The Power of Movies

by Colin Mcginn

  I agree entirely with this point: the dream theory is not a complete theory. But it was never meant to be; it is just one part of an overall theory, which needs to take into account our visual relationship to the screen and the metaphysical meaning of the movie image. Obviously, we need something in order to be able to distinguish dreams from films— something that separates them, since they are clearly not the same thing. The point of the dream theory is just to insist that dreams and films are deeply united despite their formal differences. The dream theory captures a component of the movie experience, not its entire nature; for that we need to investigate also our perceptual and cognitive relation to the screen image. Summing up, somewhat crudely, we can say that movie watching involves looking into the screen image, finding there the dematerialized body, and undergoing an experience with many of the features of dreams. The total experience is an amalgam of these three elements.

  Six

  HOW TO MAKE A DREAM

  PRODUCTION

  Heretofore, the topic has been how we experience films and dreams once they have come into existence. In this chapter I shall switch attention to film and dream production—how films and dreams arrive in the world, what the gestation process is. We know quite a bit about film production— what the basic processes and components are—and I think these can shed light on what must lie behind the production of dreams. Dreams, in other words, are created in ways significantly analogous to the production of movies. Dream production can be illuminatingly compared to film production.

  Let me start from our normal, naive view of how dreams come into being. We suppose, I think, that they arise spontaneously, with little preparation, more or less at the time they are “viewed.” We assume that the brain manufactures them with minimal effort and in short order; they just “pop into our heads” as we sleep. All we consciously experience of dreams is the end product; whatever lies behind their production is hidden from conscious view. We therefore tend to suppose that nothing much is hidden—what we see is what there is. Compare ordinary visual perception: all we experience of this process is the end product, the conscious percept that occurs in our mind as a result of the brain's processing of incoming stimuli. At the most naive level we tend to think that the external object just makes its way into the mind, that it meets consciousness without any intermediate processing—rather as if consciousness were a box into which the object is placed. Certainly, our percepts arise without any conscious effort, apparently spontaneously, just as if the process were a simple matter of the object imprinting itself on the sensorium—a primitive stamping operation. But, of course, this naive view is just that: the visual process is far more complex than the deliverances of introspection suggest. First, the object itself does not strike the eye at all; it emits light rays that enter the eyeball and impinge on the retina. The image formed on the retina, which is fragmentary and two-dimensional, then stimulates the optic nerve, and an elaborate process of interpretation is set in motion, eventuating in a conscious visual percept. Vision scientists have studied this process, noting how complex the transformations of the retinal image must be, and how multiple the computations that lead to the final percept—which then gives the misleading impression of having just popped into consciousness from nowhere. The causal background to even the simplest percept is immensely complex and multi-layered—this much is now accepted fact.1 No one has any idea how to build a machine that can generate percepts in the way brains can—full-blown 3-D percepts from 2-D patterns of light striking the retina. In a way, this is a miraculous act of creation, not something like taking a fingerprint. It is easy to draw people's attention to the elementary facts of visual perception, because the eye is out there to be observed and people know what the retina is. It is clear enough that the visual system must go from a mere spot of electromagnetic energy on the retina to a rich conscious awareness of the external world. A sense of wonder at this process is therefore readily produced. But in the case of dream production the underlying apparatus is not out there for all to see; it is hidden in the recesses of the brain. The case for complexity must therefore be circumstantial. The best way to get a sense of this complexity is to compare what comes in through the senses with what comes out in the form of dreams—the input and output of the dream-production process. Dreams are, in some sense, based on sensory experience: the images they present are rooted in perception, and we dream about people and things that have been retained in memory from our perceptual encounters with them. When you have an image of a red truck, say, in your dream, this originated in experiences of red objects in the perceptible world, and dreams tend to be about familiar objects and situations. The input, then, consists of our ordinary perceptual experience. But the output far exceeds the input. Consider a dream I had only recently, which is still fresh in my mind; it is entirely typical of dreams generally, not especially ingenious or remarkable. I was back at my old Oxford college, in the senior common room chatting to some colleagues, just before dinner was due to start. Everyone was drinking—some drinking wine, some martinis, some whiskey. I couldn't manage to secure a drink for myself as I watched everyone else happily imbibing, and particularly coveted a martini I saw someone sipping. Finally, I managed to get hold of a carafe of wine, but all the glasses I could see had been used already, and I was embarrassed to be seen drinking out of someone else's glass—I must have examined fifty glasses of every conceivable shape and size, looking for a clean one. In the end I picked up a dirty glass, poured wine into it, and took a gulp—but it was some sort of strange cherry water, not wine at all. Disappointment. Everyone was now filing out to dinner, giving me funny looks. Then I realized I was wearing a scruffy sweatshirt, absurd striped pants, and a pair of muddy Wellington boots—to high-table dinner! I sought out the bathroom, hoping to make myself more presentable. As I entered the bathroom I saw that it had changed from the old days, being cavernous, full of fancy sculptures, and very expensively appointed—and yet absolutely filthy (I will spare you the details of what I saw in the toilet bowl). My eyes dwelling heavily on the besmirched porcelain of this fallen bathroom, I felt soiled myself.

  Now let us not trouble ourselves with hidden meanings and Freudian speculations; what I want to point out is how unprecedented all this was in my own sensory history. Every sight in the dream was fresh, newly minted: I have never owned a pair of pants like that (nor would I!); I have never been in a bathroom full of sculptures; that cherry drink was like nothing I've ever experienced. It was all a product of my brain's inner creative powers (and I have mentioned only a portion of the full content of the dream). In no way could we think of this dream as some sort of simple transcription from experience, a mere imprint of waking perception; it was a radical transformation of my old experiences at Corpus Christi College—which, by the way,

  I haven't visited in over ten years. I hadn't been thinking about Oxford recently, either—this dream just came “from the blue.” What must strike us, then, is the yawning chasm between input and output—between my waking experience and the dream content. Somehow the brain must cross this gap, delivering the end product from the initial materials. That is, the brain must be capable of remarkable creative feats in producing a dream like this. Nor could I, at will, consciously generate the same experiences in the form of a daydream—the detail, vividness, and sense of reality would not be there. The pants alone, with their lurid stripes, their air of effete dishevelment, were a masterpiece of artistic production—perfectly designed to humiliate their unfortunate wearer, the very zenith of sartorial inappropriateness. Whoever dressed me in those pants knew what he was doing. And I've never seen a martini looking as inviting as the one being poured just out of my reach for some gray-bearded professor or other.

  The first point to be made, then, is that dream production must involve some very elaborate processes of image and scene construction. Clearly, the intention of the dream was to make me feel frustrated and embarrassed—this was its emotional crux—and the sensory contents (the
awful pants, the inviting drink) were designed to bring this about. Thus the dream has a means-end structure—an emotional end and a sensory means. It exhibits intelligent design. The interweaving of the sensory and emotional (discussed in chapter 4) aids this design. The dream-production process is creative, purposeful, and resourceful. How should we conceptualize this process? What kinds of stages might it involve? How might it be analyzed?

  FILMMAKING

  Here is where I want to invoke the analogy with film. Film production can be broken down, roughly speaking, into the following stages: concept, script, green-lighting, casting, design, sets, filming, editing, and distribution. This division is not intended as definitive or uniquely correct; further subdivisions might be made, and I haven't said anything about raising the money or paying the actors and crew, not to mention catering and so on. The stages I have listed are basic to the process. Someone has to have an idea of what he wants to make a film about; someone has to put together a written script, more or less complete; someone has to agree to make the film, to put resources into it; the film has to be cast; design, sets, and makeup must be arranged; the actual filming must take place, the movie being stored on a medium that can later be activated; the editor must step in to shape the film; finally, the film must be distributed and viewed. The entire process can take months or years and is obviously very complex, requiring the interplay of many separate agencies. A naive viewer, such as a child, might suppose that the action just happened before the camera spontaneously, that it just sprang into being somehow (after all, many movie watchers don't even know what a director or producer is, and some confuse actors with characters). A film, however, is an elaborate artifact, a product of intelligent design; what eventually appears on the screen is just the end product of a complex process of construction. My question, then, is whether this constructive process can be usefully compared to the process of dream construction.

  I can only speculate here, since we know so little about how dreams are made, but I want to suggest that the film model provides a useful way to think about dream production. Let us then think of the dream machinery as a kind of team of separate agencies, each assigned its own job. The ideas man starts the process: he wants to make a dream revolving around a certain theme—say, my embarrassment at an academic gathering. Why the hell he would want to do that we don't know; we really don't understand why the brain generates dreams and why it generates the particular types of dreams it does. But it does, and the dreams seem to have recurrent themes—notably, anxiety and wish fulfillment. Perhaps the ideas man knows I am sensitive about this type ofthing and wants to torment me; or maybe he thinks it will be good therapy to let me get these feelings out. In any case, emotions seem to be the primary source of dream ideas. But an emotion without a script is pretty pointless, so some sort of narrative has to be constructed to express the emotion. Thus the concept is turned into a script, a specification of what the dream will contain. But converting the script into an actual dream uses brain resources, and other ideas and scripts are competing for attention (since I have many emotions that are seeking some sort of dream outlet). So some scripts might languish in the offices of the dream team as others are given a robustly sensory form. Suppose, though, that a particular script strikes the executives in the dream hierarchy as ready to be made, that the time is ripe for a dream like this. Perhaps the dreamer has been preoccupied with a certain emotion lately, or she has neglected a certain emotion. Then the script might be green-lighted, slated for production. Thus the script about me in an Oxford senior common room was green-lighted for production, with other competing scripts turned down or held in abeyance.

  The casting must begin (or maybe casting preceded green-lighting, if particularly colorful characters have already been penciled in). In the case of my dream, my twenty-four-year-old son, Bruno, was also with me (though I didn't mention this earlier); the gray-bearded professor had an especially smug and irritating look; and the bathroom attendant was a man who seemed both officious and downtrodden—his face spoke volumes. Each of these figures had a specific sensory form, a particular aura. One of the other members of the college present was someone I remember from the old days, not an invented character, and I recall him now as a little strict about college decorum. So each of these figures had to be selected to play a role described in the script—to give that role sensory presence. Then there is the matter of set design: everyone else seemed sensibly dressed (nice white jacket for the bathroom attendant), but clearly the scene won't work unless I am clad suitably outrageously— hence the wretched pants and Wellington boots. The director of the dream tells the wardrobe people to get something really horrible for me to wear, and they come up with tight striped pants from 1960s Carnaby Street. These matters having been settled, the dream can now be shot, with all the elements brought together. In other words, a record is laid down of how the dream images are to be sequenced when the dream occurs. But notice that this isn't the same as showing the dream. That is a matter for further deliberation, a question of timing and rival projects. Finally, a date is selected, the dream screened, and I wake up having had the dream I have been reporting.

  TIME LAG

  The last phase of the dream-production process requires further comment. As I described it, there was a delay between filming and distribution—between making the dream and showing it. Why did I introduce such a delay? Why isn't the dream shown while it is being shot? Indeed, why do I assume that the whole process is temporally extended? Couldn't everything from concept to distribution happen at the same moment? The model I am working with is that the brain contains a library of potential dreams, waiting to be shown on a given night, instead of dreams being projected off the cuff each night. But why do I think of the matter this way? The answer is twofold. First, the process is just too complex to be accomplished on the fly: the dream content is too intelligent and cunning to be brought about instantly. It would take me a while to consciously think up the dream I described; it is hard to believe that the dream work itself could proceed in a sudden leap forward. That dream took planning, and malice aforethought. Maybe it had been in the process of construction over a number of years, with many a tweak and twist to get it just right. There is really no reason to believe that the entire creative process behind the dream took place on the night of the dream, the moment before the dream entered my dreaming consciousness; it is more plausible to suppose that it was spread out over time, perhaps going back as far as my days in Oxford (certainly the images of Oxford seemed remarkably fresh). That dream had been in the works for a while.

  Secondly, the fact that dreams are so often rooted in the past suggests that they have been knocking around in our heads for some time. My dream about Oxford seemed like a remnant from the past, not just about the past. And why should my brain have created such a dream only just before its recent occurrence? Doesn't it make more sense to suppose that it created the dream at a time closer to the time of the experience being referred to? In fact, doesn't it make sense to suppose that the brain is continually creating dreams in response to the experiences of waking life? It takes an experience in, toys with it, tries out various dream scenarios, and maybe lays down a track for later use—all this happening unconsciously. The dream work is accordingly going on all the time, day and night, not just at the very time of dreaming. Dreams are a reworking of experience, so it makes sense to suppose that they are reworking it all the time. When a dream refers to the past, then, it is likely that at least some of the initial dream work was done at or around the time referred to. Some dream themes relating to Oxford may have been laid down in my unconscious years ago when I was still there; only now do they emerge into the light of night in the form of a conscious dream. As with film production, there is a considerable time lag between the early stages of production and the final viewing.

  A further piece of evidence for the hypothesis of time lag comes from empirical findings about the relation between the time of an experience and the time of dream
ing about it. J. Allan Hobson reports that the optimal time lag here is about six or seven days: it takes that long for a dream to be constructed that refers to an earlier waking experience (not the day before, as some had claimed).2 This suggests that the process requires such a time period: that's how long the brain needs to put together a decent dream. But we often dream of experiences that go back much farther, so presumably the dream work was initiated long ago. Thus our brains house a library of potential dreams, some still in production, not quite complete, and on a given night a decision has to be made about which dreams are going to take up valuable screening time. It is not that each night the brain enters the sleeping state and has no idea what dreams it will generate that night, hoping for inspiration as REM time approaches. No, it dips into its back catalogue. There is a scheduling of the dream calendar, with particular dreams tagged for release on given dates. The reasons for this timing are often obscure, though sometimes it is quite obvious why now is the time you are dreaming about such-and-such (as when you dream about an exam the day before you take one). However, it seems to me fantastically unlikely that dreams are cooked up on the spur of the moment, just when they are about to be unveiled, as if the dream machinery were idle for the rest of the time. It is far more plausible to suppose that dream creation is an ongoing activity, as much in the day as in the night. The unconscious is hard at work on its dream projects for much (most? all?) of the time.