). Creativity: The magic synthesis. New York: Basic Books.
Auld, F., Goldenberg, G., & Weiss, J. (1968). Measurement of primary proccss thinking in dream reports.
Journal of Personality and Social Psychology, 8, 418-426.
Benak, L. (1967). Toward systematic consideration of the nature of the genesis of the creative process.
In D. P. Spence (Ed.), The broad scope of psychoanalysis. New York: Grune & Stratton.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

308

Livingston and Levin

Breger, L. (1967). Function of dreams. Joul71al of Abnormal Psychology, Monograph, 72, (5, pt. 2, Whole
No. 641).
Dallet, J. (1973). Theories of dream function. Psychological Bulletin, 79, 408-416.
Dave, R. (1968). Effects of hypnotically induced dreams on creative problem solving. Joul71al ofAbnormal
Psychology, 88, 293-302.
Domino, G. (1976). Primary process thinking in dream reports as related to creative achievement. Journal of Consulting and Clinical Psychology, 44, 929-932.
Domino, G. (1982). Attitudes towards dreams, sex differences, and creativity. Journal of Creative
Behavior, 16, 112-123.
Dreistadt, R. (1971). An analysis of how dreams are used in creative behavior. Psychology, 8, 24-50.
Erikson, E. (1954). The dream specimen of psychoanalysis. Journal of the American Psychoanalytic
Association, 2, 5-56.
Fiss, H. (1986). An empirical foundation for a self psychology of dreaming. JOlll71al of Mind and Behavior,
4, 161-192.
Fosshage, J. (1983). The psychological function of dreams: A revised psychoanalytic perspective.
Psychoanalysis and Contemporary Thought, 6, 641-669.
Fosshage, J. (1986). Dream interpretation: A comparative study. New York: Spectrum Publications.
Freud, S. (1900). The intetpretation of dreams. (J. Strachey, trans.) New York: Macmillan, 1933.
Fred, S. (1910). Leonardo da Vinci and a memory of his childhood. In J. Strachey (Ed. and trans.),
The standard edition of the complete psychological works of Sigmund Freud (Vol. 11). London:
Hogarth Press, 1953.
Freud, S. (1915). The unconscious. In Sigmund Freud, Collected papers, 4, 98-137. New York: Basic
Books.
Freud, S. (1924). A general introduction 10 psychoanalysis. New York: Liveright Publishing.
Fu, V., Kelso, G., & Moran, J. (1984). The effects of stimulus dimension and mode of exploration on original thinking in preschool children. Educational and PsycllOlogical Measurement, 44, 431-440.
Giovacehinni, P. L. (1966). Dreams and the creative process. Bri/ish Journal of Medical Psychology, 39,
105-115.
Greenberg, R. (1987). The dream problem and problems in dreams. In M. Glucksman & S. Warner
(Eds), Dreams in new perspective. New York: Human Sciences Press.
Guilford, J. P. (1959). The three faces of intellect. American Plychologist, 14, 469-479.
Hocevar, D. (1981). Measurement of creativity: Review and critique. JOl//'I1al of Personality Assessment,
5, 450-465.
Hock, A. (1960). Reason and genius. New York: Philosophical Library.
Holt, R. (1977). A method for assessing primary process manifestations and their control in Rorschach responses. In. M. Rickers (Ed.), Rorschach psychology (2nd Ed.), pp. 375-419. New York: Kreiger
Publishing Company.
Jung, C. (1966). On the relation of analytical psychology to poetry. In The spirit in man, art and literature.
New York: Pantheon.
Krippner, S. (1981). Access to hidden reserves of the unconscious through dreams in creative problem solving. Joul71al of Creative BehaviO/; 15, 11-23.
Kris, E. (1952). Psychoanalytic explorations in arl. New York: International Universities Press.
Kubie, L. S. (1958). Neurotic distortions of the creative process. New York: Noonday Press.
Levin, R., Galin, J., & Zywiak, B. (1991). Nightmares, boundaries, and creativity. Dreaming, 1, 63-74.
Levin, R., & Livingston, G. (1991). Concordance between two measures of dream distortion. Perceptual and Motor Skills, 72, 837-838.
Lewin, B. (1984). Remarks on creativity, imagery, and the dream. Journal of Nervous and Mental Disease,
149, 115-121.
Livingston, G. (1989). Creativity, the QAD, and primmy process ill dream reports. Unpublished master's thesis. Yeshiva University, New York.
MacKenzie, N. (1965). Dreams and dreaming. New York: Vanguard Press.
Noy, P. (1969). A revision of the psychoanalytic theory of the primary process. In/emational Joumal of
Psychoanalysis, 50, 155-178.
Palombo, S. (1983). The genius of the dream. American Journal of Plychoanalysis, 43, 301-353.
Palombo, S. (1984). The poet as dreamer. American Academy of Psychoanalysis, 12, 59-73.
Rothenberg, A. (1984). Creativity and psychotherapy. Psychoanalysis and Contemporary Thought, 3,
233-268.
Rothenberg, A. (1990). Creativity and madness. Baltimore: John Hopkins University Press.
Sladeczek, I., & Domino, G. (1985). Creativity, sleep, and primary process thinking in dreams. Joumal of Creative Behavior, 19, 38-46.
Suler, J. (1980). Primary process thinking and creativity. Psychological Bulletin, 88, 144-165.

This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Primary Process, Dreams, ami Creativity

309

Sylvia, W., Clark, P., & Monroe, L. (1978). Dream reports of subjects high and low in creative ability.
Journal of General Psychology, 99, 205-211.
UUman, M. (1962). Dreaming, life style, and physiology: A comment on Adler's view of the dream.
Journal of Individual Psychology, 18, 18-25.
Wallach, M. (1970). Creativity. In P. H. Mussen (Ed.), Carmichel's manual of child psychology (pp.
1211-1269). New York: Wiley.
Wallach, M., & Kogan, N. (1965). A new look at the creativity-intelligence distinction. Journal of
Personality, 33, 348-369.
Wood, J., Sebba, D., & Domino, G. (1989-90). Do creative people have more bizarre dreams? A reconsideration. Imagination, Cognition, and Personality, 9, 3-16.
Woodman, R. (1981). Creativity as a concept in personality theory. Journal of Creative Behavior, 15,
43-66.