The Dream as Text, The Dream as Narrative - Patricia Kilroe

Dreams are intriguing phenomena of the mind. There seem to be certain shared processes, for example, that are important to both dreaming and language. Some of these have been well-explored in the century since Freud published The Interpretation of Dreams. In my own work work (Kilroe 1997, 1998, 1999a) I have explored the role of metaphor, metonymy, and punning in the formation of dream imagery.

Now I would like to turn my attention to consideration of the question, Is the dream a text? And following that, a narrower question, Can the dream be properly thought of as a narrative? Although it has been assumed in past decades that text and narrative are terms appropriately applied to dreams, I would like look closely at these assumptions. I am going to propose that all dreams are texts, although this proposal rests on a particular use of the term dream, which is to be clarified below. I also propose that some but not necessarily all dream texts are narratives. I am interested here in the form of dreams, in their narrativity, by way of inquiring into the cognitive aspects of narrative structure and its relation to language. I am also interested in the role of language in the generation of dreams. I am not here attempting to formulate an interpretative system for dreams, although I believe an encounter with potential meaning is unavoidable, accepting as I do that the dream is a metaphor in motion (Ullman 1969), and that in narrative the functional units are already units of content: “It is what a statement ‘means’ which constitutes it as a functional unit” (Barthes 1994:105).

To begin with, the term text has been understood broadly in recent decades, and it is this broad understanding which permits the notion of text to be extended to dreams. We may say that texts “can be any physical structure at all made to embody ideas in the semiotic sense,” (Deely 1990:64-5) and that text is “the primary element (basic unit) of culture,” generated by the systems of cultural codes (Uspenskij et al. 1973:6). Hence the application of text semiotics to a broad range of cultural phenomena. Danesi (1999:30) for example, includes “routine conversations,” along with “musical compositions, stage plays, dance styles, and ceremonies,” as examples of “products of our text-making capacity.”

Although the claim has been made that the concept of text “in its broadest sense refers to messages of any code” (Nöth 1990:331), it will be critical to my argument to recognize that “the terms message and text are not synonymous. A message refers to what one wishes to communicate; a text refers to how the message is constructed” (Danesi 1999:30). Whether or not a text must convey a message is a question of relevance to the study of dreams. For purposes of the present discussion I will define a text by formal rather than semantic criteria. A text must have definable parameters; it must be separable from other phenomena of experience by spatiotemporal boundaries. Thus a text is a formal unit. It also has coherence, the determination of which is made by the experiencer of the text; it is content expressed through a formal code which makes use of cohesive links to create an impression of unity.

A text, then, is a form that must have content, but that content does not necessarily have a message. We could say, for example, that Lewis Carroll’s Jabberwocky is a text because as a poem it is a formally cohesive unit, but that its content is semantically empty and hence conveys no message. (If there is a message, it is arrived at only when an interpreter views the poem as a signifier in a broader cultural context.) Conversely, a specific message such as “children are lovable” could be formalized through a variety of codes, e.g., a photo exhibit in a museum, a book without illustrations, or a televised public service announcement. Each different product of encoding this message would have to be considered a separate text despite the identity of the message of their content. In general, however, a text without a message seems a rather pointless exercise. Communicators that they are, people seem to pay more attention to meaning than to form. Because a form whose content is semantically empty is unlikely to hold our interest, the great majority of texts will contain a message of some sort. So I will take form as the primary marker in identifying a text, but meaning—the message—as the overwhelmingly likely motive for the creation of a text.

The etymology of the word text, ‘something woven’ (from the past participle of the Latin verb texere, ‘to weave’), refers to a characteristic of textuality that might be defined as a “coherent whole” (Nöth 1990:332). For Danesi (1999:6), “A text is, literally, a ‘weaving together’ of the elements taken from a specific code in order to communicate something.” For example, “When someone says something to someone or writes a letter, he or she is engaged in making a verbal text; when someone selects clothing items to dress for an occasion, he or she is making a bodily text...” The notion of a woven, that is, coherent, product is pivotal in the consideration of dreams as texts.

Is the dream a text? My response is a qualified “yes” which is dependent upon a particular use of the term dream. Texts have boundaries; as products separated from the stream of continuous experience, they allow us to focus on them extracted from other phenomena. By my use of the term dream I mean ‘the dream once we have experienced it’. To say “I had a dream” is to acknowledge an isolatable experience with temporal boundaries that begins where we can recall it beginning, and it ends where we can recall it ending. Through criteria of form we recognize a dream as a subjective text, and if the formal properties of texts are their defining factor, then the question of whether or not the dream conveys a message can be treated separately. As a discrete unit of form, the experienced dream coheres and so can be considered a text.

Semantically, the dream may or may not seem to cohere, although a lack of semantic coherence may be a result of not knowing how to read the dream text, to paraphrase Jung (1974:97). It seems reasonable to suppose that if dreams are indeed texts as I am asserting, then they generally have a message-based content; this follows from the observation that meaningless texts would rarely attract our interest, and clearly many people are intensely interested in the presumed meaning of their own dream-texts. But of course this view rests on my argument that experienced dreams are texts. Moreover, I do not limit the term message to the sense of ‘lesson’, but rather accept the use of message for any event sequence that can broadly be said to constitute the plot of the dream.

The dream while it is being dreamed is experience, not text. Our memory of that experience, whether we report is or not, is the text of the dream. So there is something between the initial experience of the dream and the dream report, just as there is something between waking experience and our report of that experience. The report in both cases is a text; the experience itself is not. The experience becomes a text once it is a completed product; we recognize it as a cohesive phenomenon bounded in space and time, having form as well as content. So the dream becomes a text the moment the initial experience of it has ended, just as a waking experience can become a text as soon as we are able to reflect on it as “something that happened” to us.

The more difficult question concerns the relation of the experienced dream to the dream report, and which of them is the appropriate object of inquiry into dream textuality and narrativity. That the dream report is a text seems uncontroversial. The report takes form through a medium, whether visual (pictorial), or verbal (spoken or written). It has a beginning, middle, and end —formally speaking, it has spatiotemporal boundaries. It also has cohesive links: in a verbal report we find, for example, the use of anaphoric pronouns (we, her...), adverbials (there, suddenly...), and other common devices of textuality. There are often near-formulaic statements, e.g., for beginnings, “It starts out with...”, endings, “...And that’s all I can remember” and even midstream transitions, “Then, suddenly, the scene shifted”.

Obviously, when we work with dreams, we can only work with the report of a dream, which is an objective product. But we are doing the same when we report a waking experience. In order to share any experience, the experiencer must encode it in a representational system such as language, which is both enabling and limiting in its capacity for full and accurate representation. But we don’t discount the value of a report because it isn’t identical to the experience; rather, we accept reports of waking experience as simply the best means available for representing and relating them. We can think of dream reports in the same way. If a report of waking experience can be considered a workable representation of that experience, then the dream report can analogously be considered a workable representation of the dreamer’s subjective experience of the dream. The report—and from here on I will be considering only verbal (spoken or written) dream reports—is the best means we have for representing our dream experience in a way that allows it to be anchored in time and communicated to others. We can hardly produce a dream report, moreover, if the experienced dream is not already a unit of text.

It is of interest to note here Jung’s apparent lack of concern about the gap between the experienced dream text and the dream report. In his argument against Freud’s manifest dream being a facade for the latent dream, Jung seems to presuppose the adequacy of the dream report as a valid object of textual inquiry:

The “manifest” dream-picture is the dream itself and contains the whole meaning of the dream... What Freud calls the “dream-facade” is the dream’s obscurity, and this is really only a projection of our own lack of understanding. We say that the dream has a false front only because we fail to see into it. We would do better to say that we are dealing with something like a text that is unintelligible not because it has a facade—a text has no facade—but simply because we cannot read it. We do not have to get behind such a text, but must first learn to read it. (Jung 1974:97)

Support for the adequacy of reports also comes from laboratory evidence. Kramer (1993:157-8), for example, provides a summary of work that justifies reliance on dream reports as sufficiently faithful representations of the dream:

...experimental laboratory research has provided support, it not confirmation, to the idea that there is significant similarity between the dream experience and the dream report (Taub, Kramer, Arand, and Jacobs, 1978). Eye movements and dream action during REM sleep are relatable (Roffwarg, Dement, Muzio, and Fisher, 1962). The intensity of the psychological experience during REM sleep and the dream report of that experience covary (Kramer et al., 1975). Experiments in which stimuli presented during sleep are incorporated into dreams suggest a relationship between the dream experience and the dream report (Kramer, Kinney, and Scharf, 1983).²

Although no dream report can ever be as satisfying as direct access to the dream itself, the latter is impossible, and the former is what there is to work with, and that in abundance. Freud’s notion of secondary revision notwithstanding (discussed in the following section), it may be impossible to know precisely what modifications the reporting process imposes on the experienced dream. So it is necessary to begin, at least, by examining dream reports for evidence of textuality and narrativity in the experienced dream, as indirect an approach as this may be. In addition to positing that experienced dreams are texts, I consider the verbal dream report to be a particular form of text which transforms the experienced-dream text into an objective product, analogously to the way a given story can be told through different media. For the remainder of this paper, then, I will employ the term dream text in reference to both the experienced dream and the verbal report of that dream. At the very least, from studying verbal dream reports we can learn something about textuality and narrativity as mental processes that provide frameworks for the representation of experience. At best, however, we may come to better understand the creation and structuring of unconscious dream texts.

I now consider whether dream texts can properly be called narratives. Perhaps no discussion of narrative can proceed without being prefaced by Roland Barthes’ well-known statement:

Numberless are the world’s narratives. First of all in a prodigious variety of genres, themselves distributed among different substances, as if any material were appropriate for man to entrust his stories to it: narrative can be supported by articulated speech, oral or written, by image, fixed or moving, by gesture, and by the organized mixture of all these substances; it is present in myth, legend, fable, tale, tragedy, comedy, epic, history, pantomime, painting..., stained-glass window, cinema, comic book, news item, conversation. (Barthes 1994:95)

Various scholars of narratology have contributed what amounts to a definition of narrative. I cite only several, selected on the basis of their potential relevance to the structure of dreams. Toolan (1988:7), for example offers what he terms a “first attempt at a minimalist definition of narrative” as “a perceived sequence of non-randomly connected events”.

Sarbin (1986:9) defines narrative as “a way of organizing episodes, actions, and accounts of actions; it is an achievement that brings together mundane facts and fantastic creations; time and place are incorporated.” For Prince (1982:1), narrative “may be defined as the representation of real or fictive events and situations in a time sequence.” And among the best-known minimalist statements on narrative is surely that of Todorov, who defined narrative as “the shift from one equilibrium to another . . ., separated by a period of imbalance” (1986:328). This statement merits attention, having been a source of disagreement with respect to dreams in States (1988:152):

Dreams do not proceed on such a structure, partly because they do not (at least narratively) begin and end. A dream seems to be a steady disequilibrium, with no functional or thematic interest in solving or rounding out a problem. The narrative of the dream is concerned with ramifications of a tension, ... not with getting me into trouble (or pleasure) and out of it, but with extending the trouble (or pleasure) to the boundaries of the feeling that produced the dream.

While this statement may be characterize certain dreams, it relies on a narrow concept of narrative appropriate to the study of literary texts.³ It also conflicts with Jung’s formulation of the structure of the average dream (for which, see below). But as with the term text, once again it is a broad conception of narrative which permits us to include dreams reports as a type of narrative, for as I will illustrate, dream reports do represent events and situations in a sequence, bringing together both mundane facts and fantastic creations into a perceived sequence of non-randomly connected events, which are characterizable as the shift from one equilibrium to another. It is also worth noting that Barthes (1994:147) saw the dream in contrast to the classical narrative because it is “removed from the logico-temporal order,” while the classical narrative is readable because it contains a sequence of events narrated in an irreversible (logico-temporal) order; yet through the ages the narrative “subverts itself (modernizes itself) by intensifying in its general structure the work of reversibility.”

Before turning to the basic elements of narrative, mention may be made here of the problem of reliance on dream reports in connection with Freud’s secondary revision. Although Freud was concerned not with the gap between the remembered dream and the report of that dream but rather with the “manifest” dream and the supposed latent thoughts that lay concealed behind it, his notion of secondary revision is nevertheless relevant to the question of the dream’s relation to the dream report. Freud wrote that

...we should disregard the apparent coherence between a dream’s constituents as an unessential illusion, and ... trace back the origin of each of its elements on its own account.... a psychical force is at work in dreams which creates this apparent connectedness, which ... submits the material produced by the dream-work to a ‘secondary revision.’ (1900:486)

Freud further explained this process as “the psychical activity which, though it does not appear to accompany the construction of dreams invariably, yet, whenever it does so, is concerned to fuse together elements in a dream which are of a disparate origin into a whole which shall make sense and be without contradiction.” (Freud 1900:496). He identified secondary revision with waking thought, because according to him it behaves in the same way, establishing order in perceptual material, setting up relations in it, and making it “conform to our expectations of an intelligible whole” (Freud 1900:537).

In short, Freud seemed to recognize that the same text-making, narratizing capacity that helps us make sense of waking experience is operative in the dream, and that the dream only achieves an impression of coherence thanks to this capacity. He also assumed that secondary revision operates simultaneously with the other aspects of dream formation (condensation, representability, censorship), and not subsequent to them (Freud 1900:537), which lends support to my claim that text-making, narratizing processes are at work as the dream is created; they are not simply imposed by the linguistic constraints of the dream report.

This view would seem to contradict Hartmann’s (1996:12) statement that in a dream “what is experienced generally is images...”, which leads in turn to the caution that “though we are often forced to work with verbal dream reports we need to keep in mind that these are only attempts to render the dream experience in a preservable and reproducible form.” This warning notwithstanding, I am proceeding under the assumption that there is a narratizing principle at work in the dream formation process which helps to organize perceptual material into a coherent text. I suggest that, while it may be true that “the potential to narratize dreams is as surely wired into the human brain as is the potential to speak language” (Foulkes 1982:276), narratizing is integral to the formation of many dreams, and narrative form is further enhanced and intensified when the dream text is rendered as a verbal report.

Many dream texts would seem to lend themselves to analysis along the lines developed by narratologists for other sorts of texts. The basic elements of narrative have been put forward by, for example, Chatman (1978), in reference to Todorov’s formulation of the structuralist view of narrative in the 1960s (cf. Barthes 1994:102). Chatman divides narratives into two basic parts, story and discourse.

A story (histoire) [is] the content or chain of events (actions, happenings), plus what may be called the existents (characters, items of setting); and a discourse (discours) ... [is] the expression, the means by which the content is communicated. In simple terms, the story is the what in a narrative that is depicted, discourse is the how. (Chatman 1978:19)

Narrative discourse, in turn, consists of the narrative form (narrative voice, point of view, etc.) and the manifestation of the narrative in a specific medium (Chatman 1978:22).

I will keep to Chatman’s basic formulation of story and discourse, incorporating into it Bal’s (1985) usage of the terms event, actors and to act: “An event is the transition from one state to another state. Actors are agents that perform actions. They are not necessarily human. To act is defined ... as to cause or to experience an event.” (Bal 1985: 5)

Half a century ago, Jung provided a succinct formulation of the structure of the average dream, borrowing from the classification of the elements of dramatic plot found in Aristotle’s Poetics. My focus here will be on the story component of narrative rather than the discourse, and what I will explore will be a blend of ideas gleaned from Jung, Chatman and Bal. Jung identified an opening Exposition phase, consisting of a statement of place, and a statement about the protagonists, and less frequently a statement of time; this phase also often indicates the dreamer’s initial situation. This first phase is followed by the Development of the plot, in which tension develops and the situation in the dream becomes complicated. In the third phase there is a Culmination (peripeteia), in which something decisive occurs or changes completely. The final phase is the Solution or Result (lysis), which shows the final situation; this phase is sometimes lacking (Jung 1974:80-1).

The adequacy of Jung’s formulation for dream reports and its relation to the elements of narrative just mentioned will be taken up in the following section. Here it may be said that Jung’s Exposition phase corresponds roughly to a presentation of the existents of narrative structure (setting, actors) and Jung’s remaining phases, Development, Culmination, and Result, are all aspects of the content of narrative (events acted by actors). In addition to the term actor for agents that perform actions in dreams, I will use the term character to encompass, additionally, animate beings who appear as presences in dreams without acting.

I now turn to analysis of dream texts. I have chosen data from my dream journal in order to investigate Jung’s proposed narrative structure of dreams.⁴ First I will look at a set of texts in order to see how well the Exposition phase, i.e. the identification of existents, characterizes the texts in the set. Next I will look at the narrative structure of a complete dream text.

Beginning with the Exposition phase, which may consist of the spatial and possibly temporal setting, the characters involved, and the initial situation, my journal contains numerous reports which begin by situating the dreamer in a certain location, although some are quite simple, others more elaborate. For example:

1. I walk into a large old hall of a college building. It seems to be Steven Pinker’s office, or a classroom where he teaches.

3. In a house with others (someone else’s). On a couch in a den with J. to my left and a man to his left.

4. I am in a meadow at twilight with J. It is huge, open, green, with gently rolling hills at the edges. There is a pedestrian tunnel or passageway.

5. I come out of a building at dark, early evening in winter, onto a slippery backyard slope with someone else.

6. I am lying on a table, like a hospital operating table, in a room with cognitive science and linguistics faculty.

7. At a show of some kind, like the Spectrum or other indoor amphitheatre. I sit next to J. in a sideways-facing mezzanine level.

9. I am in a large house/building with others. A room with a huge scorpion running around.

11. I am with some man--J., V., a composite, or someone–. We arrive at a museum parking lot.

As diverse as the surface features of these settings described by these statements may be, each one describes the dreamer’s initial location. Sometimes the opening statement incorporates other characters in the dream. In (4), for example, the first sentence consists of a statement of physical setting, temporal setting, and actors. The first sentence in (3), by contrast, situates the dreamer in a location that includes other people who are not individualized, although it is made clear that this is not the dreamer’s domicile. The second sentence specifies the dreamer’s position in the setting, and includes information about the dreamer’s position with respect to other actors, now depicted as individuals. In (5) the initial sentence shows the dreamer and an unnamed character actually entering upon the scene, which is described in terms of both time and place.

Occasionally the introduction of actors precedes that of location, as in e.g., (11). Finally, sometimes the opening sentence describes location, characters, and a situation, all three, as in (12).

The second (and third) sentence of the Exposition phase of these dream texts may provide further description of that place, e.g. (4), “It is huge, open, green, with gently rolling hills at the edges. There is a pedestrian tunnel or passageway”, or it may, additionally, incorporate a situation in progress, as in (9), “A room with a huge scorpion running around”.

Almost as easily found in my journal, however, are dream reports in which characters and setting seem to be of secondary or no importance, for the dream text begins directly with a statement of the situation, occasionally incorporating the dreamer as actor. For example:

With someone else driving, we go to meet this person at a designated spot on the highway. It appears to be San Francisco-like, and we have to drive onto a bridge that is partly –mostly– underwater. Several cars have either been wrecked or stranded off to the left.

The kids and I go out to see where, and after some confused searching, turning in this direction and that, I see the fire blazing on top of a (suburban) hill, burning down a house, I think. Later we seem to be closer to it and I see some kind of glass office building in ruins, post-fire smouldering.

The flight is for 5 p.m. tomorrow, and I want to confirm it. I keep looking in the stack of papers inside my top drawer, but I can’t see close up anymore, so I don’t find it --perhaps because I just keep missing seeing it.

For some reason Mickey Hart (the only drummer for the Dead in the dream), is sick or otherwise unavailable, and I have been selected to play in his stead.

The stainless steel, I think, but especially my cast iron skillet, which more closely resembles V.’s iron pan. There is a wavy, “bite-sized” chip in it about an inch or so long.

In (13) and (14), subsequent sentences do provide information on setting and characters, although in (14) they are incorporated into an elaboration of the situation. In (15), the opening statement of situation is followed by one of explanation for the activity, which is followed in turn by a sentence providing minimal information on the dreamer’s location (before the top drawer). In (16), the dream report opens with a statement of the dreamer’s situation, which is followed by an explanation for this state of affairs, recognized as unusual even in the dream. Setting and finally characters are introduced in this dream only after several more sentences which expand upon the initial situation.

The dream text of (17), presented in its brief entirety above, consists of a static image of a set of objects. Aside from the dreamer as witness to the image, there are no characters and there is no setting or action, merely this one situation.

To summarize, there are considerable possibilities for variation within the Exposition phase which nevertheless preserve its structure. I found no examples of dreams that did not have any of the elements identified by Jung as pertaining to this phase within the first few statements of the text; even dreams such as (17) which consisted of a single image had at least the element of a situation in some, albeit unidentified, space. The initial statements of dream reports seem to fit Chafe’s statement on narratives in general --they “give evidence that the mind has a need for orientation in terms of space, time, social context, and ongoing events” (Chafe 1990:97), this evidence being that narratives typically begin with “a statement of the particular place, time, characters, and background activity against which the events of the narrative proper then unfold” (Chafe 1990:94). Or, to use Chatman’s term, this orientation is achieved through a statement of existents.

Turning now to the remaining phases of the dream as suggested by Jung (Development,

Culmination, and Solution or Result), I present a dream report in its entirety in order to investigate the presence of these phases and their narrative elements. The numbering of the statements is my division into the phases.⁵ The Exposition phase of this dream text, consisting of a statement of place and a statement about the protagonists, was introduced as (7); it is repeated here as (18a):

18a. At a show of some kind, like the Spectrum or other indoor amphitheatre. I sit next to J. in a sideways-facing mezzanine level.

18b. During intermission, I go out and realize the music can be heard out here. Standing around one of the entrances, I sing along with the music –it sounds like the Grateful Dead, at least vaguely. I note with someone passing from out to in how you can hear the show outside [it’s in some big city, but which ?], but we agree we’d prefer to be inside.

18c. It looks like it’s E., and he adds, as he passes me going in, how its only five dollars anyway. I think for a moment, because I paid more, and realize that was for reserved seats. Did I really need them?

18d. I make my way back to my seat. J. already seated in his. A woman like G. [from UT] comes along and we try to figure out which is her seat, since she’s trying to sit in mine.

In (18a) there is both a statement of setting (indoor amphitheatre), and a statement of actors (dreamer and dreamer’s spouse), as well as further information on the situation in progress (a show of some kind) and of the actors’ specific position in the amphitheatre: not only are they at the mezzanine level of this amphitheatre, but they are facing sideways. Thus the Exposition positions the actors as spectators to the performance of some ongoing action, yet they are not facing this action directly, but rather indirectly, i.e. sideways (possibly a pun that they are not “facing the music” with respect to a situation).

The next, Development, phase of the dream I have identified as (18b). Something is supposed to occur; tension develops and the situation becomes complicated in the Development phase. The development in the present text appears to be rather ordinary. The dreamer leaves her initial position and exits the structure, a shift of position which causes her to become an “outsider” after having been an “insider” adjacent to a spouse. The dreamer is still participating in the action of the performance from this exterior position, not as a spectator facing the stage sideways, but as an actor in the open air, with greater freedom of movement, who from this fringe position actually participates in the action of the show. Using terms from narratology, the dream text depicts actors acting in events at this phase.

The Culmination phase of this dream, which would see something decisive occurring or changing suddenly, I have identified as (18c). The dreamer is speaking here with another “outsider” who is on his way “in”. This brief exchange of talk introduces a puzzle into the narrative. The dreamer learns that some people, at least, paid less for their seats than she did. Why did she pay more? The question seems to be resolved upon the realization that the dreamer’s seats were reserved. But a further concern is then introduced: it seems now as though it might not have been necessary to pay extra for reserved seats, although no definitive resolution about this is offered at this stage. But the matter of reserved seating seems to be pivotal, for a doubt has been introduced. Was the dreamer duped through naivety? Perhaps she has merely been overly cautious. Or it may be that the purchase of reserved seating was wise. Perhaps the dreamer is more vulnerable than others to the loss of her seat and needs the added protection of reserved seating, or perhaps she simply had to pay a higher price than others to guarantee her place. Events at this phase allow for these questions to be posed, although they remain unanswered here.

The final part of the dream presents the Solution or Result, identified as (18d). The dreamer has left her seat, exited the building, learned something about seating in the amphitheater, and now returns to her seat with this new information. To find someone else occupying one’s seat presents a response to the question of whether or not it was necessary to pay for reserved seating. For here is someone from the dreamer’s past in her seat, and only her reserved ticket proves that the dreamer, and not this acquaintance from her graduate school days, is entitled to occupy this particular seat. So it was a good idea to have a reserved seat after all.

This dream, while not the stuff of high drama, nevertheless conforms to the general characterization of narrative as “the representation of ... events and situations in a time sequence” (Prince) with actors acting and experiencing events. It also conforms to Jung’s phases and can be said to have an identifiable beginning, middle, and end, wherein, following orientation to the existents, there is a development of the situation, a problem identified, a question or puzzle to which the problem gives rise, and a response to that puzzle, resulting ultimately in a “shift from one equilibrium to another” (Todorov).

Although many additional dream reports I have reviewed also seem to conform to this suggested narrative pattern, not all do. Short dream fragments, for example, seem to resemble snapshots more than stories. If some dream reports have narrative structure, but others do not, then it is not the case that we inevitably impose narrative form on dreams in order to relate them, since we are capable of reporting dreams that do not conform to this structure. Recall that for Freud, secondary revision “does not appear to accompany the construction of dreams invariably” (Freud 1900:496). This in turn lends support to the view that our dream reports may be fair representations of our dreams after all.

First, I hope to have demonstrated convincingly that the dream-once-dreamed is a text, one that often but not always has narrative structure. The dream text is made up of strands of disparate thoughts weaving their way through our minds as we sleep. The loom which allows these threads to be woven together is a narratizing process that is part of human cognition. The dream report is also a text which, as a representation, will reflect the narrative structure of the experienced dream or its absence. But as a linguistic representation, the verbal dream report will also enhance the narrative organization of the dream text. An analogy may be made with the structuring that occurs when we express our waking thoughts in spoken or written language: the content of ideas is in essence the same whether it is expressed or only thought, but expressing thoughts through linguistic representation structures them in specific ways.

Further, dream texts will vary in their degree of narrativity, ranging from fragmentary snapshot to epic tale.⁶ Some remembered dreams seem less coherent than others, but it is still the case that we remember “a dream”, no matter how bizarre or tenuous the connection between images seems to the waking mind. So, we do not necessarily impose narrative structure on the dream, since we are as capable of reporting a dream with narrative form as without.

For a dream to have narrative form is necessarily for it to have structured content, but not necessarily to have a message in the sense of a lesson.⁷ However minimally, simple action sequences which begin and end and are ordered with respect to each other, involving characters, settings, and usually the tension of a dilemma, do count as narratives. What we usually call a story is a product of having imposed conventional narrative structures on selected content. But I don’t see that dreams are “trying to become stories” (Hunt 1989:177) so much as they are rudimentary, unrefined stories, a shifting “from one equilibrium to another” (Todorov).

My second point has to do with the relation of narrative structure to language. Two possibilities present themselves concerning the nature of this relationship. One is that narratizing is a cognitive process, but not a linguistic one. This seems to be the position of Turner (1996), for example, who claims that “narrative imagining—story—is the fundamental instrument of thought”, that “it is our chief means of looking into the future, of predicting, of planning, and of explaining” and that it is a capacity “indispensable to human cognition generally” (pp. 4-5). Turner speculates that “the linguistic mind is a consequence and subcategory of the literary mind.” (p.141) That is, “the ability to recognize and execute small spatial stories” (p.25) precedes linguistic expression.

The other possibility is that narratizing is a linguistic process, which means that all narratives are the product of verbal thought, and even phenomena such as mime acts and wordless comic strips, while legitimately qualifying as narratives, can nevertheless only be generated and in turn interpreted by language users. I am not here committing to one or the other possibility concerning the role of language in the narrative process, but I mention the issue because it is relevant to my final point.

My final point concerns the relation of language to dreaming, a complex issue which can only be mentioned here (see Kilroe 1999b). If narratizing is a cognitive process independent of language, then narrative-like dreams are not necessarily generated by language. But if narratizing is a linguistic process, then dreams with a narrative quality are either essentially unconscious verbal thoughts illustrated by imagery, or they are a weave of unconscious presentational imagery and representational verbal thought (cf. Hunt 1989).

So one type of dream may consist of a sequence of images illustrating unconscious verbal thoughts, which are the subliminal continuation of the mind chatter that we experience while awake. Not all dreams are necessarily structured this way, but those that are can often be matched to verbally established metaphors, puns, and other linguistic phenomena (Kilroe 1999a). And although dream images may be drawn from and recombined out of nonverbal perceptions stored in memory, it would still be verbal thought that motivates these images to come together to form a text.

From this point of view, disparate strands of verbal thought make up the dreamer’s mental discourse, from scattered impressions to focused preoccupations. These strands are unconsciously woven into narrative form while we sleep, resulting in a discourse of imagery that forms a dream text which we report, usually verbally, as our dream. Opacity between the dream report and the unconscious verbal thoughts prevents the straightforward retrieval of the latter into consciousness.⁸ It is like trying to tell a story from the illustrations alone and not having access to the verbal text that gave rise to the illustrations in the first place —we may be correct in saying what the author’s story is, or partially correct, or simply mistaken.

In any case, whether dreams are generated by linguistic processes or by nonlinguistic cognitive processes, the study of the form of dreams helps to show that narrative structure is not an artful invention but rather a natural process of the mind.

Bal, M. (1985). Narratology: Introduction to the theory of narrative. Toronto: University of Toronto Press.

Barthes, R. (1994). The semiotic challenge. Berkeley and Los Angeles: University of California Press. Trans. Richard Howard. [Originally published in French as L’aventure sémiologique, 1985, Paris, Éditions du Seuil.]

Chafe, W. (1990). Some things that narratives tell us about the mind. In B.K. Britton & A.D. Pellegrini (Eds.), Narrative thought and narrative language (pp. 79-98). Hillsdale, NJ: Lawrence Erlbaum.

Chatman, S. (1978). Story and discourse: Narrative structure in fiction and film. Ithaca and London: Cornell University Press.

Corti, M. (1978). An introduction to literary semiotics. Bloomington: Indiana University Press.

Danesi, M. (1999). Of cigarettes, high heels, and other interesting things. New York: St. Martin’s Press.

Freud, S. (1900). The Interpretation of dreams. Trans. J. Strachey, 1965. New York: Avon Books.

Hartmann, E. (1996). Outline for a theory on the nature and functions of dreaming. Dreaming, vol. 6, no. 2.

Hunt, H. (1989). The multiplicity of dreams: Memory, imagination, and consciousness. New Haven: Yale University Press.

Jung, C. G.(1974). Dreams. From The collected works of C.G. Jung, vols. 4,8,12,16. Trans. R.F.C. Hull. Princeton: Princeton University Press.

Kilroe, P. (1999a). What is a play on words without words? Consider the dream pun. Ms. under review.

Kilroe, P. (1999b). The role of language in dreaming. Paper presented through the University of Louisiana-Lafayette Mind & Matter colloquium series, November 5, 1999.

Kilroe, P. (1998). Freudian artifacts and imprints of imagination: Signification in language and dreaming, presented at the 23^rd annual meeting of the Semiotic Society of America, Toronto, October 18, 1998.

Kilroe, P. (1997). Toward a semiotics of dreams. Paper presented at the 22^nd annual meeting of the Semiotic Society of America, Louisville, Kentucky, October 24, 1997.

Kramer, M. (1993). Dream translation: An approach to understanding dreams. In G. Delaney

(Ed.), New directions in dream interpretation (pp.155-194). Albany: SUNY Press.

Kramer, M., L. Kinney, and M. Scharf. 1983. Dream incorporation and dream function. In W. Koella (Ed.), Sleep 1982: The 6^th European congress on sleep research, Zurich, 1982 (pp. 369-71). Basel: Karger.

Kramer, M., T. Roth, and J. Czaya. (1975). Dream development within a REM period. In P. Levin and W. Koella (Eds.), Sleep 1974: The 2^nd European congress on sleep research, Rome,1974 (pp. 406-08). Basel: Karger.

Nöth, W. (1990). Handbook of semiotics. Bloomington: Indiana University Press.

Prince, G. (1982). Narratology: The form and functioning of narrative. Berlin: Mouton.

Roffwarg, H., W. Dement, J. Muzio, and C. Fisher. (1962). “Dream imagery: Relationship to rapid

Sarbin, T.R. (1986). The narrative as a root metaphor for psychology. In T.R. Sarbin (Ed.), Narrative psychology: The storied nature of human conduct (pp.3-21). New York: Praeger.

Solms, M. (1997). The neuropsychology of dreams. Mahwah, NJ: Lawrence Erlbaum.

States, B. (1993). Dreaming and storytelling. Ithaca and London: Cornell University Press.

States, B. (1988). The rhetoric of dreams. Ithaca and London: Cornell University Press.

Taub, J., M. Kramer, D. Arand, and G. Jacobs. (1978). Nightmare dreams and nightmare

Todorov, T. (1986). Structural analysis of narrative. In R.C. Davis (Ed.), Contemporary literary criticism: Modernism through poststructuralism. New York: Longman.

Toolan, M. J. (1988). Narrative: A critical linguistic introduction. London and New York: Routledge.

Turner, M. (1996). The literary mind: The origins of thought and language. New York: Oxford University Press.

Ullman, M. (1969). Dreaming as metaphor in motion. Archives of General Psychiatry, 21, 696-703.

Uspenskij, B.A., et al. (1973). Theses on the semiotic study of cultures. In J. van der Eng & M. Grygar (Eds.), Structure of texts and semiotics of culture (pp. 1-28).The Hague: Mouton.

¹Thanks go to the three anonymous reviewers of this article for their numerous helpful comments, which among other things prompted me to rethink the concepts of text and narrative central to this paper. Whatever errors and obscurities remain are of course due to no one’s shortcomings but my own.

² But cf., e.g., Solms (1997), which scrutinizes the relation between dreaming and REM sleep.

³ Also cf. the discussion in States (1993:75) of how unlike stories dreams are, e.g. that dream reports “do not read like stories; on the contrary, most dream reports preserve the main structural features of dream narrative, including the in medias res beginning, the unexplained gaps and shifts in scene, and the inconclusive ending.” Here I take States’ remarks as a plea against the dream being comparable to a literary text, but presupposing some kind of narrative form for dreams.

⁴ Despite my reluctance to expose personal material, the ready access these data permit to a network of associations in my waking life is a compelling reason to draw from them. I refer to myself as “the dreamer” in this analysis because in looking at my dream texts, I do not feel identical to the “I” of the dream so much as represented by a dream character who seems to be “me”.

⁵ Although it has no direct bearing on the question of narrative structure, readers interested in the story content of the dream may find it useful to know that at the time of the dream, I had recently begun a full-time faculty position at a university in a Southern state, after teaching part-time in the Midwest since graduate school (UT). I was also about to be divorced from someone (J.) whose academic career had taken precedence over mine and as a result was the main reason I had remained for years in what I considered a one-down position of university adjunct. E. is a person of brief, distant acquaintance to me who had recently become a professor in a field where training is in demand by students but which I consider academically “soft”, so in my opinion E. found a faculty position with less struggle than I did. G. is an acquaintance from my graduate school days who at the time was apparently content to be an adjunct lecturer supporting her husband, a graduate student preparing for an academic career.

⁶ Possibly the degree of narrativity in dreams is connected with whether dreams are experienced in REM or non-REM sleep (see Hartmann 1996 on REM sleep as the source of most or our memorable dreams). Or perhaps the relative dominance of left or right brain hemispheric functions during dreaming can account for varying degrees of dream narrativity (see discussion in Hunt 1989).

⁷ Alternatively, a dream-text may have multiple messages, and what Corti has said of the literary text may apply equally to the dream-text: “Every text is many texts in that the very nature of its polysemic complexity prevents identically repetitive readings even in the same cultural context” (1978:42). And it is the interpreter of a text who at least partially constructs its meaning and supplies its coherence.

⁸ But note that while this echoes Freud in agreeing that dreams are formed from thoughts of which we are unaware, it is not identical to claiming that the “true” dream consists of latent thoughts that both motivate and hide behind a manifest dream. Following Jung, I take the dream-as-dreamed to be a valid object of study. This does not prevent us from inquiring into the source of dreams, however, including, in this case, the role of verbal thinking in dream formation.