Main Article:
Quantitative and Qualitative Analysis of Reported Dreams and the Problem of Double Hermeneutics in Clinical Research
The aim of this article is to show that statistical analysis and hermeneutics are not mutually exclusive. Although statistical analysis may capture some patterns and regularities, statistical methods may themselves generate different types of interpretation and, in turn, give rise to even more interpretations. The discussion is lodged within the context of a quantitative analysis of dream content. I attempted to examine the dialogical texts of reported dreams monologically, but soon found myself returning to dialogic contexts to make sense of statistical patterns. One could cogently argue that the reported statistical relationships in this study, rather than pointing to any interaction among the “signifieds,” speak only to the relationships among the “signifiers” that were being played out through various actors on the analytic or scientific stage, since all of the constructs used in theorizing about, interpreting, and telling dreams come from the same discursive system.
Index Terms: epistemology; double hermeneutics; methodological debate; statistical analysis; statistical relationship; interpretation; hermeneutics; psychoanalytic research; knowledge claim; dream narrative; linguistic structure; discursive performance
Suggested Citation: Movahedi, S. (2012). Quantitative and qualitative analysis of reported dreams and the problem of double hermeneutics in clinical research. Journal of Research Practice, 8(2), Article M12. Retrieved from http://jrp.icaap.org/index.php/jrp/article/view/303/274
The debate in the human sciences today over their status as science or hermeneutics may seem to be a case of the return of the repressed. Some, however, may consider it to be the periodic intrusion of Hartmann’s (1964, chap. 18) ghost that never went away. We can still hear the old voices ringing in the ongoing methodological dispute between those who call themselves “natural” scientists and those who maintain that social and behavioral fields of inquiry are inherently interpretive and hermeneutic and should be studied with that in mind. Those adopting the natural-science position are hopeful that, by reducing meaning to some form of brain functioning, they can become the biologists of the mind rather than the analysts of the soul. In turn, members of the hermeneutic circle reduce research in the social and behavioral sciences to textual analysis, subject only to the requirement of internal coherence. There are also those who agree with the interpretive tradition, but maintain that textual analysis goes beyond the hermeneutic method, in that interpretation of the text changes the text itself. This is what I call “the problem of double hermeneutics.”
I would like to explore this methodological debate within the context of empirical research in psychoanalysis. I view psychoanalysis as a method of listening to the unconscious that, like any other form of social-scientific endeavor, is subject to theoretically grounded observation. There is no inherent limitation in the psychoanalytic data that renders it unsuitable for any form of analysis; any observation or communication can be analyzed by some qualitative or quantitative method. Statistical analysis allows a researcher to search for recurring patterns or structural regularities in the data. These patterns or structures are not inherent properties of the phenomenon under investigation; rather, they are a function of both the measuring instruments and statistical methods used for data analysis. Structures are matters of inference and construction. They are theoretically imposed--interpreted--rather than discovered. It is in this sense that even more rigid types of quantitative research are forms of interpretation.
This article is an attempt to transcend the false dichotomy between observation and interpretation--with its epistemological distinction between, respectively, explanation and understanding--in psychoanalysis, by suggesting that any observation is an interpretation and any explanation is an understanding.
I shall begin my discussion with the report of a quantitative study of dream narratives, which would be considered a sacrilegious endeavor in psychoanalysis. The project emerged from a psychoanalytic research workshop attended by 37 analysts and psychoanalytic psychotherapists. One critical aim of the workshop was to address the hotly debated controversy over the “proper” criteria for the evaluation of knowledge claims in psychoanalysis. We are all aware of the call for new approaches to research and new forms of data analysis in psychoanalysis. Yet epistemological confusion about the nature of empirical knowledge continues to fuel this methodological controversy. Empirical knowledge has come to be associated with quantification, and empirical research has been understood as any systematic study that allows for some form of statistical analysis (Cooper, 1993). One of my aims in leading the discussion at the workshop was to show that statistical analysis and hermeneutics are not mutually exclusive. At the same time, I hoped to demonstrate that if dreams--intentional dialogic textual discourse, in Bakhtin’s (1986) words--are transformed into voiceless variables, correlations among these variables would themselves require interpretation within the context of the original discourse.
The project examined patients’ unconscious fantasies and the patterns of object relations that emerged in dreams through analysis of manifest dream contents. To accomplish this, workshop participants all coded a sample of patients’ dreams in terms of theoretically imposed categories such as the relationship between self and others, the number of characters, the intensity of interactions among them, and the dreams’ spatial settings. The major goal was to examine, hermeneutically, any pattern that emerged and any cultural voice that was heard. Similar to the clinical vignettes presented in psychoanalytic papers, findings on the complex patterns in relationships among various variables are presented here only to facilitate discussion and not to draw any statistical inference about a particular population.
Participants in the workshop were asked ahead of time to bring four or five dreams (“raw” dreams, with no interpretation) from four or five different patients. They were instructed not to look for exotic dreams in particular; any dream, no matter how short or meaningless, would be acceptable. Participants were also asked to complete a short inventory, consisting of eight questions, for each dream. Some of the items were primarily diagnostic and assessed the dreamer’s level of projection, narcissism, and tolerance for others, and the rest were demographic (age, gender, socioeconomic status, and length of time the dreamer had been in psychoanalysis or psychotherapy). All questions were to be answered by the analysts or therapists based on clinical data or their perceptions of their patient-dreamers.
A sample of dreams was read in the workshop to develop a set of theoretically informed coding categories. A number of categories were agreed on to assess each manifest dream: level of object relations, number of people, explicit levels of wish or fantasy, level of problem-solving, emotional qualities--such as levels and types of expressed or exhibited affects--spatial settings, and so forth. The whole group then coded a number of dreams as an exercise in calibrating the coding protocol. When all participants felt they understood the explicit set of criteria for coding the manifest content of dreams, they were divided into groups of three or four to jointly code their patients’ dreams. The members of each group had to agree on any specific coding category; if they failed to achieve consensus on a specific category, the judgment of the majority was to be adopted. The numerically coded dreams were then collected, computerized, and subjected to various statistical analyses.
At the risk of being redundant, I should emphasize that this work is presented as a collective exercise in quantifying dreams, which includes the interface of various fantasies at different levels. Theoretical coding and statistical analyses can be taken metaphorically as what a practicing analyst does implicitly in listening to or making sense of dreams. The extent to which the researcher’s implicit assumptions, theories, and expectations (i.e., ideologies or fantasies) shape their findings is a major issue in the sociology and philosophy of science. A dataset cannot speak for itself; we speak for data, and we speak in different languages and from different positions.
Analysis of the manifest content of dreams has been of much interest to psychologists and sociologists. Although Freud himself pioneered the analysis of manifest dream content, most psychoanalysts have shied away from such research. It should be noted that Foucault, both in his paper on dreams, imagination, and existence and in the last volume of the History of Sexuality tries to reverse Freud’s position on the relative importance of the manifest content of the dream, arguing that manifest content, social-level, cosmos koinonia are the significant elements of the dream--not the latent content, private-level, cosmos idios. Analysts have generally maintained that the manifest content of dreams has its own structure, which is intimately linked to the dreamer’s intrapsychic functioning and to his or her mode of object relations.
Analysis of our data, however, reveals some interesting and theoretically meaningful patterns of multivariate relationships. Without questioning the personal and private domain of dreams, we wonder how to account for their statistically significant common patterns.
(a) In this study, women’s dreams exhibited more wishful thinking than men’s dreams.
(b) Men’s dreams were judged to be more unreal than women’s dreams.
(c) There was no relationship between space and gender. This is contrary to the stereotype and to previous findings in dream research, according to which women’s dreams tend to be staged indoors or in closed areas.
(d) Women’s dreams exhibited a higher level of feeling than men’s dreams.
(e) Women’s dreams scored higher on the level of object relationship than men’s dreams.
(f) Women’s dreams were more populated than men’s dreams.
(a) There was no relationship between the dreamer’s social class--as rated by the analyst--and the number of people in the dream. This finding is also contrary to other manifest-dream research, according to which lower-class subjects report a greater frequency of human characters in their dreams (Brenneis, 1975).
(b) There was, however, a significant relationship between the dreamer’s social class and the analyst’s evaluation of the dreamer’s level of projection and reality testing. That is, higher-status analysands (i.e., persons whose dreams are being analyzed here) were perceived as less projective and more realistic in their perception of others than were lower-status analysands.
(c) At the same time, there was a significant relationship between the dreamer’s social class and the analyst’s evaluation of his or her level of narcissism (lack of need for or tolerance of others): The lower the social status of the analysand, the more likely the “diagnosis” of narcissism. Since social class was measured by the analyst’s subjective estimate, this may simply mean that the analyst assigned less narcissistic (or pathological) patients to a higher social class.
(d) The dominant feelings among higher-status dreamers were fear and happiness and, among lower-status analysands, anger and confusion. This may suggest something about the kinds of affects that are more socially acceptable in different classes.
(a) The intrapsychic world of the narcissist, as projected on the dream screen, is thinly populated (Kernberg, 1975, p. 85). The number of people in the dream and the types of feeling may say a great deal about the level of narcissism. In this study, the presence of others in the dream was significantly related to the types of feeling in the dream. When there was no one else in the dream, the dominant feeling was fear (39%), with a little anger (5.4%) or sadness (5.4%), which implies that anger and sadness are more in need of objects than fear is.
(b) The presence of others in the dream was significantly related to the dreamer’s level of narcissism, as independently diagnosed by the analyst reporting on the dreamer. It was also related to dreamers’ levels of conflict resolution, object relatedness, and reality orientation in their dreams. That is, narcissism, a reality sense of the dream, and object relations all co-varied with the presence of others in dreams.
(c) There was a significant relationship between the analyst’s rating of the dreamer’s level of narcissism--as operationalized by his or her degree of tolerance of, or need for, others--and the level of object relationship in dreams. This may speak to the validity of the analysts’ diagnostic perceptions.
(d) Since gender was significantly correlated with the analyst’s estimate of the dreamer’s need for others--for instance, significantly more women were diagnosed as having a high need for others--a partial correlation was computed between the need for others and the level of object relationship in dreams, controlled for the variable gender. Nevertheless, the correlation continued to be significant.
(e) There was also a significant relationship between the analyst’s rating of the dreamer’s need for others and the quality of feelings in dream (negative, positive). When we used a moderate level of narcissism (i.e., need for others) as our baseline, both positive and negative feelings changed at both ends of the continuum. That is, when there was a high level of need for others, there was also a great deal of both positive and negative feelings in the dream. Similarly, when there was little or no need for others, the level of negative feelings rose while the level of positive feelings fell drastically. However, positive feelings--as reported by the analysts--again changed drastically with a change in the need for others. The higher the need for others, based on the analyst’s report, the more likely that positive feelings would be present in the dream.
Since reporting a dream is a communication to the listener, we examined the relationship between the analyst’s gender and other variables.
(a) In general, more than 60% of the feelings in dreams that were reported to both men and women analysts were negative (anger, fear, sadness, etc.). Yet the dominant feeling of dreams reported to female analysts was fear, while the dominant feeling reported to male analysts was sadness. Wishes in dreams were also more clearly expressed to male analysts (46%) than to female analysts (33%), and the difference was statistically significant.
(b) Dreams reported to male analysts tended to exhibit more conflict resolution (59%) than those reported to female analysts (41%).
(c) Female analysts were more likely to rank their patients lower on reality-testing than their male counterparts were.
(a) As the number of years the dreamer had been in psychotherapy or psychoanalysis increased, the number of people present in the person’s dreams also increased.
(b) The longer the length of the therapy, the more realistic the dream.
(c) However, length of therapy did not affect the level of object relationship in the dream.
(d) The level of wishful fantasy changed inversely with length of treatment (i.e., wishful fantasy decreased as time in treatment increased).
(e) The level of feeling in reported dreams increased with length of treatment; suggesting that the patient increasingly learns to express his or her feelings or becomes more versed in the discourse of psychotherapy. The level of feeling in dreams reached its peak at the end of 2 years of therapy then began to drop. Of those with little or no psychotherapy, 64% of their dreams were evaluated as exhibiting no feelings, in comparison with 5% of those who had had 1 to 2 years of therapy.
(f) The same pattern seemed to hold for the relationship between length of psychotherapy and level of conflict resolution in dreams. The relation was curvilinear. Dreams of 65% of the beginners, versus 31% of those who had had 1 to 2 years of psychoanalysis, showed no conflict resolution. The level of conflict resolution in dreams increased, however, with the length of treatment, reached its maximum at the end of the 3rd year, and then decreased.
(g) Length of psychotherapy was also related to whether feelings in dreams were positive or negative. Beginners’ dreams contained a lot of negative feelings. The feeling tone then became positive but, after 3 years of therapy, reverted to negative.
(h) Types of feelings were also related to length of treatment. At the beginning of treatment, the dominant feeling in dreams was fear (39%). Within the 1st year, this segued into confusion (46%), changed into happiness within the 2nd year (37%), and, after 3 years, ended up as almost equally distributed feeling types.
(i) Length of treatment was also related to the level of wishful thinking, with the maximum occurring within the 2nd and 3rd years (55%) and then dropping back (36%) to almost the beginning level.
(j) Length of psychotherapy was also related to the dreams’ realism, becoming steadily more realistic with greater time in psychotherapy, peaking at the end of the 3rd year, and dropping again after that.
Interpretation of the dream space, or the spatial structure of dream narratives, is complex: In dreams, is the meaning of space independent from the meaning of time? Space and time in dreams have nothing to do with physical space and time; rather, they are part of the private discourse of emotional experience. In a therapeutic situation, in which the 50-minute analytic time is a function of the analyst’s office space, space may signify an emotional communication as to the differential level of desire for closeness. Indeed, most of the reported dreams in this study had been staged indoors (57.5% versus 21% outdoors and 21.5% with no specified space)--a pattern that may be different from reported or solicited dreams in nontherapeutic situations.
In this connection, it may be of interest to point out that to Klein (1926), displacement in space--“the change from intra-uterine to extra-uterine existence”--is the foundation of orientation in time. In psychosis, similar to dreams, time and space are interchangeable; the psychotic may try to go back in time by taking steps backward in space (Movahedi, 1998). The spatial pattern of a person’s recurring dreams may also speak to the dreamer’s cognitive style, level of adaptive or defensive functioning, or the self’s sense of existential grounding.
Having detected no specific pattern in reference to spatial setting, we decided to re-code the variable space into a dichotomy of spatial and nonspatial dreams. We hypothesized that space--any space--says something about the emotional imbedding of the experience and about the existential grounding of the dreamer’s self. This is similar to Foucault’s (1954) claim that the form of spatiality in dreams speaks to the meaning and direction of the dreamer’s existence.
The relationships between the spatial structure of dreams and other variables in the study were as follows:
(a) The types of feelings in dreams were significantly related to the dream’s spatial structure. Fear and sadness were the dominant feelings in spatial dreams and confusion and happiness in nonspatial dreams. The level of feeling related to the dream’s spatial structure: there were significantly fewer feelings in nonspatial dreams.
(b) The analyst’s diagnosis of the patient’s level of reality testing was significantly related to the dream’s spatial structure: The higher the degree of reality testing, the higher the likelihood that the dream was spatial.
(c) There was a significant relationship between age and spatial structure. Among those 13-17 years old, 63% of the dreams were nonspatial.
This study was based on the theoretical assumption that individuals’ linguistically constructed unconscious fantasies would dominate their attitudes and expectancies about the external world. Such fantasies reflect relationships between the self and others that are re-projected onto the external world. Internal self-other dialogues that are emotionally experienced emerge in dreams and are taken as a reflection of such attitudes and expectancies. Between the dreamer’s imagery and the narrated dream, however, there can be a vast and complicated hermeneutic gap. This gap may be somewhat similar to that between Saussure’s (1974) langue and parole--i.e., between the images in a private psychic system and a performance that involves emotional communication to an analyst within a particular discursive context. In this regard, I agree with Gray (1992) and Pulver (1987), who, in Pulver’s words, believe that “there is no such thing as the manifest dream” (p. 102). Instead, the manifest dream varies each time a dream is reported, conveying the dreamer’s context-specific immediate feelings, wishes, and fantasies. In that sense, every so-called manifest dream is a discourse of the unconscious.
Although the quantitative approach used for the analysis of dreams in this article attempted to study dialogical text monologically, we must return to the original dialogic contexts to make sense of statistical patterns. We have to convert the data back to its multiauthored and polyphonic status. To begin with, the dreams examined here, coming from the analytic couch, should be viewed as part of the analytic exchange, which is an enactment of passion textually symbolized in a discourse of fantasy between two subjects. It is, as Kristeva (1987) puts it, a discourse of love. It is also a discourse of fantasy, not unlike a waking dream. The function of this exchange and the goal of this dialogue are, in Ricoeur’s (1977) description, the restoration of the “original” latent text in desire. The patient’s report of a dream is itself an act of textual restoration or self-interpretation. For instance, as I reported above, dreams reported to male and female analysts differed in emotional texture. Because of their gender, analysts may elicit different feelings from patients--or, alternatively, analysts may be more sensitive to various feelings. Patients who easily detect their analyst’s generalized affective state may unconsciously produce dreams or fantasies that that will align them emotionally. Women analysts may be more sensitive to fear, for instance, and male analysts to depression. One may also surmise that the analytic discourse with a female analyst is different from the analytic discourse with a male analyst. Also, since these dreams were reported by analysts, the dreams may communicate something about the analysts’ own feeling states. Why would male analysts report dreams with feeling tones that differ from those reported by female analysts? Perhaps female analysts have been communicating about their own fears, while male analysts may have been communicating about their own depression. In this sense, an analyst’s choice of dreams to report or to remember may itself be autobiographical.
We may also note that the level of feeling and the level of realism in reported dreams were significantly correlated with the length of time in psychotherapy or analysis. The level of wishful fantasy changed inversely with the length of the treatment, decreasing as years in treatment increased. Hence, the question here is: Do patients in psychoanalysis or psychoanalytic psychotherapy tend to become more aware of their own feelings the longer they stay in therapy, or do they learn a new language in talking about their mental images? And do they learn this new language for talking about their mental states or dreams in the context of the discursive system?
The relationship between the spatial structure of dreams and age is also intriguing; it even jibes with the youth culture’s lingo of being “spaced out.” But again, the question is whether the spatial structure of youths’ reported dreams reflects their alienation and crisis in identity or the developmental mode within which they organize their story lines. According to Bruner (1992), 10-year-olds tend to organize their stories in plots that are acted out by the protagonists’ subjective states; there seems to be a little disjunction between the inner landscape of consciousness and the outer one. Teenagers depict the world as a series of time-pressed plights, in which inner state and external events are in a race with each other and a sense of subjective urgency permeates their stories. Adults, on the other hand, tend to depict their experiences in a dramaturgic mode: Plight is organized in terms of agent, action, scene, goal, and instrumentality. A collision between two or more of these elements creates trouble (Bruner, 1992).
A reported dream is hardly a description of images, or of photographs, or a film of fantasies that have been played out on the stage of the internal theater. To Barthes (1977), we cannot describe even a photograph without imposing a code on it. The photograph has a denotative status, containing a first-order message that exhausts its analogical content. By being absolutely analogical--that is, lying “outside of any recourse to a code”--the message is “neutral” and “objective.” The press photograph, however, is connoted; it is reworked in terms of aesthetic or ideological codes, and the “objective” message, paradoxically, becomes “invested.” In dreams, images are all invested from the very beginning of narration. There is no such thing as purely analogical content in dreams. We doubt whether there is such a thing as an image without a code even in photography. A patient who had just been divorced was admonishing herself for having been a poor observer of their wedding pictures. “Looking at our wedding pictures,” she said, “you can clearly see how he [the husband] is distancing himself from me. I don’t know how I could have missed this. ” A photograph becomes “invested” by the fact of being a photograph, a selected image of literal “reality.” There is no need for an accompanying text--parasitic text, according to Barthes--to carry out the signification. Texts that need other supporting materials to be intelligible, or to have some relatively coherent interpretation, are called parasitic.
In later work, Barthes (1982) admits that the distinction between the literal image and the symbolic image is an arbitrary one, introduced only for the operational reason that “we never . . . encounter a literal image in the pure state; even if an entirely ‘native’ image were to be achieved, it would immediately join the sign of naiveté and be completed by a third, symbolic image” (p. 31).
Nevertheless, the difference between the images in dreams and photographic images in the press is that the latter images are observed in the context of written materials that structure their meanings, while the former images come to us--the non-dreamers--only as texts that must draw on our own fantasies, ideas, or images to be interpreted. Listening to other people’s dreams, we may have to conjure images or fantasies in our mind to link the dreamer’s reported images to our own. This means that when patients’ dreams are reported in psychoanalytic literature or in conferences, they have already been, in some sense, invested with the analyst’s own fantasies.
The same holds true in this study. The dreams that analysts brought to the workshop were themselves reconstructed texts. They were not the verbatim reports of their patients; they were the analysts’ verbatim reports of their patients’ reports. They had all, therefore, been edited. Whether we like it or not, editing is itself a hermeneutic exercise, a form of interpretation that carries all the ideological and countertransferential baggage of any other interpretation. In that sense, one could even claim that we have studied not the patients’ dreams but rather the analysts’ public presentations of their interpretations of their patients’ dreams.
I should add that the story line and the structure of the reported dreams in the workshop nicely matched the grammatical structure of psychoanalytic interpretation. Many psychoanalytic writers (e.g., Grotstein, 1979; Heynick, 1981) speak about the linguistic structure of dreams as though they are dealing with the original text of the dream as it had appeared in the patient’s mind or as some kind of “real” photographic entity.
Dreams reported in analytic sessions are not independent manifestations of the unconscious of one subject (the patient) as understood by another subject (the analyst) who are both constituted outside the analytic discourse. The analytic patient, the presented dreams, the unconscious, and the deciphering subject all belong to the same epistemic system. The unconscious is not outside that system, which renders sensibility, legitimacy, or credibility to an interpretation. The Russian linguist Bakhtin (1986) would perhaps find the dream images themselves in the private psychic system to be dialogic, that is, intimate inner conversations among different voices--intrapsychic representations--in a space located between the self and other.
Reported dreams follow the rules of spoken language. They are verbal speech produced for the ear of the other, the analyst, in the interpersonal context of the analytic situation. When a patient’s dream is reported by an analyst, however, the presence of the patient is filtered though the presence of the analyst (Olinick, 1984). In psychoanalytic reports, papers, or presentations, we rarely hear the “voice” of the patient. The voices of participants are often heard through one another’s transference-countertransference filters. Nevertheless, the clinical vignette is written by the analyst. Similar to reported dreams, clinical presentations are a form of secondary elaboration, except that here the manifest content (the patient’s reported “voice”) “hides” the latent content (the analyst’s “voice”) (Movahedi & Wagner, 2005). Thus, instead of talking about the structure of dreams, we should be talking about the structure of the analyst’s listening.
A similar point has been made by Bartlett (1932). In his experimental study of memory and recall, Bartlett gave his English subjects a story to read and reproduce. The story was a North American Indian folktale, “The War of the Ghosts.” He noted that his experimental subjects unwittingly introduced much transformation, omission, and reconstruction in the content and form of the story to normalize it and fit it into the English narrative structure. Subjects frequently remarked that the story was “not an English tale.” Labeling a narrative as “not English” or calling it a “dream” rendered it acceptable. “When an Englishman calls a tale ‘not English’ he can at once proceed to accept odd, out of the way, and perhaps even inconsistent material, with very little resistance” (Bartlett, 1932, p. 85).
We are faced with yet another question. We do not know why the participating analysts presented those particular dreams. If a dream is an instance of self-other communication, may we say that reciting someone else’s dream is also a self-other communication? How much do such dreams communicate about the analyst, and how much about the analysand? If every analytic case presentation is an instance of countertransferential enactment, as Michels (2000) has eloquently argued, why can we not say the same thing about the analyst’s presentation of the patient’s dreams, that is, the analyst’s choice of dreams for the workshop? Does the patient’s dreams--which the analyst remembers, reports, or writes about--come to represent the analyst’s own dreams? Also, if in the narration of dreams, the individual’s voice is audible through a public performance, addressed to a particular self-object, within a particular discourse, and in a particular dialogue, who is the author of the dream? That is, who owns the dream? Whose fantasy does it represent? I have had the pleasure of being present at a number of presentations by an internationally famous analyst who often makes reference to a fantasy or a vignette about a “cutting vagina.” Although the analyst is presenting the fantasies of patients with malignant narcissistic personality disorders, I wonder whether the pattern of reported fantasies is itself a countertransferential enactment.
Some analysts may insist that dreams have their own intrapsychic meanings that are independent from their analytic, social, and cultural surrounds. That assumption, however, requires the operation of a non-corrupting privileged language in which dreams can be captured. Translation of the dream language into the ordinary language is itself a form of interpretation--and it is reasonable to argue that dreams in their “private” culturalized language are interpreted fantasies. We may even take Thomas Mann’s (Saal, 1982) position that dreams are dreamt because they have already been interpreted. As Wittgenstein has argued, “The idea that there is a hidden meaning, which is the meaning of the dream, can, in fact, only be the result of a decision about the kind of interpretation we are willing to consider” (Bouveresse, 1995, p. 116). In other words, “It is the acknowledgement of the interpretation that determines and defines what we are looking for in our search for meaning” (Bouveresse, 1995, p. 117).
Free association may be a strategy or incentive to get the analysand directly involved in the construction of dreams, or in re-dreaming a dream in the analytic context. However, the construction of an interpretation on the basis of free association does not logically give us a better translation or a “truer” narrative.
Is there even such thing as the “original text”--the “latent content”--of the dream to be excavated by free association? The role of free association is to provide a discursive context for such construction. In terms of Foucault’s (1970, p. xiv) methodology in his own analysis of The Order of Things, Freud’s analysis of dreams is based “not on a theory of the knowing subject [the dreamer or the interpreter], but rather on a theory of discursive practices.” What is a “hidden unconscious discourse” as opposed to a “superficial manifest conversation” has to do with discursive rules that structure what can and cannot be thought and expressed in an analytic session and prescribe who is, and who is not, in a position to decide on a particular narrative--among many--as the favorite unconscious communiqué.
Bertram Lewin used to ask the members of his dream seminar to interpret the latent meaning of a dream without knowing the dreamer, her association, or the context of the dream. He would do this by asking them to free-associate collectively to the elements of a dream’s manifest content. The seminar members’ interpretation would closely match the “actual” latent meaning of the dream that had previously been arrived at by the dreamer’s analyst, based on the patient’s free associations and years of analysis (Allison, Loeb, & Spain, 1993). To test the validity of Lewin’s method of dream analysis, Allison and colleagues (1993) conducted a double-blind study by asking 21 analytic subjects to free-associate to the manifest content of two dreams. The dreams came from the file of an experienced analyst who had discovered the latent meaning of these dreams based on the patients’ free association to elements of the manifest dreams. Allison et al. (1993), argue that their study’s findings corroborated Lewin’s method of group free-association: There was a close correspondence between the subjects’ opinions and the treating analyst’s opinion as to the latent meanings of the dreams. The authors further contend that their data show that, without the dreamer’s associations, the context in which the dream was reported, or the dreamer’s associations to the dream, “some individuals could sometimes arrive at the principal latent meanings of manifest dreams” (p. 147).
“Some individuals” were analysts or analytic candidates who believed in the same psychoanalytic theory and belonged to the same analytic institute. In Allison et al.’s (1993) study, neither the sole Klienian analyst nor any of the “analytically naïve laypersons” in the original sample rendered an acceptable interpretation, and their responses were not included in the data analysis. Accordingly, didn’t the researchers’ data simply reflect a particular psychoanalytic theory’s rules for the analytic interpretation of dreams? I believe this is an excellent example of Wittgenstein’s view of textual interpretation, which posits that the “meaning” of dreams is not independent from the “rules” for their interpretation. The notion of an objective meaning of a dream at a latent or manifest level should be replaced by engagement in the psychoanalytic language game, which is engagement in a specific linguistic practice in a particular social context. What we have in dreams is the individual’s fantasy communicated through role-specific discursive performance. Discursive performances are governed by rules, and the rules reside in a shared symbolic space that may account for much consistency across individuals. With no private language for the individual to express his or her “inner reality” (inner speech), we are at the mercy of our intuition to listen to the person’s private voice through the public performance. And as Rorty (1991), quoting Wittgenstein, argues, “intuition is never anything more or less than familiarity with a language-game” (p. 34).
In sum, statistical analysis may capture some patterns and regularities, but statistical methods of analysis are themselves different forms of interpretation that provide grounds for yet more interpretations. The patterns and regularities identified by statistical methods may also speak to some dream genres. Following Bakhtin’s (1986) analysis of speech genres, we may introduce a distinction between primary (simple) and secondary (complex) dream genres. Freud’s (1900) discussion of recurring dreams, like flying dreams, falling dreams, death dreams, losing-a-tooth dreams, and so forth, may exemplify simple symbolic frames for interpreting dreams.
Dreams presented in psychoanalysis have their own, more complex, genres. Perhaps this is why there is great emphasis on the patient’s first dream in analysis, when it is relatively uncorrupted by the analytic discourse. This does not mean, however, that the dreamer’s authorship is absent in reported dreams. Similar to novels written in the same historical and literary genre, every reported dream is a psychic construction of the individual and represents the particular style, or individuality, of the dreamer. But, as Bakhtin believes:
A work’s author is present only in the whole of the work, not in one separate aspect of this whole, and least of all in content that is severed from the whole. He [the author] is located in that inseparable aspect of the work where content and form merge inseparably, and we feel his presence most of all in form. (Bakhtin, 1986, p. 160)
One could even argue--and I believe quite cogently--that the reported statistical relationships in this study, rather than pointing to any interaction among the signifieds, speak only to the relationships among the signifiers that are being played out by various actors on the analytic or scientific stage: All of the constructs that were used in theorizing about, interpreting, and retelling dreams come from the same grand symbolic space.
Allison, G. H., Loeb, F., & Spain, D. H. (1993). Lewin’s manifest dream exercise revisited. Journal of the American Psychoanalytical Association, 41, 127-150.
Bakhtin, M. M. (1986). Speech genres & other late essays (C. Emerson & M. Hoquist, Eds., V. W. McGee, Trans.). Austin, TX: University of Texas Press.
Barthes, R. (1977). Image, music, text. New York: Hill & Wang.
Barthes, R. (1982). The responsibility of forms. Los Angeles: University of California Press.
Bartlett, F. C. (1932). Remembering: A study of experimental and social psychology. New York: Macmillan.
Bouveresse, J. (1995). Wittgenstein reads Freud. Princeton, NJ: Princeton University Press.
Brenneis, C. B. (1975). Theoretical notes on the manifest dream. International Journal of Psychoanalysis, 56, 197-206.
Bruner, J. (Commentator). (1992, April). The original story and the considered story. Invited Symposium at the Twelfth Annual Meeting of the Division of Psychoanalysis, American Psychological Association, Philadelphia, PA.
Cooper, A. (1993). Discussion: On empirical research. Journal of the American Psychoanalytical Association, 41S, 381-392.
Foucault, M. (1954). Dream, imagination and existence. In K. Hoeller (Ed.), Dream and existence (pp. 31-78). Atlantic Highlands, NJ: Humanities Press.
Foucault, M. (1970). The order of things: An archaeology of the human sciences (A. Sheridan, Trans.). New York: Random House. (Original work published 1966)
Freud, S. (1900). The interpretation of dreams. In J. Strachey (Ed. & Trans.), The standard edition of the complete psychological works of Sigmund Freud (Vols. 4-5). New York: Norton.
Gray, P. (1992). Memory as resistance, and the telling of a dream. Journal of the American Psychoanalytical Association, 40, 307-326.
Grotstein, J. S. (1979). Who is the dreamer who dreams the dream and who is the dreamer who understands it: A psychoanalytic inquiry into the ultimate nature of being. Contemporary Psychoanalysis, 15, 110-169.
Hartmann, H. (1964). Essays on ego psychology. New York: International Universities Press. (Original work published 1927)
Heynick, F. (1981). Linguistic aspects of Freud’s dream model. International Review of Psycho-Analysis, 8, 299-314.
Kernberg, O. (1975). Borderline conditions and pathological narcissism. New York: Jason Aronson.
Klein, M. (1926). Infant analysis. International Journal of Psycho-Analysis, 7, 31-63.
Kristeva, J. (1987). In the beginning was love: Psychoanalysis and faith. New York: Columbia University Press.
Michels, R. (2000). The case history. Journal of the American Psychoanalytical Association, 48, 355-375.
Movahedi, S. (1998). The discourse of time and the structure of psychic reality. Modern Psychoanalysis, 23, 197-209.
Movahedi, S., & Wagner, A. (2005). The “voice” of the analysand and the “subject” of diagnosis. Contemporary Psychoanalysis, 41(2), 281-305.
Olinick, S. L. (1984). Psychoanalysis and language. Journal of the American Psychoanalytical Association, 32, 617-653.
Pulver, S. E. (1987). The manifest dream in psychoanalysis: A clarification. Journal of the American Psychoanalytical Association, 35, 99-118.
Ricoeur, P. (1977). The question of proofs in Freud’s psychoanalytic writings. Journal of the American Psychoanalytical Association, 25, 835-871.
Rorty, R. (1991). Objectivity, relativism, and truth. New York: Cambridge University Press.
Saal, F. (1982). El lenguaje en la obra de Freud [Language in the work of Freud]. In N. A. Braunstein (Ed.), El lenguaje y el inconsciente Freudiano [Language and the Freudian unconscious] (pp. 11-65). Delegación Coyoacán, Mexico: Siglo XXI Editores.
Saussure, F. (1974). Course in general linguistics (W. Baskin, Trans.). London: Fontana/Collins.
Received 1 March 2012 | Accepted 1 December 2012 | Published 16 December 2012
Copyright © 2012 Journal of Research Practice and the author