Ernest Hartmann, Robert Basile
Tufts University School of Medicine, Newton Wellesley Hospital
Verkauf Graduate School of Psychology, Yeshiva University
There is a general consensus, based on anecdotal and clinical reports, that stress and trauma have an effect on dreams, but this has been difficult to study systematically. We have presented evidence that the CI Intensity Score, measuring the intensity of the dream's central image, is significantly higher for students who report any abuse than those who do not. Also, CI Intensity was higher in subjects who had recently experienced trauma than in control groups. 1, 2 A systematic study should ideally involve a study of dreams before and after trauma within individuals who have been routinely recording their dreams. It is not easy to find such dream recorders who also happen to experience an acute trauma and are willing to share their dreams.
We have suggested that the events of 9/11/01 represented a traumatic event for everyone living in the USA. Thus, anyone who has regularly been recording dreams over the past years would be appropriate for the study of dreams before and after trauma.
Methods
We were able to obtain complete dream series—the last ten dreams recorded before 9/11/01 and the first ten dreams recorded after that date—from sixteen regular dream-recorders living in various parts of the USA. A total of 324 dreams from these sixteen dreamers were assigned random numbers and scored blind for CI intensity. Fifty dreams were scored by two experienced scorers, showing inter-rater r = .76. Each dream was also scored for what emotion might be pictured by the central image of the dream, dream length, and “dream-likeness.” In addition, each dream was scored on three ad hoc scales (scored from 0 to 3) estimating dream content involving a) attacks of any kind, b) buildings resembling the World Trade Center or the Pentagon and c) airplanes.
After the code was broken, a pre-9/11 mean and a post-9/11 mean was obtained for each subject. The pre- and post-means were compared using a t-test for correlated means.
Results
Before | After | Difference | t | p | |
| CI Intensity | 1.19 | 1.35 | 0.16 | 3.21 | .003 (one-tailed) |
| Length (text lines) | 11.64 | 11.55 | -0.09 | 0.09 | NS |
| Dream-likeness | 4.5 | 4.54 | 0.04 | 0.47 | NS |
| Attacks | 0.048 | 0.135 | 0.087 | 1.74 | NS |
| Buildings | 0.069 | 0.122 | 0.053 | 0.95 | NS |
| Airplanes | 0.012 | 0 | -0.012 | 1.46 | NS |
There was a clearly significant increase in CI intensity score after 9/11. There was no significant increase in dream length, “dream-likeness” or the three ad hoc content categories. Results on the emotion thought to be pictured showed a shift towards more “fear/terror” and “vulnerability/helplessness.” However, these emotions are always frequent in all samples, and the shift did not reach statistical significance.
Discussion
In this group of regular dream-recorders, the effect of 9/11/01 was to increase dream image intensity without altering other measured aspects of dreams. Since they were selected only on the basis that they regularly recorded their dreams, we could tentatively generalize to the entire US population and suggest that the dream image intensity probably increased for the population. (Our experience, based on interviews and questionnaires, is that dream-recorders are unusual only in “sensitivity” and in interest in dreams, daydreams and mental functioning in general.)
The present results, consistent with past studies cited, suggest that the intensity of dream image may be considered a measure of the emotional arousal of the dreamer.
1 Hartmann, E., Zborowski, M., Rosen, R., Grace, N. (2001). Contextualizing images in dreams: more intense after abuse and trauma. Dreaming , 11: 115-126
2 Hartmann, E. (1998) Dreams and Nightmares . New York: Plenum Press and (2001) New York: Perseus Books.