Understanding what dreams are and interpreting what they mean has been a preoccupation of diverse cultures for millennia. The close relation between dreams and narratives is apparent and manifests itself in several ways: the use of dreams in literature; narrative reports of dreams; and dreams themselves as narratives.
The exact nature of the relation is unclear, though. If dreams are a form of hallucination, that is to say a delusory experience, where does narration come into the picture? Only retrospectively, in the dream report, or in memory? Or is the memory trace all there ever was of the dream? On the other hand, if dreaming is, or can be, an instance of narration, multiple questions arise: how do we understand the agency of narration in dreaming? What are the materials with which this narrating activity engages? What principles of coherence and intelligibility inform such activity—how is narrative sense achieved? What constitutes the medium of narration?
It is necessary to distinguish between the status of dreams as experiences, and so objects of narrative report, and the status of dreaming as itself a kind of narration. In the former capacity, dreams have been important in many periods and cultures, and their fascination has much to do with their characteristic resistance to the naturalized (but artificial) logic of narrative vraisemblance (Culler 1975: 131–60). At the same time, the suggestive power of dreams has been harnessed throughout the history of art and literature as a mode of meaning in which the semiotic force of dream events is foregrounded. Indeed, there are dream features that appear to be language-like or even predicated upon language, such as dream puns (Kilroe 2000).
Dream researchers tend nonetheless to maintain a distinction between the experiential dream and its signifying capacity, whether the latter is confined to the dream report (Marozza 2005) or credited to the memory of the dream, understood as the primary cognitive process in which the dream experience is interpreted, thus acquiring meaning and the status of text (Kilroe 2000). This “Cartesian theatre” model of dream experience has been critiqued by Dennett (1981), for whom the memory trace is all there ever was of the dream. The retrospective implication of a “memory” is not intrinsic to this view, the main force of which is to assert that the dream itself is no experience, but narrative. Whose narrative, then, and with regard to what? The notion of the dream as itself narrative appears to conflate perceptual consciousness of the “facts” of the dream with reflective consciousness about the dream.
In the Freudian model, the dream gives expression to prior, unconscious dream thoughts (Freud  1953). From a neurobiological perspective, however, there is no further regression of meaning, because dreams arise from the activation of the forebrain by periodic neuronal activity in the brain stem (Hobson & McCarley 1977). Such brain activity during sleep may be random or part of some adaptive process associated with that of sleep itself; the inception of dream mentation is just a by-product in this account. All the remarkable coherence of dreams is attributed to the mind’s subsequent cognitive efforts of synthesis, drawing upon the narrative sense-making capacities of waking life (Hobson 2002). Cognitive models of dreaming have more to say about the functioning of such sense-making processes, however. They too regard narrativizing as integral to the formation of dreams, but note that this should not be taken for granted; our storytelling capabilities develop in the course of childhood, and this development correlates with the development of children’s dreams (Foulkes 1999). Narrative logic, here, is not a given; instead, cognitive accounts foreground the creativity of dreams—their status, that is, not just as narratives but as fictions. Such approaches conceive the motive forces of dreaming as continuous with those of waking thought, whether the emphasis falls upon imaginative world-making (States 2003) or on the articulation of emotion (Hartmann 2010b).
In most cultures, the role of dreams has been spiritual or visionary. It is possible to distinguish between conceptions of the dream as experiential and the dream as meaningful (and therefore narrative); however, the two views often co-exist, and even cultures for which dreams occur in a real space (Ryan → Space) often regard them as a kind of thinking (Kracke 1992). This anthropological perspective is borne out in classical and biblical sources: Homer makes reference to the Greek personifications of dreams, the Oneiroi, in the Iliad (2.1–35); but in the Odyssey (19.560–69) he presents dreams as true or false narratives of future events, in the image of the gates of horn and ivory (Howatson ed. 1989). In the Bible the most common form of dreams is discursive (“God said to him in a dream”), but Jacob’s dream of the ladder (Gen. 28: 12) is a situated experience, if fraught with symbolism, and the major biblical dreams tend to be of this type: those of Nebuchadnezzar (Dan. 2) and Daniel (Dan. 7: 1–27) are obscurely portentous events in need of interpretation, while the events of Joseph’s dreams (Gen. 37: 5–10) are clearly legible for his brothers. Such fusions of experiential and symbolic concepts of dreaming are less surprising than they may seem, since it is only a materialist worldview (in which what exists simply is) that enforces this dichotomy. From a religious perspective, reality itself is charged with meaning, and dreams fit within the implied model of experience as itself a discursive medium.
Dreams have had a pervasive influence upon art and literature throughout history, and upon film from the very beginnings of its emergence as a narrative medium. Three examples will suffice here. The most prominent literary manifestation of the influence of dreams is the tradition of dream poetry in the Middle Ages following Le Roman de la Rose. Medieval dream vision poetry was a self-consciously literary genre, notably in Chaucer’s use of it, and in this respect dreams served as an inspirational model for imaginative fictional narrative (Spearing 1976). But dreams also functioned here as a motivational device for allegory, as in Piers Plowman and Pearl. In the dream vision tradition, dreams are more than a representational resource; they become a basis for understanding fictional narrative—to the extent that The Divine Comedy, for example, is read as a dream vision despite not being formally framed as a dream.
Dreams feature very prominently in the earliest years of film—especially as frame narratives—and not just as a topical psychological preoccupation (Marinelli 2006), but as part of early filmmakers’ self-conscious exploration and play with the affordances of the new medium. There are plausible grounds for thinking that the cognitive experience of dreams was crucial to early cinema’s transition from a cinema of visual attractions, founded upon the illusion of life, to a narrative cinema with its own conventions of storytelling (Gunning 1990). The influence of dreams may be discerned not only in foregrounded cinematic techniques such as montage, but also in early negotiations of the grammar of editing within the scene, the continuity of which we have now thoroughly naturalized. (Indeed, the question now may well be whether the conventions of filmic narrative have in turn begun to influence the form of our dreams.)
A third example is Surrealism, which in part follows on in the history of cinema (especially Buñuel), but is more broadly an encounter between dreams and the idea of art. The Surrealists’ interest in dreams was caught up with their interest in the Freudian unconscious, but the aesthetic concern behind both was the attempt to liberate the imagination from rationality. Dream logic appealed to Breton, Dali and others as a key to narrative creativity, to the primacy of the play of thought over social, moral and intellectual convention (Breton 1978).
Such appropriations of the formal qualities of dreams in art and literature correspond to a very widespread assumption about dreams themselves: that they require or invite interpretation. This assumption also cuts across any distinction between experiential and (framed) communicative models of dreaming, because it is a corollary of the recognition that the dream is not empirical fact. It makes a representational medium of the dreamer’s perceptual faculties, giving the dream the status of a text (Kilroe 2000). This calls into question the narratological consensus that equates mental representations with the non-discursive story level of narrative. More specifically, dream narratives are fictions—if we exclude possibilities such as literal foresight or a dream corresponding directly to a memory. The latter possibility seems substantial, but clinical evidence argues against it, even in the strong case of repetitive post-traumatic dreams, which consistently manifest a creative element (Hartmann 2010a). Dream interpretation, then, undertakes to motivate this fictive narration. As most modern dream research discounts any external communicant of the dream narrative (God, or some other kind of spiritual inspiration), the narration must be understood as the dreamer’s own, despite being typically characterized by novelty (to the dreamer, as against the wholesale recall of prior cognitions in episodic memory); by strangeness; and by the sense of a lack of control. Such a view of dream narration is suggestive for our understanding of narrative creativity in general.
The Freudian unconscious offered a royal road to dream interpretation, providing as it did for an expressive intentionality beyond the conscious frame of reference of the dreaming dreamer. The dream itself is for Freud a transformative articulation of prior unconscious dream thoughts (Freud  1953). This process of articulation—the dream-work—is a kind of negotiation between the unconscious and the constraints of, on the one hand, the censorship of consciousness and, on the other hand, the affordances of the perceptual medium of dreams (Freud  1953: IV–V, ch. 6). Of the four dream-work mechanisms that Freud identifies, two—condensation and displacement—bear mainly upon the symbolic potency of the manifest dream and do not directly bear upon its narrative form, although it should be noted that displacement, taken under the rubric of metonymy, has been accorded a central place in appropriations of the Freudian model to literary narrative (e.g. Brooks 1984; for the structuralist heritage of this connection, see Jakobson & Halle 1971: 90–2, and Lacan 1977: 146–78). Freud’s other two mechanisms—considerations of representability (the pressure of vraisemblance and the constraints of particularity imposed by a perceptual medium) and, especially, secondary revision (the dreamer’s efforts towards imposing global coherence and intelligibility upon the dream)—have a close relation to the typically narrative form of dreams. In this respect it is important to note that secondary revision, often invoked in relation to the narrative report of a dream, is for Freud a part of the dream-work itself; its secondariness bears upon the relation between the manifest dream and the (primary) latent dream thoughts (Freud  1953: V, 488–508).
The influence of literature upon Freud’s thinking is apparent here, as throughout his writings, and there is a risk of circularity in reading dreams according to literary protocols that may themselves owe much to the influence of dreams. Freud’s account of narrative creativity in dreaming, by invoking the agency of the unconscious, reinscribes in a covert form the transmissive model of narrative intentionality that dreams seem to problematize. For Jung, by contrast, the dream itself is a natural phenomenon in which consciousness attempts to find meaning (Marozza 2005: 697–98; Jung 1928–30). This move dissociates a psychoanalytic perspective from the specific agency of the Freudian unconscious so unappealing to scientific accounts of dreaming, but it also jettisons much of the suggestiveness, from a narratological point of view, of the dream-work. The general drift of post-Freudian thought, however, has been towards an emphasis upon the creative function of the dream-work (Marozza 2005), and to that extent there is some congruence between psychoanalytical and scientific approaches to dreaming.
The neurobiological account of dreams, the “activation-synthesis” model (Hobson & McCarley 1977), offers some answers to questions about the physiological causes of dreams and helps to specify the neurological conditions that define the characteristic qualities of dream mentation, as distinct from thought in waking consciousness (Hobson 2002). In the activation-synthesis model, dreams arise in the first place as a result of neuronal activity that occurs during REM and non-REM sleep, and which probably has (like sleep itself) an adaptive function. Such a function would relate to the ordering, updating and/or consolidating of the brain’s memory systems, and only incidentally intrudes into consciousness in the form of dream percepts (87–8). This is the ‘activation’ side of the model. The cognitive deficiencies of dream consciousness relative to waking consciousness, combined with the mind’s effort to impose coherence on initially chaotic perceptual images, result in an elaborative process that constitutes the “synthesis” component of the model, which Hobson conceives as a kind of confabulation (101) or, more broadly, narrative creativity. The mind’s cognitive sense-making efforts, in other words, are themselves progressively incorporated into the ongoing dream, allowing the dream to give a novel and emotionally significant coherence to its materials.
The focus of cognitive approaches to dreaming is this sense-making effort in itself, without reference to either a Freudian unconscious or to the neurobiological activation of dreaming. A cognitive perspective clarifies the relation between narration and experience in dreaming by distinguishing between volitional and non-volitional parts of dream mentation in terms of receptive consciousness and associational patterns; a double-mindedness, but not of the Cartesian homunculus variety (States 2000: 188). Such doubleness, combined with the perceptual medium of dreaming, explains a characteristic feedback loop: “in the dream state, owing to the peculiar simultaneity of thought and image, the arousal of an expectation almost guarantees its arrival” (States 2003: 7). These special conditions granted, the emphasis of cognitive accounts falls on the continuity between dreaming and waking thought, rather than on its cognitive deficiencies (4); the bizarreness of dreams is a reflection of the absence of the constraints upon thought characteristic of waking imagination (States 2000).
The foregoing has shown that narration is a relevant concept for dreams and that dream research affords some provocative insights into the process of narration, narrative sense and its connective logic, and the medium of narration in a cognitive context. It remains to draw out some of the implications for narratology that follow. Most obviously, dream research problematizes conventional models of narrative creativity. The standard communication model of narrative, or any model predicated upon a view of narrative as the transmission of a prior conception, cannot accommodate the case of dreams. The recursive process of elaboration in dreams, on the other hand, is suggestive as a model for the genesis of fictional narrative in general: it implies that narrative emerges from the particularization of emotions (or ideological concerns, or values) and the representational elaboration of those interests. Unlike the communication model, this account does not posit a pre-narrative meaning which the act of narration communicates, but rather takes narration itself to be the generative principle for meaning it bears as narrative. Such a model (call it the articulation model)—whether it understands narrative as emerging through an evaluative feedback loop or through a surrogate logic of representational particularization—accords well with novelists’ own accounts of the process of narrative creativity, which frequently emphasize a loss of originary creative control (Walsh 2007, ch. 7).
The equivalence of narrative content with the process of narration in dreams also calls into question the standard distinction between story and discourse, the “what” and the “how” of narrative (Chatman 1978); dreams may just be taken to compromise the universal applicability of that distinction, but must also call into question its adequacy to any form of narrative. The view that our cognitive-perceptual faculties are themselves the medium of narration in dreams disallows recent efforts to redeem the story-discourse distinction by claiming that story is conceptual and discourse is material (Shen 2005). The cognitive-perceptual medium of dreams also bears upon new media narratology and the representational status of simulations and virtual worlds. The contrast between experiential and narrative views of dreaming applies equally well in the context of a user’s interaction with a virtual environment: such an interaction may construe the simulation as a context for (virtual) behaviour; or the user may understand the simulation as representational, so that interaction with it is a semiotic activity. Only the latter involves narrative creativity and hence, properly speaking, interactive narrative.
The relation between simulation and narrative, or between worlds and narratives in general, illustrates the tension between systemic and narrative modes of understanding. Here, too, the case of dreams is significant. Dreaming involves a dynamic interaction “between linear and non-linear thought processes” (States 2000: 190) in which the representational logic of narrative sense-making constrains and stimulates the hyperconnective associational capacity of dream mentation. Dreams test the limits of narrative cognition, not as a struggle between sense and nonsense so much as between two incompatible kinds of sense—one sequential, the other systemic. This is not a problem confined to dreaming: temporal processes on all scales, across the range of disciplines, are generally better modelled in terms of the behaviour of complex systems than the sequential logic of narrative (the case of evolution by natural selection is a representative example). Yet as the form of our dreams also makes clear, narrative is not a mode of sense-making that we can shed or outgrow; it is a non-negotiable part of our cognitive heritage, and so it is only by being brought into relation with narrative that systemic phenomena become intelligible and acquire human meaning. A proper scepticism about narrative must therefore take the form of self-reflexive lucidity rather than abstinence.
Two areas of ongoing research bear upon the relation between the experience and narration of dreams, in complementary respects; both would benefit from the influence of a narratological perspective. The first is research into sleep behaviour disorders, in which persistence of muscle tone during REM sleep results in sporadic goal-oriented motor behaviour from the sleeping subject. For some researchers this disorder presents an opportunity to confirm the nature of the relation between REM and dream content, on the hypothesis that the eye movements have an experiential correlation with dream images (Leclair-Visonneau et al. 2010). The unexamined premise here appears to be a particularly literal notion of a Cartesian theatre of images for the eyes to scan—a topic that cries out for re-examination in terms of narrative cognition.
The second area of research is lucid dreaming, a topic with genuine interest that has been compromised by association with unduly speculative new age thinking. Dennett (1981) dismisses lucid dreams as recursive effects within the memory trace (you dreamt that you realized that you were dreaming), and States implicitly agrees, without subscribing to the broader no-experience explanation, by understanding lucidity in terms of an ontological distinction (dream world vs. real world) and doubting the possibility of “an independent in-sleep discovery that somehow rises above the curtain of sleep—during sleep” (States 2000: 189). There is lab-based research into lucid dreams that supports a different view, however, not only by cultivating experimental conditions in which it becomes possible for the dreamer to signal awareness of dreaming while still asleep, but also by intimating a more integrated notion of dream lucidity as lying on a continuous scale with other degrees of the reflective consciousness inherent in all dreaming (LaBerge & DeGracia 2000). Such an approach reaffirms the creativity of dream representations, offering the possibility of new insights into the nature of narrative fictionality and its affective power.