Dreaming Vol. 10,  No. 4, 2000


Bert O. States

I hope that by now the temptation to which we are all
subject, to contrast thinking with imagination is diminishing.
There remains, of course, the need to fix more definitely the
nature of the differences between relatively unimaginative and
relatively imaginative thinking
--Gilbert Ryle, On Thinking

As one who has been repeatedly critical of our approach to dream bizarreness (1993a,13-45; 1993b, 13-31; 1997, 125-31; 1998a, 1998b), I may seem to be kicking a dead unicorn in bringing up the topic again. But, as "the defining feature of many dreams" (Hunt et al, 1993, 198) bizarreness continues to be the chief descriptive element in our characterization of dream images and events. It may also, in the long run, be a key piece to the dream puzzle. Following the appearance of Owen Flanagan's Dreaming Souls (2000), I have had further thoughts on the subject which I offer here partly in the hope of arousing interest in certain complexities of the problem and partly by way of drawing some clearer continuities between dreams and other modes of thought. I am aware that a purely phenomenological orientation, like my own, may be slighting evidence of other kinds; but a limited point of view is sometimes useful in enabling us to see farther in one direction (at the expense of another) through a kind of foeval concentration. The phenomenologist looks directly at the experience (or tries to), the neuroscientist looks at the machinery beneath it, and I take it as a given that without both points of view our understanding would be incomplete.

Bizarreness, I have suggested, isn't really a corruption of thought; it is a way of thinking in its own right, though it goes by other names; and like all ways of thinking--including logic--it can be carried to extremes and become, as we say, bizarre. It is clear that dreams offer impossible conflations of things, people, places, events, and so on, and that some dreams are more bizarre in this regard than others. It is also true that people morph into other people in dreams far more frequently than they morph into animals or objects. So there seem to be some control conditions attached to bizarreness that survive the brain's descent into cholinergic modulation. As Rittenhouse, Stickgold, and Hobson suggest (1994, 100), there is a struggle between chaotic "bottom up" brainstem activation and "top-down" cortical attempts to make sense of the resulting disorder. If this is the case it would be useful to know the difference between compromises, or mutilated thought-corpses, produced by the "warring" processes (what we commonly call bizarreness) and other thought processes that resemble these compromises, my principal topic here. It is probably impossible to sort it out cleanly, given our limited access to what we actually see in the mind's eye--but I think some outside groundwork can be done along these lines, on the assumption that the brain that dreams is (more or less) the same brain that thinks.

The problem lies in our tendency to measure bizarreness against the probability ratios of waking life (what else?). In fact, Bonato et al, examined thirteen prominent definitions and scales of bizarreness and found that their only common denominator was a consistent relation of bizarreness to reality, their own scale defining reality as "natural Newtonian laws of wakefulness" (1991, 56). But I think there is a sense in which this amounts to a comparison of apples and triangles. Apples are real, triangles (apart from drafting and musical instruments) are usually conceptual or descriptive of anything with three sides (e.g., the Bermuda triangle). It isn't the triangle that's real but the thing that is shaped like one. You can even cut an apple into a three-sided shape and call it a triangle, but only in the same descriptive sense in which you might have called it a sphere before you cut it up. So there is hard core reality, so to speak, and there are "things" that pertain to reality in a descriptive way: a dream is one of these latter "things," a psychical prism, you might say, through which reality somehow gets refracted--as opposed to reflected.
A more common coupling is to compare dreams to other representations of reality. Thus you might claim that dreams tend to be less realistic than novels by Steinbeck or Hemingway and much closer to stories by Poe, Kafka or Lem, or paintings by Dali. But here again, the comparison is based on the same apples and triangles we used in the first instance: that is, we call Steinbeck and Hemingway "realists" because their novels observe Newtonian probability ratios better than Poe's or Kafka's. Hardly anyone would speak seriously of Poe or Kafka as being realists in the sense of accurately describing phantasmagoric or dream experience. Realism is a word we reserve for things we can see or expect to see in the conventional continuity of everyday life. So our conception of what constitutes a realistic representation is understandably biased at the source, and this bias continues into our third, and probably most common, practice: comparing dream mentation to waking mentation. As I see it, this is where the most interesting questions arise and where I think bizarreness can be useful to dream study as more than a descriptive tag we put on dream elements that have failed the reality test.

I can deal best with the questions by confining myself to characteristic examples from recent work on bizarreness. Obviously the literature on the subject is immense, a crude sampling of which would include Antrobus (1990), Crick and Mitchison (1986), Foulkes (1985), Hobson (1987 et al, 1988), Hall and Van de Castle (1966), Hunt (1982, 1989, 1993 et al), Meier (1993), Reinsel, Antrobus and Willman (1992), Revonsuo (1995), Revonsuo and Salmivalli (1985), Snyder (1970), and Strauch and Meier (1996). One could base a critique--at least the sort I will pursue here--on almost any attempt to isolate bizarreness from "normal" thought processing and treat it as if it were a measurable phenomenon. For convenience, my discussion will focus, first, on the Flanagan book because it most directly provoked my thinking, or re-thinking, on the topic, and as an up-to-date discussion it illustrates how persistently the idea of "scaled" bizarreness has worked its way into our study of dreams. (I hasten to add that Flanagan's book is a serious, philosophical-neurobiological theory of dreams, and what I have to say about it has no bearing on its central, and persuasive, argument that dreaming is probably an epiphenomenon, rather than an evolutionary adaptation.) The second example (Rittenhouse, Stickgold and Hobson, 1994) seems to me an attempt, among several by the Hobson group, to measure bizarreness in relation to what is probably the most widely cited and controversial neurocognitive dream theory, the Hobson-McCarley activation-synthesis hypothesis. Other studies might raise other issues, but I hope the concentration on these examples will gain in parsimony with only an incidental loss in range or nuance.

"Bizarreness," Flanagan writes, "will increase the more control mechanisms are turned down and the more you have on your mind. Quantity impacts quality. No crew of psychoanalytic dream disguisers need be posited to account for the bizarre character of adult dreams. Flanagan's Second Law of Dream Science is a corollary of the first: While awake we attend better than we do in sleep, and we apply more efficiently and reliably while awake than asleep criteria of relevance and common sense in taking the measure of things as they are presented, and in representing what is presented" (2000, 147).

Flanagan goes on to explain that bizarreness "can be measured according to a scale that plots incongruities, uncertainties, and discontinuities. For simplicity call the scale that measures dream bizarreness the IUD scale. Incongruity (I) refers to mismatches--the blue Caribbean waters viewed from the restaurant in Montreal; Socrates in a business suit. Uncertainty (U) refers to actual persons, things, and events that are not specified in the dream, one's geographical location, the person herself--maybe Beth or maybe Jane. Discontinuity (D) refers to an abnormal shift in person, place, or action--Clinton becomes Reagan; I am in New Jersey one second and I am with the same people in Paris in the next" (148).

My major difficulty arises from the notion that "wakeful mentation" is more "efficient and reliable" than dream mentation. One can easily grant the idea that while awake "we attend better than we do in sleep" to the world around us." But here again, I think the categories are skewed, if only in the assumption that efficiency, reliability, relevance and common sense--so essential in pragmatic goal-directed waking situations--are characteristic of human thinking at large and might also apply in dream situations, were it not for the disorder caused by the change in brain modulation. Again, the comparative bias is toward making sense, and the measurement of bizarreness begins in the dream's departures from this criterion as the overall "normal" state of brain functioning.

But what does it mean to say we behave less "efficiently" in dreams respecting (I assume) the events occurring in the dream world? The statement suggests that dreamers, or dreams themselves--and these are two different things--are deficient in efficiency and reliability. On the ground of the comparison, this is probably true; but are such qualities of behavior relevant to the dream world where the conditions of "actuality" have been radically changed--though unknown to the dreamer, for whom no other world exists? What could be the benefit of greater efficiency in the dream world? What might efficiency achieve? A happy ending? A life goal? A "dream" job? Or, on still another level, might "efficiency" be exerting itself in another way in dream thought?

One might argue that the real relevance of efficiency and reliability in the dream world is that they are behavioral qualities (among others) that are typically obfuscated by dream events. How can one be efficient (in the waking sense) in a world where car brakes unaccountably fail, hotel keys disappear when you need them, bridges collapse when you are crossing them, the dead come to life, and landmarks and people constantly shift their locations and identities? Well, one might say, "there's the proof, right there: there's the unreliability. You can't count on anything in a dream." But it does not close the matter to say this because there is a relentless consistency in the thoroughness with which dreams (or some dreams) are capable of arranging these frustrating reversals. This consistency, in fact, seems to me related to the consistency with which dreams do not turn people into trees or objects but into other people. In other words, there is something like a logic or a selective process at work in dreams. Bizarre dream images may, in some instances, be compared to slips in human speech production (e.g., "our queer old dean" for "Our dear old queen.") which are not simply inexplicable errors but the work of "unconscious control systems operating on internal feedback" (Sellen, 1992, 321; see also Foulkes, 1985, 60, 161). At any rate, it isn't simply a flawed logic caused by cholinergic inebriation but something verging on a reliable probability ratio, or, if you like, a reliable unreliability ratio. How else but by some sort of internally generated efficiency can we account for the persistence and dogged character of univeral dreams--chase/attack dreams, dreams of falling or drowning, poor test performance, naked dreams, being lost or trapped--dreams which are dreamt endlessly, in the same structure, all over the world (See Garfield, 1999, 25-6)?

Granted, cholinergic modulation may have created the "environmental" conditions behind this obsessive quality of dreams; but surely it is more than a product of the war between the brainstem and the cortex--unless one could show that the brainstem itself was diabolical. My inefficiency as dreamer seems to be caused not by an absence of this capacity (as dream protagonist) but by the "efficiency" of the dream (as involuntary production) in making sure that all my attempts to practice efficiency will come to naught (this is what Sartre referred to as the fatality of the dream: the sudden coming about of the thematic potential at stake in the dream (1968, 61-5). This is not a claim that all dreams behave so consistently but that considerations of a practical nature are very much a part of the dreamer's behavior and frustration: we want to survive, to obey codes, to avoid hazards, to reach goals in dreams. Moreover, how can one say we attend less efficiently in dreams when the things to be attended to aren't objectively occurring events, as in the waking world (where effect normally follows cause, and efficiency counts), but products of the dreamer's own associative thought process? This process is highly unstable in having both a feedback and feedforward effect on the narrative in progress. In fact, a dream image may be defined as the simultneous integration of all the neural events at work at a given instant. The so-called nonsense produced by dreams may not, in some cases, be the result of reduced cognitive competence but of "misplaced competence," in J. T. Reason's phrase (1984, 515), or the appearance of an image (object, person, place, etc.) that seems discontinuous in a dream report but was primed, as we say of cognitive error, by a "reminder" node in the dream context (the so-called "That reminds me" syndrome that commonly derails waking conversation). Even so, in the case of dreams, the term "misplaced" seems inaccurate because it refers to errors in thought and speech committed in everyday communication; the dream, in contrast, has no "place" to go because it is not governed, as far as we can tell, by an objective.

So waking efficiency and reliability, while thematically relevant in dreams, have nothing, or little, to do with dream logic. To assess dreams by these capacities may serve the purpose of differentiating them from the waking world (though I doubt it), but it does not tell us much about what sort of rules obtain in dream experience. To say that dreams exhibit less common sense, efficiency and reliability than waking thought seems rather like saying that objects in a vacuum fail to obey the rules for falling bodies in the open atmosphere, and ignoring the conditions responsible for the "failure."

The second text is the study mentioned above (Rittenhouse et al, 1994) which offers a detailed two-stage analysis of transformations of characters, objects, places, etc. taken from REM dream reports and submitted to six judges for evaluation of the amount of coherence or bizarreness. There were two main findings: character and object transformations (as mentioned above) "tend to occur only within classes and not between classes" (105). But "the most surprising and novel finding [was] that many discontinuous dream images are, paradoxically, coherent." For example, "A dream object does not transform randomly into another object, but into an object that shares formal associative qualities with the first." Thus, in one of the dreams, a car transformed into a bike can be linked either as modes of transportation or as sharing mechanical components (wheels, brakes, steering systems, and frames). In any case, the transformation is still bizarre, whether bike and car are related or not, because it is a discontinuity in the dream and because "in the normal (waking) state, such associations do not intrude into our consciousness and are unable to override our externally supplied sensory information" (110).

There are two issues here which I will take up separately. First, the notion that coherence in dream transformations can be assessed from dream reports by judges seems questionable on two counts: (1) the validity of dream reports as a source of such evidence, and (2) the assumption that coherence in dreams can be determined by an external agency from any sort of evidence. I suppose that the car / bike transformation could be considered coherent on almost any imaginable basis. But coherence is not always a quality that reveals itself in such obvious resemblances between objects. In fact, what may be a discontinuity in the realm of logic is not necessarily a discontinuity in the realm of association. For example, a bad syllogism often makes a good metaphor: i.e., hummingbirds flit from flower to flower; Sarah flits from project to project; therefore Sarah is a hummingbird. There is no way for a waking judge (or a sleeping one) to detect subjective coherence of this kind because it is rarely a visual part of the dream content, or even of the dreamer's awareness, and therefore it will not appear in the dream report.

For instance, suppose the dreamer is riding a bicycle and Uncle Albert suddenly appears out of nowhere (or vice versa). On the Hobson scale this would technically be described as a discontinuity, and therefore an example of bizarreness. But suppose it was Uncle Albert who gave the dreamer his first bicycle--a fact recorded only in the dreamer's implicit memory, that is, memory that is there but not voluntarily accessible. (Sellen & Norman, 1992, 318). Even on waking the dreamer has not remembered that crucial fact, having suffered "source amnesia" (See Schacter 1996, 118-21). Surely this would be a coherent development on the associational level in any dream even though it may violate good narrative standards in allowing Uncle Albert suddenly to change the drift of the dream. But what do narrative standards have to do with dreams anyway? Indeed, it seems axiomatic that the real coherence of dream events derives invariably from the significance of the event to the dreamer, and has only superficially to do with resemblant properties observable in the data that survive the dream. Moreover, the notion that dream coherence or incoherence can be determined by "logical" sequence is further undermined by what Schacter refers to as "mood-coherent [memory] retrieval," or the possibility that one dream image or event might follow another not because they are causally related but because they belong to the same emotional memory "bank" (1996, 211-12). In short, anything can call up anything else in memory if both share a common emotional bond that may have nothing to do with a similarity of content or structure of the specific images themselves. This leads us to the possibility that dreaming isn't an epiphenomenon of the brain processing memory--as a current theory runs; it is rather the brain idling (in neutral, if you will) on nuances of experience that have already been processed but still retain potency of one sort or another. In any case, to measure dream properties according to a waking report of the dream is to set one's comparison at two removes from the dream proper--that is, after the dream has undergone a linguistic reduction in which its invisible moorings in a unique psychology are lost.

At one point, in connection with changes in geographical location in a dream sample, the Rittenhouse study seems to be aware of this problem: "Certainly it would be difficult to see how a judge could decide that Dallas, rather than Boston or Virginia, should be transformed into a subject's childhood home. An effective investigation of this issue would probably require affirmative probes into the significance of [the transformation] to the subject" [110]). But why a caveat about subjective significance in this single case and not in the other 96 cases of bizarreness where coherence measurement would surely be affected by the same principle?
This brings me to the second point: the idea that the transformation is bizarre not because the objects may be unrelated but "because in the normal (waking) state, such associations do not intrude into our consciousness and are unable to override our externally supplied sensory information" (110). Let us take a simple waking instance involving the sudden switch from a car to a bike. Suppose I need to go to town. But I suddenly remember that my wife has our only car. All is lost, except--Ah ha!--the bike in the garage. Problem solved. The question to pose first is: how does this thought transformation--from car to bike-- differ from a dream transformation? Obviously, this is the sort of association that takes place almost automatically in lower order brain programs. One might even argue that it was simple common sense, a no-brainer, not a transformation. Still, the association of car and bike has intruded into my consciousness in some form; something like a bike has crossed my mind to replace the thought of the car, though probably not in a clear visual transformation such as we find in dreams when this happens. In other words, I didn't exactly see a car in my mind's eye, and then suddenly see a bike in its place, but I did get from car to bike in a big hurry. At any rate, from the standpoint of thought processing what otherwise is different about the coupling beyond this matter of seeing vividly in one instance (the dream) or weakly in another (the garage incident)? I have gone from car to bike in the flip of a neuronal switch. That dreams are so starkly visual, as opposed to waking thought, does not seem adequate reason to say that dreams are bizarre in such instances and waking thought isn't bizarre because the image is "unable to override our externally supplied sensory information." What has happened in either case is that one thought has given birth to another thought on the basis of an association. What makes dream bizarreness seem so bizarre isn't necessarily the content of the thought/image but the peculiar way the dream visually "showcases" it.

Even so, the case is still a bit if-fy. After all, I didn't get on my bike and ride off to Cleveland to find my lost unicorn (or vice versa), as I might have in a dream. So let us imagine, further, that I go into the garage to ready the bike for my trip; and suddenly, out of nowhere, I see a sharp image of Uncle Albert, all but forgotten over the years--dear Uncle Albert who (now that I think about it) gave me my first bicycle on my tenth birthday! I stand, transfixed in recall: I can smell the scent of his tobacco, and I can see the birthday cake on the table, Uncle Albert puffing away as he wheels the bike in. Mother, dad, Norman Rockwell, The Beatles, everybody is there. And suddenly, out of nowhere, I see Miss Krider, my fourth-grade teacher who made me stand in the coat room for ten minutes because I knocked down Freddie Findley on the same bike later that year. The event springs to vivid life, morphing out my warm family scene; I smile. Where is Freddie now, I wonder? Whereupon Freddy's father pops into view carrying a garden rake! All this, and more, in perhaps twenty seconds. A common everyday mental sequence.

How would the judges score its bizarreness or coherence--bearing in mind that all I have written here was not spelled out in these ever-reductive words in my 20 second highspeed reverie in the garage? Surely such episodes in involuntary recall bring us to the very threshold of dream construction and their unique probability ratios.

Let us suppose, for a moment, that the physiology of dream sleep permitted a continuation of "normal" mentation in dreams, that there was nothing like the emergence of the cholinergic modulation phase. What would a dream be like? What would we dream about? My suspicion is that if REM continued, by one means or another, and dreams occurred, they would not differ much in kind or substance from their present nature. The basis for my suspicion is that I don't think the brain could produce an efficient, reliable, continuous, uncontaminated narrative under the most conducive conditions of dream sleep--even without the cholinergic modulation. There is, first of all, the question of why a dream would "want" to do that, what advantage there might be in creating "reliable" narratives as opposed to so-called disorderly ones. Beyond that, if the business of dreams were to tell stories with consecutive plot sequences, without serious discontinuities, we would have to re-think a lot of cognition theory because being reliable and efficient in a state of detachment from the world isn't something the brain does well--and for which benefit we can probably be thankful because at least half of the world's thinking gets done in the unreliable mode. In short, the only ground on which to compare dreams to anything would be (as I suggest elsewhere: 1998) that of imaginative thought, some kindred forms of which might be daydreaming (my bike fantasy above), trance, doing science or mathematic problems (at the initial stage), composing art of any kind, any thinking about any project whose nature, as Gilbert Ryle says, "is not preordained." Any case, in short, in which "we have to [or simply do by choice] originate or innovate" and there is "room for option, invention or preference" (1979: 56-57). Beneath such instances of thinking we may list others that are related and though less boundless still involve matters of "invention or preference"--those cases in which our attention is devoured by an imagined world: reading, watching films, viewing or hearing artistic compositions, and so on--what we might refer to as supervised dreaming. Nocturnal dreaming is simply a more radical instance of such states of "distraction" from the empirical world.

Let me emphasize that I am not comparing dreams to books or films or other products of human thought; I am comparing the dream state to the state in which imagination, like the cholinergic system, rises to the fore and "modulates" our thought, either voluntarily or involuntarily, and works toward its own ends. These ends may be to produce something (a scientific hypothesis, a work of art) or they may have no conscious purpose (day dreams, recall, trance). There are all sorts of variations, but all of them have one thing in common: they are all dominated, more or less, by the phenomenon of bizarreness, or to put it more carefully they are all dominated by an absence of constraint that encourages bizarre image formation, or radical discontinuity. Such images are not experienced as bizarre (nor are they in dreams), but as perfectly normal, for the good reason that they are exactly that. Moreover, for the most part one is not seeing the images, as in an REM dream, but using or manipulating them, as a kind of multiple-sensory thought-tool. For example, when Einstein was asked by Jacques Hadamard about the nature of his thinking while doing mathematics, he said that "the combinatory play [of images] seems to be the essential feature in productive thought--before there is any connection with logical construction in words or other kinds of signs which can be communicated to others. . . . [These] have to be sought for laboriously only in the secondary stage, when the . . . associative play is sufficiently established and can be reproduced at will." Einstein says that his images are primarily of the "visual and some, of [the] muscular type" (1954, 142; see also Sokolov, 1972: 31).

This is hardly a description of bizarreness but it is obvious that the "combinatory play of images" requires a freedom of mind unconstrained by what we would call deduction, induction, or processes of thought that follow efficient "pre-ordained" courses. These, for Einstein, come later (and "laboriously"), when the play of images has, as Polonius says, by indirection found directions out. Thought starts from scratch, which is to say that it has little--as yet--to be logical or pragmatic about; at most it has a problem to solve, a "guess" or a "hunch." The mind must wait, as it were, for assistance from intuition and its handmaiden, metaphor, which is to intuition what the syllogism is to logic. As Hadamard concludes from his study of the psychology of invention in mathematicians, "There is hardly any completely logical discovery. Some intervention of intuition issuing from the unconscious is necessary at least to initiate the logical work" (112).

An even more detailed expression of this idea is found in Stephen Pepper's classic book on World Hypotheses: A Study in Evidence. All world theories, Pepper argues, begin in analogical thought, or what he calls a root-metaphor. "A man desiring to understand the world looks about for a clue to its comprehension. He pitches upon some area of common-sense fact and tries [to see] if he cannot understand other areas in terms of this one. The original area becomes then his basic analogy or root-metaphor" (1970, 91). The process is much more involved than this brief description suggests; the main idea is that hypotheses about uncharted worlds--or systems within a world--emerge only from finding some salient characteristic of its system, some "precritical" aspect of its form, and making it a homing device, or "vehicle," with which one seeks "structural corroboration" with the other parts; and if you're lucky you arrive at a set of consistent data, or the law that makes the system a system. It is a trial and error process, and it doesn't solve all the problems of hypothesis-making, but it is, like Einstein's associative play, the way thought gets "off the ground." (See also Gibbs, 1994, 169-79, for an excellent discussion of metaphor in scientific thought.)

Obviously dreams have nothing remotely to do with hypotheses-making or the analytical goal of Einstein's thinking process. Moreover, it is probably safe to assume that Einstein didn't see unicorns or fly through the air in the discovery stage of his thought. But he must surely have conceived visual images of a universe "muscularly" pushing itself to the edges of infinity then rounding back on itself like a balloon skin. Some such image, for example, must have led him to the conception of a universe completely self-contained but without boundaries. My only claim is that dreaming involves the same mental operation as Einstein's "associative play" and my bicycle reverie; and if this is true it seems any explanation of dream bizarreness has to begin with its profound similarities to non-logical forms of thinking. You may eventually want to separate these two processes on various grounds, and to take into consideration the role brain chemistry plays in either case; I'm arguing only that you are obliged to put the processes side by side to see what they have in common before you can differentiate them; first genus, then species or differentia. The present scale of measurement used in bizarre studies seems to run directly from dream bizarreness to waking "directed" thinking--in short, what in opposition theory is called a Cut rather than a Scale--without reference to any "middle" ground of thought. My suggestion is that a better ground of comparison would be the non-directed thinking that dominates mental life in the waking world--the kind of brain state described by Lewis Thomas in The Fragile Species:

We like to think of our minds as containing trains of thought, or streams of consciousness, as though they were orderly arrangements of linear events, one notion leading in a cause-and-effect way to the next notion . . . . I can acknowledge openly that my own mind is, at most times, a muddled jumble of notions, most of them in the form of questions, never lined up in any proper order to be selected and dealt with when time allows, most of the time popping into my head unpredictably and jostling against any other ideas that happen to be floating along, each new disturbance amplifying the disorder of all the others, creating new geometric shapes of chaos imposed on chaos. (1996, 111)

One of the best ways to appreciate the fluidity of this state of mind, and to see it in action, is to attend closely to the image parade that occurs in the (pre-cholinergic) descent into sleep. Hypnogogia is the true Land of Bizarre (on good nights anyway), if only because you get better "graphics" of what is going through your head than in daytime reveries (where the environment gives you too many signals) or in dreams when you are so personally involved with the bombardment of internally generated signals that you can't appreciate how easy it is for the brain to turn an apple into a triangle or a car into a bike. It does not seem an exaggeration to say that transformations during the hypnogogic interval are far more bizarre than those occurring within an REM dream. It is the perfect synthesis of visual and imaginative thought.

To be more technical, we are involved here in the province of thinking cognitive science refers to as inner speech or inner thought, or more recently "blending" or "mapping" (See Fauconnier, 1997, Turner and Fauconnier 1995) . For convenience, I will concentrate here on the highly influential theory of the Soviet scientist-philosopher, Lev Vygotsky. Though Vygotsky refers to speech it is clear from his discussion that he is concerned with the intersection between thought and speech, or "verbal thought" (1986, 254), and he intends his theory of inner speech to stand at "the threshold of a wider and deeper subject, i.e., the problem of . . . consciousness" (256). In any case, I select Vygotsky's theory, over others, because it explains inner thought succinctly in a way that sheds light on dream mentation.
Vigotsky refers to inner speech as "speech-for-oneself" (225), which "cannot find expression in external speech" (230). Its "main characteristic trait is its peculiar syntax" (235) which is "disconnected and incomplete . . ., abbreviated and incoherent." The three main semantic peculiarities of inner speech (244) are: first, the preponderance of the sense of a word over its meaning. Sense refers to "the sum of all the psychological events aroused in our consciousness by the word [or the thought image]. It is a dynamic, fluid, complex whole, [with] several zones of unequal stability." (244-5) "A word may sometimes be replaced by another without any change in sense. Words and senses are relatively independent of each other" (246).

Second, inner thought, by its very nature, is heavily combinative, making free use of agglutination, or word fusing, in the manner of German nouns which are composites of several words. "When several words are merged into one word, the new word not only expresses a rather complex idea, but designates all the separate elements contained in that idea" (246).

Third, as a consesquence, in inner speech "a word is so saturated with sense that . . . it becomes a concentrate of sense" (247); ". . . one word stands for a number of thoughts and feelings, and sometimes substitutes for a long and profound discourse. And naturally this unique inner sense of the chosen word cannot be translated into ordinary external speech. Inner sense turns out to be incommensurable with the external meaning of the same word" (248).

In some respects, this is reminiscent of Freud's principles of image formation (condensation, overdetermination, etc.); I leave it to the reader to assess its obvious relevance to dream thought and bizarreness, because I want to move on to a more basic application. The natural form of inner speech, Vygotsky says, is Predication. "Psychologically, it consists of predicates only. It is as much a law of inner speech to omit subjects as it is a law of written speech to contain both subjects and predicates . . . . We know what we are thinking about: i.e., we always know the subject and the situation. And since [it] is already known, we may just imply it" (243).

Can we say that the "natural form" of dreams, like inner speech, is predication, that we always know the subject and the situation and therefore we abbreviate it, leaving out such things as causal connection? If so, what is it that we know about what we see in dreams that does not actually appear in them?

To approach the problem, it is necessary to recover our distinction between the compositional and the receptive-participatory aspects of dreaming. It is clear that there is a difference between the "part" of consciousness that experiences a dream and the part of consciousness that composes it, though in the deep sense they are complicitous in making the dream. This is far from an argument for double-mindedness in the sense of a mind within a mind (homunculus). But it does make a distinction between volitional and non-volitional parts of dream mentation, which is equally the case in waking thought and speech. The composition of the dream is non-volitional; it does not involve the conscious or intentional attempt to create a dream, any more than the images that occur during inner thought are provoked intentionally by the thinker; rather it takes place as a consequence of the associational patterns rising involuntarily from memory into an hospitable thought-environment and spun out on a visual field as experiences. Put simply: to have a dream is to exist fully in a world refracted from spatial and temporal continuity by memory, purified of contingency and all existential restrictions. In a dream, J. T. Fraser says, "my wish (or fear) becomes the unquestioned law that binds events together" (1990, 291). However frightening the dream world may become, it rarely surprises the dreamer, in point of its probability. "The situation," Fraser goes on, "rather resembles the second reading of a story . . .: we are simultaneously surprised and not surprised." If it appears illogical, on waking, that the dreamer has accepted the monstrosities that occur in dreams without "blinking" it is not a matter of deficient reasoning within the dream; it is a matter of what Sartre describes as the "descent to credulity" (1965, 44), or the descent into a genuine world with its own internal validity--the only validity, for the nonce, that exists.

What survives the descent into the dream sleep, then, is the creatural experience of being at home in the world in which we find ourselves. Here again, it is useful to return to Fraser's concept of Umwelt, or species specific universe: "Final reality is relative to the perceiver." There is no "immaculate perception" of reality, no superior point of view (75-6). Or, as Humberto Maturana puts the point, "We live in a domain of subject-dependent knowledge and subject-dependent reality . . . . The logic of the description is isomorphic to the logic of the operation of the describing system" (1978, 60). To live in a world, therefore, is to accept it as real. If we lived in a world with different probability ratios, we would involuntarily use them as the yardstick for measuring the reality of other possible worlds (Cf. bizarreness measuring scales). This is effectively what happens in the dream world: there is no refuting its actuality. In plain terms: you cannot have an unreal experience, any more than you can have an unreal emotion, and whether it occurs in a dream or in broad daylight is incidental to its status as an experience; therefore the reality of any "immaculate" or transcendental world outside the experience is completely beside the point. The experience itself predicates the world in which it takes place.
What this means, simply, is that in a dream you can't tell the difference between a real and an imaginary unicorn, because there is none. On the other hand, in the waking world you can make such a distinction. That is, when we are awake we can tell the difference between the world at hand and the memory of having dreamed a world that does not, in empirical fact, exist; we can confidently say, "That happened in a dream last night, whereas this moment is occurring now, in reality." We can make such a distinction for all sorts of reasons: for example, nothing that has occurred in the dream--a discovery of hidden treasure, an injury, a falling out of friends--survives it in the waking state. There is every proof that the dream was an imagined world--unless, of course, you raise the endlessly intriguing question posed by Chuang-Tzu's butterfly-philosopher.

But in the dream state no such distinction can be made between possible worlds, between this world and some other, not because we are deluded or illogical, but because the validity of the dream world does not include the possibility or proof of another world's existence. What is lost in the dream state is all standard of comparison. It is obviously possible for the dreamer to say to herself within the dream, "This is a dream" or "I must be dreaming," but this is not the same thing as saying "I must have been dreaming" when you awaken from a dream. Even when such an awareness occurs in a so-called lucid dream it is not proof that a true distinction has been made between worlds, for one can never tell whether the awareness-of-dreaming within the dream was not programed into the dream as part of the dream incentive, rather than being an independent in-sleep discovery that somehow rises above the curtain of sleep--during sleep. The belief that one is dreaming may, in short, be part of the dream, having the same status as a unicorn, regardless of its empirical truth, and to that extent it can have only dream-validity. It isn't that reason or logic have fled the coup--they are, in fact, very much intact, with certain "local" adjustments, in the dream--what has fled the coup is any confirmatory proof that can distinguish dream from waking reality. The ultimate bizarre "trip" would be to find yourself in a dream and, like Moliere's misanthrope, be utterly unable to accept it as an actual world.

In The Lives of a Cell, Lewis Thomas cites the case of the Sphex wasp which, at egg-laying time, deposits her dead prey (future food for her hatchlings) directly in front of the door of her burrow. She then inspects the inside of the burrow thoroughly, and only then pulls the prey inside. If you move the prey a short distance from the door while she is inside, Sphex will find it on emerging, bring it back to the doorway, and carry out the whole inspection routine again---and again, as Thomas says, "for as long as you have the patience or the heart for [frustrating] it" (1975, 109).

Is the wasp deluded? Is its bahavior bizarre? Certainly the answer in both cases must be no--no more deluded than the frog is in not seeing that its motionless prey is sitting on the very next lily pad within easy tongue reach; no more deluded than we may be in failing to see that a rich and endless source of food is growing in our back yards, or that our entire species is a vast experiment in frustration tolerance being conducted by a still "higher" species. It is all a matter of acceptance of the conditions in one's um-welt, as Fraser would say. Or, again, Maturana: "all that an organism can do constitutes its cognitive domain" (60). Otherwise, one must say that all creatures are deluded, all of the time, because all creatures, including physicists and mathematicians, live in a world constrained by limited cognitive domains. The wasp is being quite "logical" in dealing with its problem. Our "higher" perspective on its logic may reveal a truth of which the wasp is unaware, but to imply a state of delusion is to assume the wasp is capable of saying to itself, "Something funny is going on here," and fails to do so. The Sphex has no more capability in this regard than the dreamer: it must solve its problem its own waspish way and that is to do it over again, as often as necessary, in the very same way.
What happens, then, when we fall asleep and nature adds a strong dose of cholinergic modulation to our homespun process of inner thought? This is the province of the brain scientist, but from a purely phenomenological standpoint I offer a friendly amendment to the brainstem/cortex interaction advanced by Rittenhouse, Stickgold and Hobson: in the dream, thought becomes at once more non-directional and more directional. That is, the associational thought mode continues to unfold as it does in waking life, except that the process, well-oiled by acetylcholine, is open to an even wider range of associations, giving the dreaming brain, as Kahn and Hobson say (in another paper), "the ability to jump from one class of images/thoughts to another (1993, 158). If this is true, and I see no reason to doubt it, all that I have argued seems equally true: that such an expansion of cognitive boundaries is bound to produce an entirely different sort of associational power. On the other hand, the process, now having a habitation in a sustained imagined world (thanks to the isolation of the sleep state and the activation of the visual cortex), simultaneously becomes more directed, more like waking life, more situation-bound--in a word, bogged down in a plot whose subject is always, to one degree or another, the dreamer's personal life. The dream is the resulting tension of the two, a "cooperative competition," as Kenneth Burke would say (1964, 167), between linear and non-linear thought processes: that is, between the brain's need to create meaningful causal sequences and its equally urgent need to explore correspondences or variations on its discoveries--in short, to see things as belonging to categories. It is not simply an interaction between the chaotic and the orderly but between basic forms of organizationaal imperative that are indispensable to any narrative system. The "predicate" of the dream, what the dream "knows"--its emotional or anxietal tension--remains relatively consistent throughout, but it pursues its "verb" under the influence of the associative mechanism which intrepidly converts new images into "class" variations through the principle of resemblance. Thus, in trying to get someplace in a dream you are likely to find yourself driving a car, then a bicycle or a sled, or a rowboat, or an "impossible" amalgam of all of them, on the assumption that all forms of transportation are, so to speak, created equal (until one proves more equal than another), or that images invariably bring their class relatives along with them, if only as ghosts. Moreover, there is no guarantee of your arriving at your destination. The dream frankly doesn't care whether you arrive or not because it is less a directed plot (with beginning, middle, and end) than an exercise, as Pepper might say, in the structural corroboration of memory "parts." This all defies waking standards of coherence until you fold into the picture the idea that a dream is a world with its own unimpeachable truth, its events and facts as valid as the maxims of logic or the hypotheses of science whose sole value, as Pepper puts it, "is as a means of facilitating human thought" (72).

This is, as I say, another way to look at the matter, and I'm sure it is far too simply expressed. Its advantage is that it makes the cortical response to chaotic brainstem signals more than simply a matter of "trying to make sense" of disorder. Making sense, on either an elementary or a complex level, involves ramification, or further elaboration, as well as advancement along logical lines: a diagonal compromise, if you will, between horizontal advancement and vertical depth. Without the two principles working in tandem thought would be hopelessly impoverished and waspish. In a way, it is the same grammatical necessity that requires adjectives which "slow down" the speed of nouns caught in the gravity of the verb. My argument, in any event, is not that everything in a dream is coherent, any more than everything in inner thought is coherent. How could such thought be coherent when it is, in its very nature, a search for coherence, for a way through a problem or situation that has not yet defined itself in narrative lineaments? My point is really twofold: first, you cannot establish coherence by what amounts to a backward projection from from dream report to dream and, second, standards of coherence derived from one world-order do not apply in another. The coherence or noncoherence of dreams is simply unavailable to study except on the most rudimentary and conjectural level. Even if you got total agreement from a hundred judges, you would end up with an indeterminate truth factor because the judges are playing a different game in a different ball park. Barring a miracle in nanotechnology, this is probably a limitation we will have to live with, and it is the major nemesis of all dream interpretation.


Antrobus, J. S. (1990). The neurocognition of sleep mentation: Rapid eye movements, visual imagery, and dreaming. In R. Bootzin, J. Kihlstrom, & D. Schacter (Eds.), Sleep and Cognition (pp. 3-24. Washington, DC: American Psychological Association.

Bonato, R. A., A. R. Moffitt, R. F. Hoffmann, M.A. Cuddy, & F. L. Wimmer (1991).
"Bizarreness in Dreams and Nightmares," Dreaming, 1 (March), 53-61.

Burke, K. (1964). Perspectives by incongruity: Ed. S. E. Hyman. Bloomington, Ind.:
Indiana University Press.

Crick, F & G. Mitchison (1983). The function of dream sleep. Nature, 304, 111-14.

Fauconnier, Gilles (1997). Mappings in thought and language. Cambridge, UK: Cambridge University Press.

Flanagan, O. (2000). Dreaming souls: Sleep, dreams, and the evolution of the conscious mind. New
York: Oxford.

Foulkes, D. (1985). Dreaming: A cognitive-psychological analysis. Hillsdale, N.J.: Lawrence Erlbaum.

Fraser, J. T. (1990). Of time, passion, and knowledge: Reflections on the strategy of existence.
Princeton, N. J.: Princeton.

Garfield, P. (1999). The Universal Dreams. Dream Time, 16 (Winter/Spring), 1, 25-6.

Gibbs, R. W. Jr. (1994). The poetics of mind: Figurative thought, language, and understanding. New
York: Cambridge University Press.

Hadamard, J. (1954). An Essay on the psychology of invention in the mathematical field. N.p.,

Hall, C. S. and Van de Castle, R. L. (1966). The content analysis of dreams. New York: Appleton-

Hobson, J. A. (1988). The dreaming brain. New York: Basic Books.

Hobson, A, S. Hoffmann, R. Helfand & Kostner, D.(1987). Dream bizarreness and the activation
synthesis hypothesis. Human Neurobiology, 6, 157-164.

Hunt H. T. (1982). Forms of dreaming. Perceptual and Motor Skills , 54 (suppl.), 559- 633.

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

Hunt, H., K. Ruzycki-Hunt, D. Pariak, & K. Belicki (1993). The relationship between dream
bizarreness and imagination: artifact or essence. Dreaming, 3, 179-99.

Kahn, D. & J. A. Hobson (1993). Self-organization theory of dreaming. Dreaming, 3, 151-178.

Maturana, H. R. (1978). Biology of language: The epistemology of eeality. In G. A. Miller & E.
Lenneberg (Eds.), Psychology and biology of language and thought: Essays in honor of eric
lenneberg. New York, San Francisco, London: Academic Press.

Meier, B. (1993). Speech and thinking in dreams. C. Cavallero & D. Foulkes (Eds.) Dreaming as
Cognition (pp. 58-76). New York: Harvester Wheatsheaf.

Pepper, S. C. (1970). World hypotheses: A study in evidence. Berkeley, Los Angeles, London:
University of California Press.

Reason, J. T. (1984). Lapses of attention in everyday life. In R. Arasuraman & D. R. Davies
(Eds.), Varieties of Attention (pp. 515-549). New York: Academic Press.

Reinsel, R., J. Antrobus & M. Wollman (!992). Bizarreness in dreams and waking fantasy. In J.
Antrobus & M. Bertini (Eds.). Neuropsychology of Sleep and Dreaming (pp. 197-214).
Hillsdale, N. J: Lawrence Erlbaum.

Revonsuo, A. (1995). Consciousness, dreams and virtual realities. Philosophical Psychology, 8, 35

Revonsuo, A. & C. Salmivalli (1995). A content analysis of bizarre elements in dreams.
Dreaming, 5 (September), 169-187.

Rittenhouse, C. D., R. Stickgold, and J. A. Hobson (1994). Constraint on the transformation of
characters, objects, and settings in dream reports. Consciousness and Cognition, 3, 100-13.
(This entire "Special Issue" is devoted to pertinent essays on bizarreness in relation to the
Activation-Synthesis hypothesis.)

Ryle, G. (1979). On thinking. Oxford: Basil Blackwell.

Sartre, J. P. (1965). What Is literature. Trans. Bernard Frechtman. New York: Harper & Row.

Sartre, J. P. (1968). The psychology of imagination. Trans. Bernard Frechtman.
New York: Washington Square Press.

Schacter, D. (1996). Searching for memory: The brain, the mind, and the past.
New York: Basic Books.

Sellen, A., and D. A. Norman (1992). The psychology of slips, in Experimental
Slips and human error: Exploring the architecture of volition. Ed. B. J. Baars.
New York and London: Plenum.

Snyder, F. (1970). The phenomenology of dreaming. In L. Madow & L. H. Snow (Eds.). The
psychodynamic implications of the physiological studies on dreams. Springfield IL: Chas. S.
Thomas. (pp. 124-151).

Sokolov, A. N. (1972). Inner speech and thought. Trans by G. T. Onischenko. New York and London: Plenum.

States, B. O. (1993a). Dreaming and storytelling. Ithaca and London: Cornell.

________ (1993b). Bizarreness in dreams and other fictions, in C. S. Rupprecht (Ed.), The dream and
the text: Essays in language and literature, Albany: State University of New York Press, pp. 6-

_______ (1997). Seeing in the dark: Reflections on dreams and dreaming. New Haven and London: Yale University Press.

Strauch, I. and B. Meier (1996). In search of dreams: Results of experimental dream research.
(M. Ebon, Trans.). New york: State University of New York Press.

Sutton, J. P., C. D. Rittenhouse, E. Pace-Schott, R. Stickgold, and J. A. Hobson (1994). A new approach to dream bizarreness: Graphing continuity and discontinuity of visual attention in
narrative reports. Consciousness and Cognition, 3, 61-88.

Thomas, L. (1975). The lives of a cell: Notes of a biology watcher. New York: Bantam.

Thomas, L. (1996). The fragile species. New York: Simon & Schuster.

Turner, M. and G. Fauconnier (1995). Conceptual integrtion and formal expression.
Metaphor and Symbolic Activity 10 (3), 183-204.

Vygotsky, Lev (1996). Thought and Language. Trans. Alex Kozulin. Cambridge, MA & London:
MIT Press.

5514 Camino Contigo
Santa Barbara, CA 93111
e-mail states@humanitas.ucsb.edu


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