Dreaming is an overwhelmingly visual experience for sighted people. About half of all dreams also have auditory sensations, but in two large-scale studies less than one percent had gustatory, olfactory, or tactual sensory references (Snyder, 1970; Zadra, Nielsen, & Donderi, 1998). Kerr (1993) suggests that the extremely visual nature of dreams may be why many people wonder if blind people even dream. This wonderment may explain why the presence or absence of visual imagery in the dreams of the blind has been of scientific interest since the early nineteenth century. A series of questionnaire and interview studies since that time have led to four empirical generalizations (Kirtley, 1975):
Studies of blind participants in sleep laboratories using awakenings during REM periods to collect dream reports have shown results similar to the questionnaire and interview studies (e.g., Amadeo & Gomez, 1966; Berger, Olley, & Oswald, 1962; Kerr, Foulkes, & Schmidt, 1982).
The substantive content in the dream reports of the blind has received less attention than the presence or absence of visual imagery. However, Kerr, Foulkes, and Schmidt (1982) found few differences among four congenitally blind, two adventitiously blind, and four sighted subjects between the ages of 19 and 32 with 16 REM and 10 NREM awakenings per subject over a period of eight weeks. This conclusion is based on ratings by the subjects themselves on 22 dimensions of the dream reports.
Studies by Kirtley (1975) and Kirtley and Sabo (1983, 1984), using dream journals from blind participants ranging in age from 20 to 56, generally found few differences from sighted subjects in their social interactions in dreams when compared to the Hall/Van de Castle norms, except that women were below the norms on two aggression indicators and above the norms on a friendly interactions indicator. However, the results from their content analyses have to be treated with caution because the various group comparisons did not control for the fact that some participants contributed as many as 39 dreams and some as few as 10.
The main purpose of the present study is to attempt to replicate the findings on sensory references and substantive dream content reviewed in the previous paragraphs. The paper also has two subsidiary purposes. First, it presents evidence that verbs of perception, such as "see" and "look," are used metaphorically by the blind in their dream reports in the same way they are used by sighted subjects (Matlock, 1988; Matlock & Sweetser, 1989; Sweetser, 1990). Second, it presents evidence to suggest that one of the primary waking concerns of the blind, traveling from place to place, may manifest itself in dreams through a higher percentage of locomotion/transportation dreams with a misfortune in them than is found for such dreams in the Hall/Van de Castle norms (Hurovitz, 1997).
The participants in the study were 15 congenitally and adventitiously blind men and women, ages 24 to 73, with 11 of them between the ages of 44 and 60 (M=46.2; SD=12.5). The ten women and five men in the study were chosen from a larger pool of 16 women and nine men gathered by Hurovitz (1997). They were selected because they contributed at least six recent dream reports that were not labeled as "earlier," "recurrent," or "childhood" dreams over the two-month period they recorded their dreams. In all, they provided 372 dream reports, 236 from the women, 136 from the men.
None of the 15 was presently or in the previous 12 months undergoing psychotherapy or under psychiatric care. The only slight exception to this generalization was the inclusion of a female army veteran who had been in rehabilitation training at a VA hospital, where she received a routine psychological evaluation upon entering the program and some brief optional adjustment counseling during her training.
Participants were recruited primarily through personal contacts by the first author. A few were found through an announcement of the project on the Internet. All were recruited through means which protected the privacy of potential participants and provided a maximum level of volunteerism. They were not paid for their participation. Most of them resided in Connecticut, but some lived in Pennsylvania, New York, Massachusetts, Maine, or California.
All of the participants had at least a high school education, and most were college graduates. Several had postgraduate training as well. They worked in or retired from a range of occupations, most of which are with social service agencies or industries for the blind. Table 1 presents basic demographic and occupation information on each participant along with the nature and degree of their blindness and the number of dream reports each person contributed.
The participants recorded their dreams on audio tape for a two-month period. They were asked to keep their tape recorders within easy reach of their beds and to record their dreams immediately upon awakening. For this population, it was much easier to reach for a small cassette recorder than to get out of bed and turn on a typewriter or computer. The first author picked up the tapes from participants who lived in Connecticut at the end of the collection period. The other participants mailed their tapes to the first author, who had all the dreams transcribed into a word processor and stored on a disk.
The dream reports were studied for any sensory references with DreamSearch, a new program designed for dream research by Adam Schneider and available on the web at www.dreambank.net (Schneider & Domhoff, 1999). The program can find any word, list of words, or phrases that are entered into the query box. It then provides a list of the dreams with at least one mention of the word, words, or phrases that are requested. It underlines the requested words or phrases and puts them in boldface when the dream reports are called up on the screen or printed. Thus, DreamSearch makes it feasible to do instantaneous retrievals and fast analyses in large databases that would not have been possible in the past.
For this particular study, all forms of the words "see," "saw," "watch," "look," and "notice" were used to provide a starting point for coding visual imagery. All forms of "hear" and "listen" were used to locate possible auditory imagery. For the three senses that were combined as "taste/smell/touch" for the purposes of this paper, all forms of "taste," "smell," "aroma," "scent," "feel," "felt," and "touch" were used as a starting point. This list of words was based on a preliminary reading of the dream reports for sensory terms by two coders.
After the possible sensory references were located by DreamSearch, the two coders studied each boldface word and its context to determine which sensory mentions seemed to be literal and which might be metaphorical in nature. For example, "I was feeling hurt," "I hadn't seen him in years," and "I never saw a dress like it" were flagged as likely metaphoric expressions. The importance of searching for possible metaphoric constructions is shown in unpublished interviews with an 18-year-old congenitally blind woman in which she was asked detailed questions about the imagery and sensations in her written dream reports (Hall, 1948). In addition, many words or phrases in her dreams that seemed to imply sight were in fact based on information from other senses, particularly auditory cues. As Kerr (1993) notes, terms or phrases that seem visual in dream reports from blind subjects may represent spatial constructions that do not require visual information.
Once the criteria were established through discussion and the coding of a few examples, the two coders agreed on virtually every coding. The few ambiguous instances, such as the "I never saw a dress like it" example in the previous paragraph, are examined in the results section. Every ambiguous instance involved words related to visual perception, which makes them especially important to resolve because they bear upon the issue of whether there was visual imagery in the dreams of participants who were born blind or became blind before the age of five.
Sensory references were divided into three categories: visual, auditory, and taste/smell/touch. The percentage of sensory references for each of the three categories was determined for each subject and then compared in a general way with (1) the age of blindness; (2) the degree of blindness; and (3) the percentage of a person's life that he or she has been blind. Since both sighted and blind subjects are known to include a significant number of auditory sensations in their dreams, the greatest emphasis was put on the contrast between visual and taste/smell/touch percentages. The "percent of life blind" ranged from 100% for the congenitally blind subjects to 14% for a 73-year-old woman who developed macular degeneration at the age of 63 and was legally blind even though she still had some peripheral vision.
For the second analysis, which focused on the substantive content of the dream reports, seven Hall and Van de Castle (1966) categories were used: (1) aggression; (2) friendliness; (3) misfortune; (4) success; (5) failure; (6) animal characters; and (7) food and eating. The aggression and friendliness categories were employed because they are often revealing and had been used in the studies by Kirtley (1975) and Kirtley and Sabo (1983, 1984). The misfortune, success, and failure categories were utilized because it was thought that blind people might suffer more misfortune and failure in their dreams due to their lack of vision. The "animal characters" and "food and eating" categories were included because Hurovitz's (1997) work with the blind suggested that such categories might be higher in blind dreamers due to their use of guide dogs and a tendency to obesity in older blind people.
Three types of analysis were made once the dream reports were coded for the seven categories. First, the percentage of dream reports with "at least one" instance of each category was determined for each participant. Second, the aggression and friendliness codings were used to determine four standard indicators discussed in Domhoff (1996, 1999):
The third and final analysis of the substantive content concerned the occurrence of misfortune in dreams in which the dreamer moved from place to place, whether by foot or vehicle. Such dreams were called "locomotion/transportation" dreams for purposes of this study. This analysis was included because in the post-collection interview 19 of the original 25 participants said that the "most difficult aspect" of their blindness was in moving from place to place (Hurovitz, 1997, p. 31).
In comparing the findings for the substantive categories with the Hall/Van de Castle norms, the h statistic developed by Cohen (1977) for use with proportional (percentage) comparisons was employed to determine both effect sizes and significance levels. The application of this statistic to the study of dream content is explained in Domhoff (1996, 1999). The table for determining statistical significance from h can be found in Domhoff (1996, p. 317). The DreamSat spreadsheet (Schneider & Domhoff, 1999) computes effect sizes and significance levels from h as part of its regular output.
There was a strong relationship between the lack of visual sensory references and (1) congenital or early childhood blindness; (2) total blindness; and (3) a high percentage of the person's total life spent in blindness. As shown in Table 2, four of the seven congenitally blind subjects who were totally blind had no indications of visual imagery in their dream reports (participants 11, 7, 10, and 2). Nor did a 49-year-old woman totally blind since 18 months or a 58-year-old woman blind since age 12 (participants 8 and 9).
At the other end of the continuum defined by the age and degree of blindness, 89% of the sensory references were visual for the partially blind 73-year-old woman who developed macular degeneration at age 63 (participant 3). The percentage of visual sensory references was 83 for a 27-year-old woman who became totally blind at age eight and 53 for a 50-year-old man who became partially blind at age seven (participants 1 and 12). A 50-year-old congenitally blind woman who was able to perceive very bright light had a visual references percent of 24 (participant 4). For a 45-year-old man totally blind since age 28, 9% of his sensory references were visual. All five of these participants with some visual imagery in their dreams were consistent with previous findings.
However, there also were three participants (13, 5, 15) whose references to visual imagery seemed potentially inconsistent with previous findings. Each of them is therefore discussed in more detail. Participant 13, a 24-year-old man totally blind since age four, was atypical in that two of the five sensory references in his 20 dream reports were visual. Neither of these visual references seemed metaphorical. In one dream he was sailing with his parents in a boat that was headed for a rock formation; he stated in his report that he "could see them [the rocks] clearly." In another dream he stepped out of a bus to "look" at some mountains and reported that he could "see them plainly." Thus, it seems possible that there are visual scenes in the dreams of some people who became blind before the age of five.
On the other hand, the four visual references coded for the two participants who had been totally blind since birth were very likely metaphoric expressions. In the case of participant 5, a 52-year-old woman, the metaphoric nature of her "looking" and "seeing" seemed apparent because in a dream where she and her husband (also blind) visited Thomas Jefferson at Monticello she reported that Jefferson "was glad to meet us and he didn't care if we couldn't see." However, she also said that Jefferson took them to "see" the plants in his garden. A month earlier, she had a similar dream in which she was "looking" at clocks and plants at Monticello, but without the guidance of Jefferson himself. This participant is also the one who reported that she "never saw a dress like it," the example of an ambiguous sensory reference provided in the previous section. Based on the evidence in her dream series as a whole, this ambiguous case is very likely a metaphorical expression.
Participant 15, a 46-year-old man, had an ambiguous visual reference in a dream based on a recent experience. It concerned his arrival at a maternity hospital to "see" his first grandchild. "Just like in real life," he reported, "we received the phone call, left the house, going down there, walking into the room and seeing the baby, holding the baby, it was a wonderful experience." As with participant 5, this visual reference seems to be metaphoric in nature, employing a frequently used conceptual metaphor in which "knowing" or "experiencing" is "seeing" (Matlock, 1988; Matlock & Sweetser, 1989; Sweetser, 1990).
Turning now to the findings on non-visual sensory references, the percentage of taste/smell/touch sensations was very high in the dream reports of participants who had little or no visual imagery. For the four participants who had no visual imagery in 13 or more sensory references (participants 11, 7, 10, and 2), 55% of the sensory references were taste/smell/touch and 45% were auditory. For participant 14, the 45-year-old male totally blind since age 28, 22 of his 34 sensory references (65%) were in the taste/smell/touch category, as compared to the 9% for visual references reported earlier. The sensations in such dreams were very strong. The participants "felt" the warmth of the sun, the texture of a coat, the edge of a knife, the slope of the ground, vibrations, snow, or the soft fur of a dog. They "smelled" fire, tobacco, aftershave lotion, fresh air, food, or coffee. They noted the "taste" of a cigar, a cup of coffee, or an orange. These dream sensations seemed to reflect their use of or pleasure in these sense modalities in waking life.
The findings for taste/smell/touch sensations in participants with little or no visual imagery in their dream reports contrasted dramatically with those for participant 1, the 27-year-old woman who became blind at age eight, and participant 3, the 73-year-old woman who developed macular degeneration at age 63. Out of 36 sensory references in the dreams of these two women, only one was in the taste/smell/touch category. Five of the sensory references were auditory, and the rest (86%) were visual. In short, their dream sensations and imagery seemed to be much like those of presently sighted people (Snyder, 1970; Zadra, Nielsen, & Donderi, 1998).
The results for the Hall/Van de Castle content categories were first examined participant by participant because the number of dream reports contributed to the overall sample by each person varied greatly. Despite some individual differences, this analysis showed that no participants were so atypical as to have created the group differences reported below for each category, except in the case of animal characters for women, where two women had animals in 44% and 24% of their dreams, as compared to 8.3% for the other eight women. These animals were most often their guide dogs. The detailed results of this individual analysis are not included in this paper.
Table 3 presents the results for the seven "at least one" analyses. It shows that the dreams of blind men and women were less likely to include either an aggressive or a friendly interaction than would be expected on the basis of the Hall and Van de Castle (1966) norms. The effect sizes were largest for the aggression category (h=-.72 for men, -.55 for women). Blind men and women also had significantly fewer dreams with either success or failure in them. There were no differences from sighted participants for the misfortune or food and eating categories, although it can be noted that a few participants were very high on food and eating dreams. As already stated, the slightly higher percentage of women's dreams with at least one animal character was due to two participants.
Table 4 presents the results for the four social interaction indicators. In three of the four categories the blind dreamers were somewhat below the norms for sighted dreamers, but the differences were not statistically significant except in the case of the aggression/friendliness percent, where the effect size h was -.38 for men and -.42 for women. In other words, unlike sighted men and women, the blind dreamers had more friendly than aggressive interactions in their dreams. This difference was due to their lower levels of aggressive interactions.
Table 3 showed that the percentage of all dream reports with at least one misfortune was very close to the Hall and Van de Castle (1966) norms. However, dreamer-involved misfortunes occurred far more frequently in the locomotion/transportation dreams of the blind than in similar dreams from the norms, as shown in Table 5. Sixty percent of the blind men's locomotion/transportation dreams and 61% of the blind women's had at least one dreamer-involved misfortune, as compared with only 31% for the male norms and 28% for the female norms. These differences are equivalent to Pearson r correlations between blindness and dreamer-involved misfortunes in locomotion/transportation dreams of .29 for the men and .33 for the women (Rosenthal & Rubin, 1982).
The findings from the analysis of sensory references in the dreams of these 15 blind participants were largely consistent with studies that Kirtley (1975) traces back to the early nineteenth century. However, participant 13, who became blind at the age of four, seems to present a small exception to the conclusion that people who become blind before age five do not retain visual imagery in their dreams.
The seeming inconsistency of visual references in the dreams of two congenitally blind participants, when looked at in the context of their overall dream series, suggests that these references were metaphoric in nature, making use of the conceptual metaphor "knowing" or "experiencing" is "seeing" (Matlock, 1988; Matlock & Sweetser, 1989; Sweetser, 1990). This conclusion is supported by Kennedy's (1993, 1997) finding of an unexpectedly close relationship between vision and touch, both of which rely on edges, junctures, and contours to make discriminations and develop an overall conception of an object.
Hall's (1948) unpublished interviews with the 18-year-old blind student concerning her sense of objects in her dreams may provide evidence for the close connection between vision and touch. In talking about a dream in which she was sitting around a table with her family in a very nice restaurant, she reported that "I knew we were at a table by kinesthetic sense and knew it was a nice place by auditory sense (thick carpets, quiet atmosphere, etc.)." Then she went on to explain: "I have a picture of a table because I know what a table felt like, not because I have seen one" (Hall, 1948, p. 3).
In another dream she described a beautiful table with two big silver candelabras on it. When asked how she knew the candelabras were silver, she replied it was because they were "very smooth to touch." When the interviewer asked her if she had noticed that she seemed to have a "preference for smooth things," she replied, "well naturally, because it's prettier to touch," thereby applying a highly visual term, "pretty," to something she likes on the basis of touch. She continued this line of thought by saying that "I think if anyone prefers rough textures it's because they are seeing them besides feeling them and the material might look pretty to them" (Hall, 1948, p. 21).
The findings on the lower percentage of dreams with at least one aggression are consistent with earlier findings for a smaller sample of dream reports from blind women (Kirtley & Sabo, 1983), but the low percentage of dreams with at least one friendly interaction does not fit with Kirtley and Sabo's (1984) finding of more friendliness in the dreams of blind women. Lacking information on the lives of the participants in either study, it is not possible to explain this difference with any certainty, but as noted in the introduction, the Kirtley and Sabo studies did not control for the fact that a large number of dream reports from one or two participants might be distorting their general results.
In the present study the participants were also low in the percentage of dreams with at least one success or at least one failure. When considered in conjunction with the findings on friendliness and aggression, the overall findings suggest that the participants have less social interaction and less striving in their dreams than the sighted normative men and women. Due to the lack of information on the amount of social interaction and striving in the waking lives of the participants, nothing more can be made of these findings unless they are replicated in future studies of blind dreamers.
Although it is not possible to draw any theoretical conclusions from most of the findings in this paper, some of them are consistent with the idea that there is continuity between dream content and waking cognition (see Domhoff, 1996, for a summary). First of all, the imagery and sensations in the dreams of the blind are generally continuous with the senses they use in their awaking lives, as also noted by Kerr (1993). Those born totally blind or who lost all of their sight very early in childhood usually have little or no visual imagery, but show the same detailed attention to sound, smell, touch, and taste that they do in waking life. The findings on those who lost sight at varying ages after early childhood suggest that visual imagery is gradually replaced by the sensations that come to be more important in their waking lives.
Second, the greater percentage of locomotion/transportation dreams with at least one dreamer-involved misfortune may be continuous with the waking-life concerns about traveling from place to place that were expressed in interviews with the first author (Hurovitz, 1997, p. 31). The specific content of these dreams supports this hypothesis as well. For example, one dreamer found himself on his hands and knees with his ear to the ground listening for traffic at a crosswalk. Another reported that she and her guide dog were lost, but then she pretended that she wasn't lost so no one would know.
Participant 7, a 44-year-old woman completely blind from birth, who dreamed of her guide dog in 14 of 32 dreams (43.8%), is relevant to this point. The dreams clearly suggested that she was extremely afraid she was going to lose the dog because of its advanced age. They also showed her dreaming of a new dog that was too big and thus unsatisfactory. Although studies of the great attachment that many blind people have to their guide dogs reveal that they are companions as well as important aids to independence and mobility, the great concern and mourning over the anticipated or actual loss of a guide dog suggests that there may be a continuity between dreams and waking life for participant 7 (Nicholson, Kemp-Wheeler, & Griffiths, 1995; Steffens & Bergler, 1998).
For the most part, as Kerr, Foulkes, and Schmidt (1982) also stress, the dreams of the blind reveal no great surprises in comparison to those of sighted dreamers, but they are just different enough on some dimensions to be of potential theoretical interest. Perhaps the findings from this study and the availability of DreamSearch make it worthwhile for other dream researchers to replicate the findings and develop new theoretical insights.
Go back to the Dream Library index.