Dreaming, Vol. 13 No. 1, March 2003.

Dreams, Art and Virtual Worldmaking


Bert O. States1

This paper examines the possible role of dreams and other forms of virtual worldmaking (chiefly fictions) in forming and maintaining our adaptive systems. I posit no exclusive function for the dream. Rather, I treat it as an extension of fiction’s preoccupation with our daily concerns, desires and fears. I suggest that narratives help us to enlarge and revise our perceptual and response systems, not by offering us moral or ethical propositions to live by but by increasing certain skills in our mental organization. Departing from John Paulos’ idea that fictions and mathematics (narratives and numbers) work in similar ways, I further examine the role that probability ratios might play in dreams, despite the seeming bizarreness of many dreams. The overall idea is that narratives of all sorts are one cognitive means, among many, by which we accumulate "sums" of knowledge and expectation, and maintain and revise our notions of what goes with what in human experience. I also look briefly at fictional archetypes (Oedipus, Orestes/Hamlet, etc.) and universal dreams (falling, being lost or attacked, etc.) as master plots in our probability systems.

KEY WORDS: narrative; probability ratios; adaptive structures; archetypes; universal dreams.

1Correspondence should be directed to Bert States, 5514 Camino Contigo, Santa Barbara, CA 93111; e-mail: bnstates@cox.net


In general, I share Wittgenstein’s idea, in his critique of Freud, that there is no reason to think that dreams have a single purpose or an "essence," given that there are so many different kinds of dreams. After all, he argues, there is no essence to talking or playing, so why should dreams be any different? (1966, 47-50) However, a good deal of our commentary still treats the dream as one might treat a mysterious body organ; that is, it must be doing something for us because it’s there, but right now we don’t know what this something is.

We have many theories, of course (e.g., dreams process memory, act as sentinels, fine-tune our vigilance/fear systems, reduce tension, create counterfactual simulations, perform self-therapy, self-formation or neuronal dumping, rehearse our survival and predation routines, and so on). But we have no measuring devices for determining how—or if—these so-called functions influence our behavior because we dream.

Still, such functions must be commonly carried out, consciously and unconsciously, in waking life, and it would therefore seem natural that they occur in our dreams as well. Since it is widely agreed that dreams reflect our daily concerns, perhaps our efforts to form better adaptive structures come into dreams automatically with the concerns. If this is true, however, in what sense can we claim that dreams have a special purpose? This is not to suggest that they don’t, or to dismiss our theories as invalid, but that the issue of function might better be approached at what we might call a pan-cognitive level. By this I mean simply that there seems to be a "state continuum" in mentation, as Hartmann et al. have suggested, "running from focused waking thought (e.g. doing arithmetic problems) to looser thought (e.g. reverie) to daydreaming and finally to dreaming" (2001,103; see also Domhoff, 2001, 19). If this is the case, perhaps some of these functions attributed to the dream run through the entire spectrum of thought. Therefore, "closed" studies of dream function, studies that concentrate solely on the dream, run the risk of attributing to the dream what belongs to thought in general.2

What is obscured in such cases, it seems to me, is the dream’s place in the continuity of mind and consciousness and its possible cooperative value with other forms of virtual worldmaking, which presumably have deep adaptive value to the species. I am thinking of virtual worlds as imaginatively conceived worlds that bear at least a metaphorical relationship to what we call "the real world"; in short, reverie, art of various kinds, dreams, tribal history, mythology, gossip, etc. This is where Wittgenstein’s argument may be in need of qualification: the validity of attributing a purpose to the dream might depend on how wide a cognitive net you cast in conducting the search. Here I want to cast a rather wide one in the context of the dream’s relation with art. My purpose is not to advance a new theory of dreams or of art, but to see how these possible functions might be interrelated at a basic level.

Let me concentrate here on fictions, which I will treat as a synecdoche for art in general. Fictions are particularly apt because they share so many structural and narrative features with dreams—more so, say, than painting, music or sculpture. We have long debated the question of why fictions appeal to us, and why we create them. It has been the prevailing idea since Horace, if not before, that poetry is both useful and pleasing (utile et dulce), and there have been countless notions of what these terms mean. In the light of modern evolutionary theory, this may even be a slightly redundant notion, if you consider the biological possibility that pleasure and usefulness may at times be aspects of the same benefit. Take sex, for example. To be more specific: the mind delights in such things as discovery, symmetry, contrast, parallelism, resemblance, the detection of causal series, reversal, crescendo, and so on. But all of these delights of mind are obviously indispensable to survival as well, if only because they keep us mindful of how the things and events of our experience resemble and differ from each other (see States, 1998). So it is possible that delight and usefulness are sometimes two different ways of looking at the same thing. In saying this I am not equating delight and usefulness. Many things may be fun to do (fast driving, drugs, rich food, etc.) but are not good for you. At any rate, I suspect that we continue to say that poetry has two functions—to delight and to instruct—only because we have no single term for the wholesale thing it does for us.

But note that usefulness at this level is not usefulness in the specific sense that Wittgenstein was probably referring to—e.g., the heart’s "job" is to circulate oxygen in the blood or the dream’s (in Freud’s view) is to be the guardian of sleep. At best all these useful delights seem an evolutionary outgrowth of mind’s encounter with experience. If art is powerfully grounded in them it suggests that we make art for much the same reason that we go through the day seeing parallels and contrasts in things, telling parts from wholes, and vice versa, feeling unity stirring in our experience (coherence), detecting order or disorder in a series of events, and watching the sunset (crescendo). Everything that interests us in art, everything that compels our attention, has its variant at most levels of our experience. Consider how statements like "I thought the characters were well drawn" or "I thought the ending seemed tacked on" are embedded in our absorbed notions of behavior and likelihood formed as much outside of art as through its conventions.

It would seem, then, that art is a kind of cognitive "looking glass," to invoke an old renaissance metaphor: it concentrates and enlarges our perceptual and response systems, creating an intensity of feeling and involvement that doesn’t normally occur in more distributed experience. Northrop Frye, for example, suggests that the instructive power of literature lies "not in illustrating moral precepts, but [in] expanding the power of vision" (2001-02, 18). This isn’t a particularly mind-catching idea, given the Romantic connotations of a word like vision. However, we can see its cogency better if we put it beside a recent variation of the same thought by evolutionary psychologists John Tooby and Leda Cosmides: "The kind of truth conveyed in art is not propositional or referential in the ordinary sense. It consists of the increased mental organization that our minds extract from experiencing it, which is why this form of truth has seemed so elusive, so difficult to articulate or explicitly define. This organization consists mostly of what might, for want of a better word, be called skills: skills of understanding and skills of valuing, skills of feeling and skills of perceiving, skills of knowing and skills of moving. . . . The truth inheres in what the experience builds in us" (2001, 24).

Thus, for Tooby and Cosmides, fictions are valuable or functional "because the mind detects that such bundles of representations have a powerfully organizing effect on our neurocognitive adaptations, even though the representations are not literally true" (21). So, apart from its noisier variants (propaganda, "position" literature, erotic novels, How to books, etc.) literature is useful in this quiet enduring way. We don’t see its long-lasting effects on people, and we can’t separate people who don’t read fictions from those who do at the "skills" level because there are still other ways to come by the same skills. Literature’s contents are imprinted and re-imprinted in our cognitive systems in the way that our linguistic skills are quietly improved by using the same words over and over and applying them to different things as new situations emerge. The "elusive" truth Tooby and Cosmides speak of is born out in the very act of reading (or making) fictions, in expanding our fund of "instances" of human behavior, and not in simply making us better "educated" or smarter people. That may indeed be an additional value, but the business of literature is "increased mental organization": better personal survival equipment for a world that is constantly mutating, even as it remains much the same. Or that is the way an evolutionist might put it.

Obviously, you don’t deliberately pick up a book to satisfy this need, nor do authors write them for that reason. But if you read, say, War and Peace, you can’t escape strengthening or changing your notions about war, peace, love, marriage, the beauty of hay fields and oak trees, friendship, national pride, and death. Or, if you read a number of those horror stories called tragedies you begin, among other things, to understand more clearly the connection between ego and disaster, and so on. Truth of this sort is simply what passes into literature from life experience; it is what we expect to find when we read, what delights us when we find it, and what we reject when literature’s imaginary "lies" become falsehoods when reflected against the facts of life.

While thinking about this problem, I happened to read John Allen Paulos’s Once Upon a Number (1998), a book devoted to the mathematical logic of storytelling. We live, Paulos says, according to systems of probability, that is, things that will "conditionally" happen or not happen on the basis of past performance. But, he goes on, our subjective probability ratios arise in "idiosyncratic ways": "we differ in the way we attach rough probabilities to happenings, and differ even more in the probabilities we assign to the associations between happenings." Our network of probability estimates forms a "map of our minds and dynamically interacts with new experiences and old stories that we constantly edit" (69). Paulos examines a number of ways in which fictions and mathematics (narratives and numbers) work in similar ways, despite their many formal differences. We need numbering systems for counting the things of the world; we need stories for counting the tendencies of the world in the sphere of human action. Stories, in short, are one cognitive means, among others, by which we accumulate "sums" of knowledge and expectation. In fact, without knowing it, we constantly use such things as Bayes’ probability theorem to maintain and revise our estimates of what goes with what, and when, as new experience blends with old. Simply thinking about something involves a kind of math or a sub-network of probabilities on which the thinking is based.

It would seem circular to say that the function of thinking is to maintain and revise our probability ratios, because the very construction of thought—at least non-hallucinatory thought—is itself based on the constant manipulation of probabilities. As T. S. Eliot’s Sweeney says, "I gotta use words when I talk to you." So too, we "gotta" use probability ratios when we think. Even when we listen quietly to music we delight (and profit by) the modulations and fulfillment of probabilities of key, harmony and melody. Or, take two common thoughts: "I suppose Sally has spoken to Richard by now" and "I wonder if Sally wants to speak to Richard." Both are garden-variety thoughts, and they are the tips of deeper understandings based directly on "statistical" samplings of Sally’s probable behavior in either case. "We are all statisticians," Paulos says, "when we make grand inferences about a person from that tiny sample of behavior known as a first impression" (11).3

It should be stressed, however, that fictions are not routine adventures in the probabilities of daily life. In fact, the distinctive tendency of fictions is that they take improbable, if not outrageous, events (a son marrying his mother, a son and daughter murdering their mother, a father sacrificing his beloved son or daughter) and show, as Aristotle said, how such events occur "according to probability or necessity." Thus the horrific extremities of fiction are made plausible by an intricate nest of likelihoods (It is probable that a man like Hamlet would have difficulty murdering his uncle-father; it is probable that Romeo would go to Mantua and that Juliet would awaken too late in the tomb; it is probable that Cordelia would refuse to flatter her father and that Lear, who has "but slenderly known himself," would be outraged by the lapse, and so on).

This, of course, is another aspect of the utile and the dulce of fictions coming together: unlikely or distant fears (incest, infanticide, revenge) are nevertheless fears, and one can see how extreme "cases" would have a value in exposing us to the danger of improbable disaster (the "What if . . . ? or "Worst case scenario" principle) in a probable way. It is not a fear that we might suffer the fate of Oedipus or Medea; what matters is the manner in which the unlooked for comes about in a probable way, a "lesson" we ignore at our peril. At the same time they are delightful and fascinating because they bring into focus the full force of our fondness for "perfect" developments which "add up to" perfect reversals (the symmetry of opposite extremes). Thus a "good book" is one that sets its stakes as high as possible, produces an engrossing development, and fulfills its promise with maximum involvement of all its parts. The same expectations we have in music, ballet, sculpture, an elegant machine, and art in general (I explore this idea in States 2001).

In short, there is little to exercise the human mind in acts that simply produce further acts, however probable (went down town + had a few beers + came home and slept). The kinds of plots to "look for," as Aristotle said, are those that produce counter-acts, terminal discrepancies involving close kin, where the stakes are at their highest. These are the most thorough stories in the world, the very cutting edge of disaster. Only in this way can art approximate the full potentialities of life, as opposed to its mundane norms. In this respect fictions are also directly related to our fondness for jokes and riddles which allow us the momentary pleasure of seeing how two seemingly incompatible trains of thought can be justified in a single "punch line" (see Koestler, 1969). Thus humor and tragedy, so different in their "attitudes" about the probabilities of life, perform essentially the same service in sharpening our cognitive awareness.

I hasten to repeat that I’m not trying to sneak in a new theory through the back door. There is nothing new about probability theory. I am trying to find the lowest common denominator of our advanced theories, or some of them, something that stretches across the continuum of thought, as a means of illustrating a point about the elusiveness—indeed the virtual invisibility—of function. The point is that art, like a cat, may have nine lives, or do "nine" different things, and there is more than one way to skin a cat, but at bottom the cat gets along by knowing the probability ratios of its environment. For humans, this is a more complex problem, if only because we make all sorts of representations—comic, tragic, frightening, soothing, abstract, etc.—of our probability systems, and it is not easy to see how our need for these things fits directly, or indirectly, into the scheme of our survival. In any case, I find myself in agreement with Nelson Goodman, in Ways of Worldmaking, who feels that all virtual worlds created by our species are, so to speak, created equal in that they contribute to our understanding of the so-called "actual" world: "the arts must be taken no less seriously than the sciences as modes of discovery, creation, and enlargement of knowledge in the broad sense of advancement of the understanding, and thus . . . the philosophy of art should be conceived as an integral part of metaphysics and epistemology" (1978, 102).

What I have said about fictional narratives and probability ratios must surely hold true for the dream as well. I want to sample dream probability at two levels, the macro and the micro—that of the narrative and that of the image. Dream narratives are notoriously given to peripeties, or to worst-case realizations. Get anywhere near a cliff, an open staircase, city traffic, a dark street, deep water, a large (or small) animal, try to back up your car, run from a beast, find a bathroom, make a speech to a group, take a test, find your way back to your hotel, avoid someone who has just entered the room, and you have major trouble. One assumes the reason this is true—or at least more probable than not—is that in life itself safety lies in anticipation of the possible, and in the dream state, owing to the peculiar simultaneity of thought and image, the arousal of an expectation almost guarantees its arrival. In a dream an expectation is not simply hypothetical ("What if the beast sees me?"); the hypothesis immediately condenses into its own proof. There are exceptions to this idea, to be sure. The dreamer might be rescued or find a way out (as in comedy or tragicomedy), or the threat might inexplicably disappear. But apparently in such cases an emotional factor weighs in—a mood shift, a surge of confidence or fearlessness perhaps—that alters the expectation, and hence the outcome. My evidence here consists largely of personal instances in which I have suddenly become courageous in the dream and confronted the threat head-on, often making a friend of my enemy (another reversal!). However, my record in achieving such triumphs of will is dwarfed by the number of times I have taken the coward’s way out and have shaken myself out of the dream in a cold sweat. To put it simply: the outcome of a dream seems to depend not on logic of narrative—roughly the logic of events in the waking world or of literary fictions—but on something like an emotional quantum: if I’m really terrified by the monster wave approaching my frail boat it will wipe me out (awakening me in the process); if I’m only, say, 75% frightened, I’ll get a heavy harmless shower; if I find myself awed or intrigued by the wave (the feeling of the sublime) I’ll so sailing over the top of it in ecstasy, hoping the ride never ends. It all depends on the psychical attitude that accompanies the threat. In short, you get out of a dream exactly what you put into it.

At any rate, the pronounced negativity of dream content is a commonplace in dream study. It would seem that negative experience, as opposed to happy birthdays, weddings, and career successes, is the dream’s natural specialty, and that dreaming, as I have suggested elsewhere, is the mental capability most clearly adapted to concerns arising from our condition of mutability (1993, 179), or the continuous disequilibrium of life. In brief, the dream presents the conceivable in terms of the real.

In a recent contribution to dream theory William Domhoff writes that dream content "is generally continuous with waking conceptions and contains a great deal of previously unrealized repetition in characters, social interactions, misfortunes, negative emotions, and themes" (2001,14). The repetitions that concern Domhoff pertain largely to repetitions within an individual’s dream history. But there is a sense in which all dreamers dream each other’s dreams in the form of so-called universal dreams, which are the oneiric equivalent of literary archetypes. We have always admired the Greeks because they seem to have created all the great plots. And so they did, though it is more accurate to say that they discovered (in Greek terms of course) the only plots available to an oxygen-dependent species that reproduces itself sexually and forms families of offspring who live in social groups called communities or tribes that are governed by sometimes willful or fallible leaders. That is to say, stories involving the jealousies, ambitions, intrigues, passions, fears and hatreds of fathers, mothers, sons and daughters, husbands and wives, uncles, aunts, neighbors and rulers. Even the gods in Greek mythology are little more than people with better weapons. Indeed, it is almost impossible to avoid the so-called archetypal stories because they constitute the ur-plots of our collective experience, each new variation confirming the truths of its avatars at the same time that it makes necessary cultural revisions. Thus the Orestes story begins in Greek tragedy, gets passed up in countless folk story variations to Saxo Grammaticus, Belleforest and Shakespeare’s Hamlet, and we find it alive and well in Dostoevsky’s 19th century St. Petersburg (Notes from Underground) and in Camus’s twentieth-century "Rome" (Caligula). The faces change, the object of the vengeance changes, but the structure of revenge remains intact, primarily because revenge, in one form or another, is a continual tendency of our species (see States, 1980).

So too, dreams have their archetypes, which is simply to say that there are certain dreams that belong as much to our species as to individuals.4 These are often referred to as universal dreams, simply because everybody seems to have them. Again, it is hard to see how it could have been otherwise, for if fictions are related to what matters to the individual then universal dreams were as inevitable as allergies, diseases, and disasters caused by inattention, nature, predators, etc. They are, in short, responses to living in an environment that bristles with assaults on the human being. No one wants to fall from a bridge into a raging river; no one wants to be alone, ignored by the world, humiliated, caught naked, unprepared, lost, or paralyzed before an oncoming menace. And because we are inescapably susceptible to these things, we dream about them, and such dreams the world over vary only in the specifics of the individual’s experiences in a particular culture or environment.

Perhaps the biggest difference between fictions and dreams in this regard is that fictional archetypes are, in most cases, addressed to the community and feature an exemplary individual (family, group) caught in a recurrent situation (mother-father revenge, child sacrifice, the exile of the gifted pariah, the confrontation of family versus civil law) and hence are as inevitable as the angle or the arc in geometry. The universal dream, on the other hand, has no communal "message," and indeed typically features the individual dreamer in hazardous situations in which nature or culture seem to play the role of nemesis. Patricia Garfield (2001, xvi) lists these as: being chased or attacked, falling or drowning, being lost or trapped, being naked, being injured, ill or dying, being in natural or man-made disasters, performing poorly on tests, having car or vehicle trouble, missing your boat or plane, property lost or damaged, machine or telephone malfunction, and being menaced by a spirit. In my own experience, these occur more as motifs than full dreams, often with several appearing in one dream (if I am caught naked in a crowd, I am not likely to make a clean getaway to my hotel room, much less by able to find it). The important point about universal dreams, in any case, is that they represent the main fears of our species, and therefore require no more explanation for their occurrence than a "collective unconscious" is needed to account for the persistence of literary archetypes.

The question about probability becomes more complicated when we leave the dream narrative and look at the single dream image (if there is such a thing). One is tempted to disqualify dream images because they are rife with improbabilities (mother morphing into Boris Karloff). But I suggest that the probabilities may not reside simply in the normalcy or "realism" of the images but, again, in the connotative and emotional systems that have aroused the image. An image is not simply itself but a "bundle" of associations. Indeed, dream images are so visible, so apparent, that we forget they are multi-sensory productions. There is, of course, no way to prove it, but perhaps the so-called improbabilities of dreams are a sign of their power to conflate different kinds of relationships and similarities—what Tooby and Cosmides call "abstract isomorphisms" (2001)—at a level too far down in the neural cellar to be detected. Why Aunt Tina should arrive at the party on a child’s tricycle (surprising no one) may not be possible to determine, but it may not be illogical either. The usual way of explaining an image of this kind is to find a connection in the dreamer’s prior experience (Aunt Tina taught elementary school, Aunt Tina was childlike, etc.). But it is possible that tricycles and Aunt Tina have nothing to do with each other at this level and are being coupled through secondary potencies arising from the dream. Also we must assume that an image is born at the formative and unedited level of thought, where thought is sorting itself out of multiple associational possibilities gathered "idiosyncratically" in a single memory. In an edited composition there are few traces of such images, though any number of them might have occurred in the composition process; in a dream they are precisely what dream-thought is producing at the moment. It would be remarkable if a dream flowed along like a real story without shifts, mergers, loose ends or conflations in imagery and character. Even if you claim that distortion occurs as the result of brain modulation, it seems likely that a tricycle (if it has appeared) has a stronger relationship with the Aunt Tina figure (whatever she may "stand for") in the dream than, say, a bowling ball or a kayak (which did not appear). In short, even in disorder likelihoods persist that cannot be determined from an exterior standpoint, disorder being characteristic of all dynamic systems.

The dream might be compared, in this regard, to a three-dimensional chess game. Pieces move not only forward, backward, sidewise, or diagonally over 64 squares, as on a normal chess board, but up and down as well over eight levels of 64 squares, only the top level of which is visible to the observing eye. Perceived as game-events on a conventional board such moves would seem absurd: pieces would disappear from view; they would suddenly turn up on squares to which there was no visible move, and so on. The dreamwork resembles such a game in that the logic of its "moves" simply cannot be confined to the visible two-dimensional "brain board" (of, say, a dream report) or the sort of logic we have come to expect in fictions. The rule of thumb would be: if dreaming is thought as it unfolds, probability-seeking must occur at many sensory levels, some of which might appear contradictory on the level of waking ratios of coherence (for example, from my own file of dream "improbabilities": music suddenly becoming palpable, my thought as dreamer becoming the thought of another character, a living spirit emanating from an "innocent" household object, an ear that talks, an eye that listens, a problem solved without having been worked out, an image being simultaneously itself and something else as well, a dream character with alternating identities, etc.). Such things are normal in the dream world and they can be accounted for as the pure thrust of imagination seizing upon the resonances of its own productions.

Finally, there is a possibility that what is visualized in a dream image is itself the confluence of more than one train of thought. For example, the coupling may be a compromise of competing thoughts aroused by prior events in the dream: they don’t "belong" together, they simply happened to "come to mind" at the same time (as opposites attract) with equal probability values, in the way that something seemingly irrelevant often pops into waking thought. Elsewhere I have discussed the relation of dream bizarreness to inner thought (States, 2000), the main idea being that a good deal of thought (if we could "photograph" all its nuances) appears disorganized because it is exploratory—that is, a consequence of first-draft-ness, which consists basically of search-and-adopt or search-and-delete mechanisms involving many blind alleys. How can a medium like the dream be expected to get its ducks in a row without knowing in advance what the ducks are or where they are headed? To paraphrase Michael Gazzaniga, among the "zillions of automatic brain processes" that go on in any feat of cognition, "the mind is the last to know" (1998, 1).

To ignore these possibilities amounts to ignoring the immense neural delicacy and evanescence of mental representations and the incredible creative freedom of mind afforded by the dream state. It is perhaps this freedom to expand associational possibilities across sensory limits that makes the dream state an extremity in the spectrum of thought—not unique in its function, perhaps, but capable of sorting out resemblances and probabilities at a sub-narrative level. The question of what dreams do for us is clearly up in the air. But it seems plausible that if the mind couldn’t put Aunt Tina on a tricycle, or confuse mother with Boris Karloff, its creative capacity might be greatly reduced. Dream distortion, with or without chemical influence, may be a form of pretend play which maximizes our associative capability far beyond its expected needs, and allows us to remain flexible in the face of a new emergency.

I have often thought that artists like De Chirico, Delvaux, Picasso, Dali, and the surrealists appeal to us not because they remind us of our dreams but because such distortions of reality are innately attractive and in some way part of a micro-extension of our possibility ratios, coaxed out by metaphorical resemblances in the object (What if that animal were a human being? What if human chests could be opened, like drawers? Is not the human being a kind of geometric arrangement of parts?). It always fascinates me that in pre-sleep darkness my mind can conjure and alter a face into monstrous disproportions on the slightest commands. It is a fickle skill at best, lasting perhaps twenty seconds before the image collapses completely. But it always astonishes me because I have the unique impression that I am watching my mind playing with its own contents at an elementary level. Here is a living mechanism, itself nowhere visible, except in its products, that seems to be consumed by vicissitude and mutation, incapable of rest or of "still photography"; it is itself a morphing machine (another word for narrative), that seems to devour its own output and move on to new images with an insatiable need to transmogrify, and I can only imagine what concoctions it goes on producing while I’m away from its "screen" doing other things.

I doubt if these concoctions are valuable in themselves or get stored anywhere in the vast neural libraries of the mind for later reference. It seems more likely that the value lies in the versatility of mind itself, the openness of thought to any conceivable production, much as language is open to the production of any word-combination possible given the right incentives. Just as poets are constantly displacing our accepted uses of the words in our vocabulary into new arrangements ("in Just- spring when the world is mud-luscious . . .") so dreams are devoted—in fact, have no other choice in the matter—to pressing possibilities as far as they will go.

I am aware that this line of argument seems like an attempt to have my improbabilities and do away with them as well. Mainly I am trying to make a case for some sort of order in dream-thought that is not derived from our notions of order in the observable world. One sharp difference is that the dream is what we may call time-warped: in obliterating the differences between past, present and future, the "present" of dream time is itself a composite which projects associations rather than facts of experience into a moment of psychic life. Obviously, there is much to learn about thought processes in this regard. In the end it is probably more complex than I have made it seem. On the other side, I have recently read still another theory of the function of dreams. There is some experimental evidence that REM sleep keeps the corneas of the eyes moist during sleep. In short: we dream so that we won’t go blind (Vision, 1999, 23). Of course as we always say, more research is needed before we can reach firm conclusions.

In any case, cornea lubrication hardly accounts for the content of dreams. At best the cornea theory is a contribution to the physiology of sleep. On a whimsical note, however, it occurs to me that the corneas probably don’t prefer specific dream plots (say childhood sexual traumas or humiliation scenes); any sequence of images would supply the oxygen, the nutrients and the lubrication to keep the corneas going during the long dry spell in which they have nothing to look at. And the same may be true for the psychical value of dreaming: it may not matter what specific kinds of experience dreams process as long as they lubricate the neural mechanisms by which we continually relearn, among other things, what goes with what. But the mind, being unflaggingly devoted to selfish concerns, tells relevant stories that have to do with desiring and fearing, dealing with what has arrived and preparing for what may be to come. So, as Hamlet says, two crafts in one line directly meet. The eyes need the dream, and the dreaming brain seizes the splendid cinematic resources of the eye to present its theatre of the self to itself.

Perhaps we could make a similar claim for waking forms of creativity like fiction, art, daydreaming, storytelling, gossip, doing mathematics, and so on. Unfortunately, this amounts to saying that new experience of any sort teaches you what goes with what. But at the bottom of the thought process this surely must be the case: we learn by doing, and re-doing, and when necessary revising what we have learned to do. Dreaming and fiction, all forms of virtual worldmaking, may (among other things) be highly intensified forms of circuit maintenance: a sort of dry run of the neurons which allows us to have an experience in a safe place, as Ernest Hartmann puts it, rather than in a dangerous one. In other words, in the virtual world of dreams and fictions we get to drive off a cliff into the sea many times in a lifetime, whereas in the actual world we can do it only once. Precisely how this is useful to us, as I’ve said, is still an open question.

My concern here has been to suggest that whatever function dreaming may have it is not necessarily different from the function of other mind activities. Or at least we shouldn’t put dreaming and other forms of cognition apart before we understand what they have in common.


Domhoff, G. William (2001). A New Neurocognitive Theory of Dreams. Dreaming 11 (1), 13-33.

Frye, Northrop (2001-02). Northrop Frye Newsletter, 9 (1). Notebook 18. Roanoke College, Salem, VA.

Garfield, Patricia (2001). The Universal Dream Key: The 12 Most Common Dream Themes Around the World. New York: HarperCollins.

Gazzaniga, Michael S. (1998). The Mind’s Past. Berkeley: U. of California Press.

Goodman, Nelson (1978). Ways of Worldmaking. Indianapolis, Cambridge: Hackett.

Hartmann, Ernest, and Robert Kunzendorf, Rachel Rosen, Nancy Gazells Grace (2001). Contextualizing Images in Dreams and Daydreams. Dreaming 11 (2), 97-104.

Koestler, Arthur (1969). The Act of Creation. New York: Macmillan.

Lakoff, George, and Rafael E. Nunez (2000). Where Mathematics Comes From: How the Embodied Mind Brings Mathematics into Being. New York: Basic Books.

Paulos, John Allen (1998). Once Upon a Number: The Hidden Mathematical Logic of Stories. New York: Basic Books.

States. Bert O. (1980). The Persistence of the Archetype. Critical Inquiry, 7 (2), 333-344.

States, Bert O. (1993). Dreaming and Storytelling. Ithaca and London: Cornell University Press.

States, Bert O. (1998). Of Paradoxes and Tautologies. American Scholar (Winter), 51-66.

States, Bert O. (2000). Dream Bizarreness and Inner Thought. Dreaming, 10 (4), 179-92.

States, Bert O. (2001). Dreams: The Royal Road to Metaphor. Substance 94/95, 104-118.

Tooby, John, and Leda Cosmides (2001). Does Beauty Build Adapted Minds? Toward an Evolutionary Theory of Aesthetics, Fiction, and the Arts. Substance 94/95, 6-27.

Vision (1999). The Johns Hopkins White Papers. Baltimore: The Johns Hopkins Medical Institutions.

Wittgenstein, L. (1966). Lectures and Conversations on Aesthetics, Psychology and Religious Belief. Ed. Cyril Barrett. Berkeley and Los Angeles: University of California.


2 To give an example, in daily life we have all sorts of strategies for keeping our vigilance system up-to-date. Now suppose we dream of driving off a cliff, or being accosted by a gunman on a dark street. Does this mean that the dream has simply acted out a fear-fulfillment; or does it mean that the purpose of the dream—and of dreaming at large—is to anticipate such dangers, "practice" them, as it were, and thereby solidify our safeguards against them? Is the dream a contribution to our fear system or to our vigilance system? If the latter then we should ask wherein the dream is doing something that differs from that of other vigilance mechanisms and practices.

3 See also Lakoff and Nunez (2000) for an extensive discussion of the metaphorical basis of mathematics.

4 I am setting aside the Jungian notion of dream archetypes as the so-called "big" dream emanating from the collective unconscious.


Dreaming: Journal of the Association for the Study of Dreams. Vol 13(1)  March 2003.

  Copyright ©2003 Association for the Study of Dreams. All Rights Reserved