The reinterpretation of dreams: An evolutionary hypothesis of the function of dreaming
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
BEHAVIORAL AND BRAIN SCIENCES (2000) 23, 793–1121
Printed in the United States of America
The reinterpretation of dreams:
An evolutionary hypothesis
of the function of dreaming
Antti Revonsuo
Department of Philosophy, Centre for Cognitive Neuroscience, University of
Turku, Turku FIN-20014, Finland
revonsuo@utu.fi www.utu.fi/research/ccn/consciousness.html
Abstract: Several theories claim that dreaming is a random by-product of REM sleep physiology and that it does not serve any natural
function. Phenomenal dream content, however, is not as disorganized as such views imply. The form and content of dreams is not ran-
dom but organized and selective: during dreaming, the brain constructs a complex model of the world in which certain types of elements,
when compared to waking life, are underrepresented whereas others are over represented. Furthermore, dream content is consistently
and powerfully modulated by certain types of waking experiences. On the basis of this evidence, I put forward the hypothesis that the
biological function of dreaming is to simulate threatening events, and to rehearse threat perception and threat avoidance. To evaluate
this hypothesis, we need to consider the original evolutionary context of dreaming and the possible traces it has left in the dream con-
tent of the present human population. In the ancestral environment human life was short and full of threats. Any behavioral advantage
in dealing with highly dangerous events would have increased the probability of reproductive success. A dream-production mechanism
that tends to select threatening waking events and simulate them over and over again in various combinations would have been valuable
for the development and maintenance of threat-avoidance skills. Empirical evidence from normative dream content, children’s dreams,
recurrent dreams, nightmares, post traumatic dreams, and the dreams of hunter-gatherers indicates that our dream-production mecha-
nisms are in fact specialized in the simulation of threatening events, and thus provides support to the threat simulation hypothesis of the
function of dreaming.
Keywords: dream content; dream function; evolution of consciousness; evolutionary psychology; fear; implicit learning; nightmares; re-
hearsal; REM; sleep; threat perception
Introduction get article leads to the slightly surprising conclusion that
dreaming does have a well-defined and clearly manifested
Dreaming is a universal feature of human experience, but biological function after all. In section 1, I clarify the nature
there is no convincing explanation as to why we should ex- of the basic question: What exactly is it that we want to un-
perience dreams during sleep. Why do we have vivid, in- derstand when we inquire about the function of dreaming?
tense, and eventful experiences while we are completely The answer is that we need a clear idea of both what the
unaware of the world that physically surrounds us? Couldn’t phenomenon of dreaming is and of the sense in which we
we just as well pass the night completely nonconscious? The are using the word “function.” In section 2, we review the
function of dreaming seems to be a persistent mystery, al- currently dominant views on the function of dreaming in
though numerous suggestions have been put forward about
the possible functions it might serve. The leading neu-
rocognitive theories, however, seem to have given up the Antti Revonsuo is a Fellow of the Academy of Fin-
hope of identifying any useful function for dreaming at all. land at the University of Turku. He has published
They cannot provide us with an answer to the question widely in cognitive neuroscience, neuropsychology, and
“Why do we dream?” Instead, they seem to imply that we consciousness studies. His research aims at under-
dream for no particular reason at all: Dreaming is biologi- standing consciousness as a natural biological phenom-
cally epiphenomenal. Dream consciousness is viewed as enon and at fruitful interaction between philosophical
some sort of random noise generated by the sleeping brain and empirical research in the study of consciousness.
He is co-editor of two books on consciousness, Con-
as it fulfills various neurophysiological functions during
sciousness in philosophy and cognitive neuroscience
REM (rapid eye movement) sleep. (Erlbaum, 1994) and Beyond dissociations: Interaction
Although the prospects for discovering useful functions between dissociated implicit and explicit processing
for dreaming look rather bleak, the empirical evidence (Benjamins, 2000). Revonsuo is the European Editor of
should be reevaluated once more from a truly multidisci- Consciousness and Cognition, and currently on the
plinary point of view, including dream content analysis, the board of Directors of the Association for the Scientific
neurophysiology of dream sleep, and evolutionary psychol- Study of Consciouness.
ogy. The exploration that I undertake in the present tar-
© 2000 Cambridge University Press 0140-525X/00 $12.50 877Revonsuo: Reinterpretation of dreams
the cognitive and neuroscientific literature as well as in the threat-simulation mechanism. In the final section, the theory
more clinically oriented dream psychology. The most com- is compared with neurocognitive theories of dreaming.
mon view in cognitive neuroscience is that dreaming has no Taken together, this target article aims to show that the
function whatsoever. In clinical literature, the function of threat-simulation theory of dreaming integrates a consider-
dreaming has been linked with problem solving and psy- able body of data from multiple sources in a theoretically
chological adaptation, but the direct empirical evidence meaningful way. The theory treats the conscious phenom-
bearing on such functions remains scarce. In section 3 we enal experience of dreaming as a natural biological phe-
point out that none of the previous theories have placed nomenon best understood from the combined viewpoints
dreaming in the appropriate context for evaluating its pos- of psychology, evolutionary biology, and cognitive neuro-
sible biological functions: the human ancestral environ- science. This multidisciplinary treatment, I hope, manages
ment in which the dreaming brain was evolving for hun- to clarify the mystery of why we dream.
dreds of thousands of years. If dreaming does have any
biologically adaptive functions, they must have been effec-
tive in the evolutionary context, if anywhere. 1. What is it that we want to understand when we
In the rest of the article I argue that switching the inquire about the function of dreaming?
context in such a way puts dreaming into an entirely new
light, which suggests that the biologically adaptive func- We should first make clear what it is we are asking when we
tion of dreaming is to simulate threatening events in order inquire about the function of dreaming. We must explicate
to rehearse threat perception and the appropriate threat- what we mean by dreaming and what we mean by function.
avoidance skills and behavioral programs. I emphasize
that to claim threat simulation as the biological function 1.1. What is dreaming?
of dreaming is not to claim that every single dream of
every single individual should realize this function. It is Dreaming refers to the subjective conscious experiences
only to claim that in certain adaptively important situa- we have during sleep. We may define a dream as a subjec-
tions with certain ecologically valid cues, the system does tive experience during sleep, consisting of complex and or-
become fully activated, and this is the principal reason why ganized images that show temporal progression (Farthing
dreaming was selected for during our evolutionary history. 1992). Questions regarding the function of dreaming must
The threat simulation theory of dreaming is expressed be clearly distinguished from those regarding the function
here in the form of six propositions, each of which is em- of REM sleep. Dreaming is a subjective conscious experi-
pirically testable. The propositions can be summarized as ence, while REM sleep is a physiologically defined stage of
follows: sleep. Furthermore, as is now clear, REM sleep is neither a
1. Dream consciousness is an organized and selective necessary nor a sufficient physiological condition for dream-
simulation of the perceptual world. ing, although it seems to be the typical and perhaps optimal
2. Dream consciousness is specialized in the simulation physiological condition in which fully realized dreams are
of threatening events. brought about (Pivik 1991). As Foulkes and Cavallero (1993,
3. Nothing but exposure to real threatening events fully p. 9) emphasize, dreaming needs a level of explanation in-
activates the threat-simulation system. dependent of the neurophysiological level at which REM
4. The threat simulations produced by the fully activated sleep is defined, because “there almost certainly is REM
system are perceptually and behaviorally realistic re- sleep without dreaming and . . . there certainly is dreaming
hearsals of threatening events. without REM sleep. No account of the distinctive physiol-
5. The realistic rehearsal of these skills can lead to en- ogy of REM sleep could provide either a necessary or a suf-
hanced performance regardless of whether or not the train- ficient explanation of dreaming.” Thus, the question we will
ing episodes are explicitly remembered. be exploring is: Does it serve any useful function to have,
6. The ancestral environment in which the human brain during sleep, the sorts of conscious subjective experiences
evolved included frequent dangerous events that consti- that dreaming consists of?
tuted extreme threats to human reproductive success. They In order to make it clear that we distinguish the level of
thus presented serious selection pressures to ancestral hu- description at which dreaming proper resides from the
man populations and fully activated the threat-simulation levels of neurophysiological description, we may say that
mechanisms. dreaming is realized at the experiential or phenomenal level
The empirical evidence relevant for the evaluation of of organization in the brain (Revonsuo 1997). We want to
each proposition is then reviewed (sect. 3). In the light of find out whether the realization of this level of organization
the currently available evidence, all of the propositions are during sleep serves any natural function. The specification
judged as likely to be true, which consequently lends sup- of the functions that lower-level neurophysiological mech-
port to the threat-simulation theory of dreaming as a whole. anisms serve during REM sleep does not constitute a spec-
In section 4, the dreams of hunter-gatherer populations and ification of the functions that the realization of the phe-
animals are considered in the light of the threat simulation nomenal level serves, for the neurophysiological functions
theory. In section 5, new predictions are derived from the can be fully specified without ever mentioning the fact that
theory and the empirical testability of the theory is evalu- subjective experience happens to be realized as well.1
ated. Finally, the theory is elaborated upon and summa-
rized in section 6.
1.2. What is it to be “functional”?
After presenting the threat simulation theory, other theo-
ries that have taken an evolutionary perspective on dreaming We must be clear about what we mean by “function” or
are reviewed. Although some of them are related to the pre- “functional.” The appropriate sense of “function” in this con-
sent view, none of them includes the idea that dreaming is a text is that of a biological, adaptive function. According to
878 BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6Revonsuo: Reinterpretation of dreams
Tooby and Cosmides (1995) the biological standard is the & McCarley 1977) emphasizes the randomness of dream
only standard of functionality that is relevant to analyzing imagery. During REM sleep, PGO waves originate in the
why brain and cognition are organized in one fashion rather pons and activate the forebrain. The forebrain attempts to
than another. A cognitive system is functional in the evolu- make sense of this random activation and it synthesizes
tionary sense if and only if it promotes the organism’s inclu- dream images to fit the patterns of internally generated
sive fitness. That is, the biologically functional system must stimulation. The forebrain selects images that isomorphi-
solve problems that will increase the probability that the or- cally correspond to the patterns of eye movements and mo-
ganism possessing the system will produce offspring, or that tor commands elicited during REM sleep. The images are
the organism’s kin will produce offspring. Evolutionary biol- loaded from memory, in which day residues are particularly
ogy gives the concept of “function” a very specific content: salient. The theory delivers no answer to the question why
The function of a system solely refers to how it systematically the brain should generate any images at all during REM
caused its own propagation in ancestral environments (Tooby sleep; it is simply assumed to be an automatic process. The
& Cosmides 1995). If dreaming has an adaptive function, narrative content of dreams remains unexplained as well.
then dreaming must solve some adaptive problems whose so- More recently, Hobson (1994) has suggested that REM-
lution tends to enhance survival and promote reproduction, dreaming might have a function in memory processing, and
thus causing the persistence of the brain’s dream-production he specifically regards the rehearsal of motor programs as
mechanisms and their spread in the population. a possible function of dreaming during REM sleep. In Hob-
If dreaming does not have any adaptive function of its son’s theory, however, dreaming as an experience with vivid
own, then it is likely to be coupled to properties that do. In phenomenal content is seen as a kind of random epiphe-
that case, dreaming is a mere by-product, a nonadaptation nomenon that merely reflects some totally different events
that was not selected for (or against) during our evolution- going on at other levels of organization where such events
ary history but was dragged along because the features to may serve useful neurobiological or mnemonic functions.
which it was coupled were actively selected for. Flanagan The Activation-Synthesis theory suggests that the experi-
(1995) makes an important distinction between “natural” ential dream imagery itself, the content of consciousness, is
and “invented” functions of dreaming. A similar distinction functionally as aimless as are the noises emitted by a com-
has been made by other dream theorists between what we puter when it processes information. The phenomenal level
do with dreams once we recall them, and what the dream of organization is not regarded as biologically functional.
can do itself (Blagrove 1996; Breger 1967). Natural func- The theory presented by Crick and Mitchinson (1983;
tions are biological, adaptive functions in the sense defined 1995) is related to Hobson’s views, but contains some orig-
above, whereas invented functions are derivative psycho- inal ideas. In this theory, memory in the brain is compared
logical or cultural functions. We can put our recalled to simple models of associative nets. When such a net gets
dreams to a variety of personal or cultural uses,2 but no mat- overloaded, it easily starts to produce outputs that are com-
ter how enlightening and meaningful such uses may be, binations of actually stored associations. In order to make
they are invented by us, not by natural selection. It is doubt- storage more efficient and avoid overloading, a process of
ful that any truly natural function of dreaming could be reverse learning can be used. The net is disconnected from
based on the conscious recollection or verbal reporting of its normal inputs and outputs, and random input is given to
dream content, for the natural functions of dreaming, if any, it. The associations that this random input produces are
must have been effective in such ancestral conditions and consequently weakened, and the process is repeated many
species in which self-reflective dream recollection or re- times with different kinds of random input. According to
porting were not likely to occur – thus, the natural functions Crick and Mitchinson (1983; 1995) this is loosely analogous
of dreaming cannot have been dependent on them. to what happens in the brain during REM sleep: the brain
Now we are in the position to state our question more is disconnected from its usual inputs and outputs, and PGO
specifically. The question we are presently interested in is waves provide it with more or less random input.3 The the-
whether dreaming serves any natural functions: Does the ory explains why REM dreams are full of bizarre intrusions,
realization of the subjective phenomenal level of organiza- consisting of mixtures of features previously stored in mem-
tion (the experience of dreaming) solve any adaptive prob- ory: these are the associations arising in an overloaded net-
lems? That is, does phenomenal dreaming in any way en- work and have to be unlearned. The reverse-learning the-
hance the prospects of the reproduction of the individual ory does not even try to explain the narrative aspect of REM
(and/or its close relatives); does dreaming increase the in- dreams, and it certainly does not assign any independent
clusive fitness of the individual? function to the phenomenology of dreaming; phenomenal
dream images merely reflect the functioning of a memory-
cleaning process.
2. Current theories of dream function David Foulkes (1985) has put forward a cognitive theory
of dreaming. He proposes that dreaming originates in dif-
2.1. Theories in cognitive neuroscience
fuse, more or less random activation of semantic and episodic
In cognitive neuroscience, recent theories and views on memory during sleep: “Since it seems that the activation of
dreaming have led to the conclusion that dreaming as a con- mnemonic elements during dreaming and their selection for
scious experience does not serve any useful biological func- dream processing is random and arbitrary, it’s not likely that
tion. Only the neurophysiological events associated with the particular content of our dreams – in and of themselves
dreaming and REM sleep are assumed to be biologically – serve any adaptive functions” (Foulkes 1985, p. 200).
functional, for they may serve important functions in the Foulkes, however, distinguishes dream content from
development of the brain and in periodically restoring the dreaming as a process. Dreaming, unlike specific dream
brain’s neurochemical balance. contents, has very predictable features. It involves an in-
The Activation-Synthesis theory (Hobson 1988b; Hobson terrelated sequence of events occurring within a “world
BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6 879Revonsuo: Reinterpretation of dreams
analog” (or a model of the world) composed of integrated current waking life. The basic assumption behind this ap-
multimodal sensory imagery; the dreamer participates in proach seems to be that dreaming is functional for the in-
these events actively and personally; the contents and events dividual if the dream in some way helps the individual cope
depicted in the dream are related to the recent or distant past with his current waking concerns, solve current problems,
of the dreamer, not as a simple replay of a past experience and to promote psychological well-being. These views can
but rather as a variation of the past as something that really be traced back to Jung (1933) who argued that dreaming
could have happened to the dreamer. Foulkes suggests that, helps to maintain the individual’s psychic balance and
since the content of dreams seems to be random, what is im- Adler (1927) who believed that dreaming serves a personal
portant about the mnemonic activation is that it is in some problem-solving function.
way unique, not the precise way in which it is unique. In These types of theories of the psychological function of
Foulkes’s theory the phenomenal level of organization is not dreaming can be divided into two categories. The first holds
regarded as functional, apart from the general feature of pro- that dreaming has a problem-solving function in an intel-
ducing novel and unique mnemonic configurations. Thus, lectual or cognitive sense: The function of dreaming is to
Foulkes’s theory is not essentially different from Hobson’s as find solutions to (or to facilitate the solving of ) intellectual
to the functionality of phenomenal dream content. problems. The second holds that the function of dreaming
Solms (1997a) has recently defended the view originally is related to emotional adjustment, not to intellectual prob-
proposed by Freud: the function of dreaming is to protect lems. Any real-life event that can be considered an emo-
sleep. According to Solms, the dream process begins when tional concern for the dreamer can be seen as presenting a
external or endogenous stimuli activate “the curiosity-in- problem for psychological adjustment, and dreaming is as-
terest-expectancy circuits.” Inhibitory mechanisms prevent sumed to contribute to the emotional or behavioral adjust-
the “appetitive interest,” aroused by stimulation, from lead- ment that is called for in order to solve the emotional prob-
ing to motor activity; therefore the activity proceeds “re- lem (e.g., Breger 1967).
gressively” in the direction of hallucinations. In anxiety
dreams this mechanism of sleep protection fails. It is clear 2.2.1. Do dreams solve intellectual problems? Some stud-
that this view does not attribute any functions to the spe- ies have directly addressed the question of whether we can
cific content of dreams: Solms regards dreams simply as solve intellectual problems in our dreams or with the help
bizarre hallucinations that the weakened frontal reflective of them. Dement (1972) reports a series of experiments in
systems mistake for real perception. which 500 undergraduate students were given a copy of a
Owen Flanagan (1995) explicitly denies that dreams as problem, and before going to bed the students were to
conscious experiences have any biological function. Dream spend exactly 15 min trying to solve the problem. In the
experience, or p-dreaming (phenomenal dreaming) as Flan- morning, they wrote down any dreams they recalled from
agan calls it, is “a likely candidate for being given epiphe- the previous night and, if the problem had not been solved,
nomenalist status from an evolutionary point of view. P- spent another 15 min trying to solve it. In 1,148 attempts,
dreaming is an interesting side effect of what the brain is the problem was solved in a dream on only seven occasions.
doing, the function(s) it is performing during sleep. To put it This means that less than 1% of the dreams were success-
in slightly different terms: p-dreams, despite being experi- ful in solving the problem. Montangero (1993) reports a
ences, have no interesting biological function. I mean in the sleep laboratory experiment with six subjects. Four subjects
first instance that p-dreaming was probably not selected for, were given a formal problem, while two were trying to solve
that p-dreaming is neither functional nor dysfunctional in an intellectual problem relating to their own professional
and of itself” (Flanagan 1995, pp. 9–11). Flanagan argues careers. Although elements of the problems appeared in
that phenomenal experience during dreaming – dream con- the dreams, none of the 29 reported dreams presented the
sciousness – has no adaptive significance, because the func- solution to the problem. However, the subjects did find the
tions of REM sleep and PGO waves, in early development of solutions to the problems with relative ease during the first
the visual system and in the restoration of neurochemicals for hour after awakening in the morning. Unfortunately, it re-
the next waking period, do not in any way require mentation mains unclear whether dreaming causally contributed to
of any sort. Furthermore, dream thoughts associated with this problem-solving success at all. Cartwright (1974a) com-
such biological functions do not seem to be worth remem- pared solutions to problems arrived at either after a period
bering. “The visual, auditory, propositional, and sensory-mo- of REM sleep or an equivalent amount of waking. She con-
tor mentation that occurs is mostly noise” (p. 24). Antrobus cluded that “There is no evidence from this study that a
(1993a) seems to agree with Flanagan’s analysis. He says that period of sleep during which dreaming occurs is regularly
since in REM sleep no sensory information is processed and followed by a better performance on intellectual tasks”
no association-motor commands are executed, it makes no (p. 454). In a study by Barrett (1993) the subjects were al-
difference what the association cortex does. Dreaming has no lowed to choose the problem that they tried to solve in their
maladaptive consequences, so it has survived. dreams. The results showed that problems of a personal na-
In conclusion, theorists in cognitive neuroscience tend to ture were much more likely to find a solution through
regard the phenomenal content of dreaming as a biological dreaming than problems of an academic or intellectual na-
epiphenomenon, although at least some of the (noncon- ture. The personal problems, however, lacked definitive cri-
scious) cognitive and/or neural activity during REM sleep teria for what should count as a solution, raising the suspi-
are regarded as serving useful functions. cion that at least some of the alleged solutions may have
been attributed to the dream during retrospective re-
flection required during the reporting rather than having
2.2. Theories in dream psychology
been arrived at within the dream itself.
In psychological theories of dream function, the emphasis Blagrove (1992a) presents a thorough review and critique
is on the individual person’s psychological adaptation to his of the problem-solving paradigm of dream function. The
880 BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6Revonsuo: Reinterpretation of dreams
assumption behind this paradigm is that the function of ing is to contain or to attempt to contain this surge. If the
dreaming is to work actively and creatively toward solutions dream is successful in fulfilling this function, it does not en-
to actual current waking problems, thus going beyond what ter awareness or memory, but protects sleep. A successful
was known prior to the dream and causally contributing to pattern of dreaming first states and then works on and re-
the solution of a real-life problem. In order to evaluate the solves the problem, which leads to a positive affective out-
evidence for such claims, Blagrove distinguishes three come and no dream recall. Kramer’s (1993) studies show
types of problem-solving dreams: (1) Dreams that actually that a successful night’s dreaming is associated with having
create a new and useful solution to a current problem in more characters in the dreams and leads to increased hap-
waking life; (2) Dreams that contain problem-solving activ- piness during the next waking period. If the problem is sim-
ity that is internal to problems encountered in the dream ply restated and not solved, as in repetitive nightmares,
world, but not relevant to waking problems; (3) Dreams then the problem remains unsolved, emotions remain neg-
that reflect solutions to waking problems, but for which atively toned, and the dream easily enters awareness. Night-
there is no evidence that such solutions have not already oc- mares and bad dreams are therefore seen as unsuccessful
curred to the waking mind (i.e., the dream does not con- attempts at solving our emotional problems. This theory is
tribute to the solution, it merely reflects the solution once called the selective mood regulatory theory of dreaming
it has already been found during waking). Blagrove (1992a) (Kramer 1993).
argues that there is little evidence for problem-solving Hartmann (1995; 1996a; 1998) has recently argued that
dreams of the first type; most of the dreams apparently solv- our dreams deal with our emotions and emotional concerns
ing problems either simply reflect solutions already known by making pictorial metaphors of them. Dreaming cross-
or solve problems only relevant in the context of the dream. connects or weaves in new material, which, according to
Although a psychological change may be correlated with a Hartmann (1998), helps us adapt to future trauma, stress,
dreamed solution to a problem, there is little reason to be- and the problems of life. Thus, dreaming and psychother-
lieve that there is a causal relationship between them. It is apy fulfill somewhat similar functions. A stressful real-life
most likely that the actual solution first arises during wak- experience can be processed in both cases in a similar way,
ing, and the consequent dreaming merely reflects the solu- essentially by “making connections in a safe place” – that is,
tion, and thus becomes correlated with whatever the bene- by associating and integrating traumatic experiences with
ficial consequences of the solution were. The conclusion the rest of life in order to facilitate psychological healing.
from Blagrove’s (1992a) review is that whatever the func- Dreaming “calms” the emotional “storm” going on in the
tion of dream experience is, it does not appear to be the mind. Hartmann calls the class of psychological adaptation
finding of new and useful solutions to the problems we face views of dreaming consistent with his theory the “contem-
in our waking reality. porary theory of the functions of dreaming.”
Punamäki (1997; 1998) has recently tested the role of
2.2.2. Do dreams solve emotional problems? Probably dreams and dream recall in protecting psychological well-
the most popular theory of dream function within psychol- being in traumatic conditions. She studied the dreams of a
ogy is the hypothesis that dreaming solves our emotional group of Palestinian children living in a violent area in Gaza
problems by helping us to adjust psychologically to, and and a control group living in a peaceful area in Galilee. She
maintain our mental health in, the real-life situations that reports that traumatized children had better dream recall
trouble us emotionally and psychologically. There is an than nontraumatized ones, and the more the children were
overwhelming amount of evidence showing that dream exposed to trauma, the more negatively emotional and the
content indeed reflects the current emotional problems of less bizarre were their dreams. Frequent dream recall was
the dreamer (Hartmann 1998; Kramer 1993). The question associated with depressive symptoms, whereas infrequent
is: Does dreaming have an effect in reducing the negative dream recall was associated with somatic and anxiety symp-
affect and other negative psychological consequences in- toms. Thus, the pattern of mental health effects associated
duced by our real-life troubles and traumas? with dream recall is not straightforward, for both good and
Cartwright (1996) argues that the best way to test this hy- bad dream recall were associated with some, although dif-
pothesis empirically is to study subjects who are undergo- ferent, psychological symptoms. Furthermore, on the basis
ing a life event that creates genuine affect. She studied sub- of this study it remains unclear whether dream recall was a
jects undergoing marital separation. Seventy subjects were cause or a consequence of these symptoms, as well as
chosen from a group of 214 potential subjects. Forty of whether frequent or infrequent dream recall in any way
them were depressed as a consequence of the divorce. All serves a positive long-term mental health function in the re-
subjects slept for three nights in the laboratory, and during covery from trauma.
the third night, REM dream reports were collected. The Thus, the literature on the possible mental health func-
depressed subjects’ dreams were emotionally more nega- tions of dreaming is inconclusive as to whether dreams truly
tive than those of the nondepressed subjects. Furthermore, solve our emotional problems, protect our mental health,
the depressed subjects were more likely than the nonde- or help us to adjust psychologically and to recover from
pressed to incorporate the about-to-be-former spouse as a traumatic experiences. The empirical evidence for such
character in the dreams. In a one-year follow-up, those de- psychologically adaptive functions appears to be relatively
pressed subjects who had dreamt about their spouse were weak and correlational at best. Furthermore, it is not en-
better adjusted than those who had not. However, it re- tirely clear what the predictions of such a theory really are
mains unclear how this correlation should be interpreted; and whether the empirical evidence confirms or discon-
on the basis of this study no causal relationship between firms them. If the idea is that dreaming “protects” our men-
dream content and adjustment can be established. tal health from negative emotional impact by turning the
Kramer (1991; 1993) argues that during REM sleep stressful emotional experience into something better and
there is a surge of emotion, and that the function of dream- by integrating it with the rest of our lives, it is surprising how
BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6 881Revonsuo: Reinterpretation of dreams
often this function deserts us when we need it most. Re- the possible adaptive function of dream consciousness is
current dreams during times of stress are accompanied the prehistoric Pleistocene environment in which humans
by negative dream content, and are associated with a defi- and their ancestors lived as hunter-gatherers for hundreds
cit in psychological well-being (Zadra et al. 1997–1998). of thousands of years. If dream consciousness is biologically
When we live under constant emotional stress or have re- functional, it should have had adaptive value at least in that
cently experienced trauma, our dream consciousness typi- original environment, under the conditions in which human
cally makes us suffer from intensive nightmares that con- ancestral populations lived. Whatever the adaptive role of
stantly remind us of the trauma by reactivating powerful dream consciousness might have been in that long-gone
negative feelings and other elements from the trauma (see original context, there is no guarantee that the average
sect. 3.5). If the real function of dreaming is psychological dreaming brain today, facing a completely different envi-
healing, shouldn’t we in fact expect exactly the opposite: ronment than the one in which it evolved, should fulfill any
pleasant, comforting, manifestly healing dreams – calming, functions that we recognize as adaptive in the present en-
not amplifying, the traumatic experience? Intuitively, reliv- vironment.
ing the emotional shocks over and over again in dreams I will simply take these answers as background assump-
would not seem to be exactly what traumatized people are tions that are reasonably well established; space does not
psychologically in need of. permit a full defense of these views here (but for more on
The usual explanation for this anomaly is that the as- consciousness see Revonsuo 1995; 1997; and for an evolu-
sumed dream function has simply “failed”; nightmares are tionary perspective in cognitive neuroscience see Cosmides
treated as “failures” of dream function (Kramer 1991). But & Tooby 1995; Tooby & Cosmides 1995).
if this is so, then dream function fails a little too regularly, When put into the proper context in this manner, the
and exactly when it would be needed most. In opposition to question “Does dream consciousness have a function?” be-
these psychological adjustment theories of dreaming, I shall comes: “Did the activation of an off-line model of the world
argue that nightmarish dreams are not ones that failed to in the ancestral human brain during sleep in some way en-
perform their function, but, by contrast, are prime examples hance the probability of reproductive success of the indi-
of the kind of dreams that fully realize their biological func- vidual living in the natural, original environment?”
tion. The view that dreams solve our emotional problems My answer is in the affirmative: The off-line model of the
and increase our happiness and psychological well-being world we call “dreaming” is specialized in the simulation of
seems to include the biologically misguided assumption certain types of events that regularly and severely threat-
that normal life is free of emotional pain and trauma. Bio- ened the reproductive success of our ancestors, in order to
logically adaptive responses to danger, such as pain and fear, enhance the probability that corresponding real events be
are not there in order to increase our happiness but to in- negotiated efficiently and successfully.
crease our reproductive success. Natural selection cares
only about fitness, not our comfort (Nesse & Williams
3.2. Dream consciousness and threat simulation
1997). If dreams are biological adaptations, they may not
care about our comfort either. We are now ready to formulate an evolutionary hypothesis
on the function of dreaming. The hypothesis I am putting
forward states that dream consciousness is essentially a
3. The biological function of dreaming mechanism for simulating threat perception and rehearsing
threat-avoidance responses and behaviors. The threat sim-
The discussion above shows that there is no convincing ev- ulation hypothesis of dreaming is presented below in the
idence that dreaming would causally contribute to the solv- form of several independent empirically testable proposi-
ing of either intellectual or emotional problems. We must tions. If each of these propositions is judged as probably
look elsewhere to discover the biological function of dream- true in the light of empirical evidence, then the threat-
ing. simulation hypothesis will receive considerable empirical
support; but if most of them are not supported by empiri-
cal evidence, then the hypothesis will be falsified. I try to
3.1. Background assumptions
show that there are good reasons to believe that each of
The construction of the appropriate context for discovering these propositions is actually true.
the biological function of dream consciousness requires
clarification of the following two questions: (Q1) What is
3.3. Proposition 1
the level of organization to which we attribute a function
when we attribute it to consciousness? (Q2) What was the Dream experience is not random or disorganized; instead, it
biological context in which dream consciousness evolved? constitutes an organized and selective simulation of the per-
Here are brief answers to these questions: ceptual world.
(A1) Consciousness can be reconceptualized as the phe- The demonstration that something is a biological adapta-
nomenal level of organization in the brain (Revonsuo tion is always “a probability assessment concerning how
1999a). A function attributed to consciousness concerns the likely a situation is to have arisen by chance” (Tooby & Cos-
causal powers and behavioral effects of events realized at mides 1992, p. 62). The content of dreams shows far too
the phenomenal level of organization. The phenomenal much organization to be produced by chance. Empirical
level forms the brain’s real-time model of the surrounding dream research has shown that dream consciousness is or-
world, of the organism’s internal state, and of its external ganized along the same lines as our waking consciousness.
position in the environment. Dreaming as a subjective ex- All sensory modalities are involved in perceptual dream ex-
perience is realized at the phenomenal level. perience, and approximately with a frequency comparable
(A2) The primary evolutionary context for considering to that of everyday waking experience (e.g., Foulkes 1985;
882 BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6Revonsuo: Reinterpretation of dreams
Strauch & Meier 1996; Zadra et al. 1998). The visual ap- According to Foulkes, dreams are credible multimodal world
pearance of dreams is for the most part identical with that analogs that are experienced as life: “The simulation of what
of the waking world (Rechtschaffen & Buchignani 1992). life is like is so nearly perfect, the real question may be, why
The dreaming brain constructs a complex, organized off- shouldn’t we believe this is real?” (Foulkes 1985, p. 37).
line model of the world in which there typically is an active Thus, all of the above shows beyond any reasonable
dream self with a body-image much like the one we expe- doubt that dreaming is an organized simulation of the per-
rience when awake, surrounded by a visuo-spatial world of ceptual world; a virtual reality (Revonsuo 1995). Even
objects, people, and animals, participating in a multitude of granted this, it could still be the case that the phenomenal
events and social interactions with other dream characters. content of dreaming is simply a random or indiscriminate
This highly predictable and organized form of dreaming sample of the phenomenal content of waking conscious-
presents a challenge to any view claiming that dream expe- ness (or the episodic memories thereof ). However, this
rience is merely an incidental by-product of neurobiologi- does not seem to be the case. There are certain experiences
cal processes operating at a different level of organization. that are very frequent contents of consciousness during
It is extremely implausible that a low-level neurochemical our waking lives but rarely or never enter our dreams.
restoration process, for example, should produce as some Hartmann (1998) describes two studies in which it was
sort of “noise” a complex and organized model of the world shown that even subjects who spend several hours daily
at a higher level of organization (cf. Foulkes 1985). If reading, writing, or calculating virtually never dream about
dreams truly were just noise, they should appear much these activities. In the first study, two judges examined 129
more noisy and disorganized than they actually are. Ran- written dream reports from several studies and found no
dom noise in the system is not likely to create organized instances of reading or writing and only one possible in-
perceptual wholes, nor is it likely to make a good story, or stance of calculating. In another study a questionnaire was
any story at all;4 it would be expected to generate disorga- mailed to 400 subjects who were frequent dreamers and
nized sensations and isolated percepts. True noise in the interested in their dreams. They reported spending an av-
brain is produced in connection with an aura of migraine erage of six hours per day engaged in reading, writing, cal-
for example. It does not generate an organized perceptual culating, or typing, but answered that they dreamed
world of objects and events; rather the contrary, it produces “never” or “almost never” about any of these activities.
for instance white or colorful phosphenes, geometric forms, They furthermore estimated on a seven-point scale how
and scintillating and negative scotomata (Sacks 1992). The frequent different activities are in dreams compared with
visual hallucinations connected with Charles Bonnet syn- waking life. Their ratings showed a remarkable dissociation
drome usually consist of static images of people, animals, between waking and dreaming life: the average rating was
buildings, and scenery (Schultz & Melzack 1991). Were our at the “far more prominent in my waking life than my
dreams closely to resemble these phenomena it would be dream life” end of the scale as to the frequency of writing,
easy to believe that dreams consist of nothing but random reading, and typing.
noise reflecting neurobiological processes at other levels of This shows that dreaming is not only an organized but
organization in the system. also a selective simulation of the world. Not every type of
It could, however, be argued that even random or disor- event or activity is simulated by the dream-production
ganized processes might activate organized schemas and mechanisms, no matter how prominent they may be in our
scripts and thus produce dreamlike phenomenology. For waking lives. Given that reading, writing, typing, and cal-
example, in Penfield’s (1975) studies the direct electrical culating are excluded from, or at least grossly underrepre-
stimulation of temporal cortex produced vivid and realistic sented in, dream experience, what kind of phenomenal
perceptual “flashbacks.” Still, these experiences were in content is overrepresented in it? Which events is dream ex-
many ways dissimilar to dreams: they were short (a few sec- perience really specialized in simulating? This question
onds) and undramatic excerpts of the patients’ previous ex- leads us to Proposition 2.
periences, like randomly chosen artificially activated mem-
ory traces: “The mechanism is capable of bringing back a
strip of past experience in complete detail without any of 3.4. Proposition 2
the fanciful elaborations that occur in a man’s dreaming” Dream experience is specialized in the simulation of threaten-
(Penfield 1975, p. 34). Thus, the activation of such traces ing events.
would not produce dreams as we know them. Conse-
quently, there is no evidence that any kind of essentially 3.4.1. Dream content shows a significant bias toward rep-
random activation could produce the phenomenology and resenting threatening elements in dreams. If dreams are
narrative structure of fully developed dreams. specialized in simulating threatening events, then we ought
Dream phenomenology, therefore, is likely to be the con- to find that dream content is biased toward including vari-
sequence of an active and organized process rather than a ous negative elements (reflecting threats) rather than posi-
passive by-product of disorganized activation. This process tive elements. Several prominent features of dream content
generates an organized world-model. Foulkes (1985) points suggest that this bias indeed exists.
out that dreams are coherently organized both momentar-
ily and sequentially. The momentary phenomenal content 3.4.1.1. Emotions in dreams. In the normative study by
of dream consciousness is comprehensible and conforms to Hall and Van de Castle (1966), 500 home dream reports
the kinds of multimodal perceptual experiences that we from female and 500 from male college students, aged 18–
have during waking perception. These momentary phe- 25, were content analyzed. Of the more than 700 emotions
nomenal contents cohere sequentially so as to constitute expressed in the dream reports, about 80% were negative
narrative stories or temporally extended episodes of expe- and only 20% positive. The figures remain similar when
rience of the same general form as our waking experience. only the dreamers’ own emotions are considered. About
BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6 883Revonsuo: Reinterpretation of dreams
half of the negative emotions experienced by the dreamers 3.4.1.4. Summary. Negative emotions, misfortunes, and ag-
were classified as “Apprehension,” the other half consisted gression are prominent in dreams. These findings indicate
of “sadness,” “anger,” and “confusion.” that normative dream content frequently contains various
In the first normative laboratory study, Snyder (1970) unpleasant and threatening elements, which supports the
collected 635 REM dream reports from students and found view that dreams are specialized in simulating threatening
that more than two-thirds of the emotions mentioned in the events.
reports were negative, fear being the most common and
anger the next most common. Strauch and Meier (1996) re- 3.4.2. Dream content is consistent with the original evo-
port a sleep laboratory study in which they not only col- lutionary environment of the human species rather than
lected REM dream reports from 44 subjects but also asked the present one
them how they had felt during the dream. The emotions de-
scribed in response to this question were analyzed. Specific 3.4.2.1. “Enemies” in our dreams. Domhoff (1996) de-
emotions were mentioned in connection with every other fines “Enemies” as those dream characters with which the
dream. Negative emotions appeared twice as often as pos- proportion of aggressive encounters of all aggressive 1
itive ones, with anger, fear, and stress being the most fre- friendly encounters is greater than 60%. This calculation on
quent types of negative emotions. In contrast to specific the Hall and Van de Castle (1966) normative sample reveals
emotions, general mood states were found to be more of- that animals and male strangers are the enemies in men’s as
ten positively than negatively toned. well as women’s dreams (Men vs. Animals 82%; Women vs.
Foulkes et al. (1988a) and Revonsuo and Salmivalli (1995) Animals 77%; Men vs. Male Strangers 72%; Women vs.
have shown that emotions in dreams are in most cases ap- Male Strangers 63%). Encounters with Female Strangers
propriate to the dreamed situations in which they are ex- are not at all so aggressive, but predominantly friendly (Men
perienced; therefore, the high proportion of negative emo- vs. Female Strangers 40%; Women vs. Female Strangers
tions is a sign of frequent unpleasant dream events that 43%) (Domhoff 1996). According to Hall and Domhoff
should be expected to produce negative emotions if they (1963), unknown males are responsible for the high pro-
were real. Emotions are evolved adaptations that increase portions of victimization and physical aggressions with male
the ability to respond appropriately in adaptively important characters.
situations. Negative emotions such as anxiety, fear, and Hall (1955) content analyzed 106 dreams of being at-
panic, can be seen as adaptive responses that increase fit- tacked and found that the attacks predominantly repre-
ness in dangerous situations threatening a loss of repro- sented situations in which the dreamer’s life or physical
ductive resources (Marks & Nesse 1994). When emotions well-being was at stake. The attacker was usually human or
are experienced or expressed in dreams, they are much a group of humans (70%) but not infrequently an animal
more likely to be negative than positive ones, and very likely (21%). When the sex of the human attacker was identified
to be appropriate to the dreamed situation. These findings it was virtually always male. The dreamer usually reacted to
are consistent with the hypothesis that dream content is bi- the attack by running, escaping, or hiding (unless she woke
ased toward simulating threatening events. up). Hall and Domhoff (1963; 1964) analyzed aggressive
and friendly interactions in more than 3,000 dream reports.
3.4.1.2. Misfortunes in dreams. “Misfortune” names a class They found that interaction was aggressive with 48% of the
of dream event in which a bad outcome happens to a char- animal characters in men’s dreams and with 29% of the an-
acter independent of anything the character has done (Hall imals in women’s dreams.
& Van de Castle 1966). Misfortunes include, for example, Van de Castle (1983) compared college students’ dreams
mishaps, dangers, and threats. The opposite is called “Good (more than 1,000 dream reports altogether) in which hu-
Fortune.” In the Hall and Van de Castle (1966) normative mans were the dominating dream characters with those in
study, there were altogether 411 cases of Misfortune in which animals predominated. He found that dreams with
1,000 dream reports, and only 58 cases of Good Fortune. animal figures typically take place in an outdoor setting,
Thus, Misfortunes in dreams are seven times more fre- have a great deal of activity that is often of a violent nature,
quent than Good Fortunes. Furthermore, about 70% of the and that the dreamer typically experiences fear. If an ani-
misfortunes happen to the dream-self, and it is accidents, mal figure initiates an interaction with the dream-self, the
losses of possession, injuries or illnesses, obstacles, and nature of the interaction is aggression 96% of the time and
threats from environment that comprise almost 90% of friendliness only 4% of the time. Van de Castle writes that
these misfortunes, whereas death and falling are rare types “almost without exception, if the animal figure initiates any
of misfortune (Domhoff 1996; Hall & Van de Castle 1966). response to the dreamer, it is some form of threat or hostil-
Misfortunes, therefore, typically reflect situations in which ity” (p. 170).
the physical well-being or the resources and goals of the Why are animals and male strangers our enemies in
dream-self are threatened. dreams? Ancestral humans lived in environments in which
many animals (e.g., large carnivores, poisonous animals,
3.4.1.3. Aggression in dreams. Aggression is the most fre- parasite-carrying animals) presented an ever-present mor-
quent type of social interaction found in dreams, the other tal threat for humans. Therefore, behavioral strategies to
classes in the Hall and Van de Castle (1966) scale being avoid contact with such animals and to escape or hide if at-
Friendliness and Sexual Interactions. About 45% of the tacked by them obviously were of high survival value. Some
dreams in the normative sample included at least one ag- deep-rooted human fears and phobias of snakes, spiders,
gressive interaction. Dreamers are involved in about 80% rats, and open spaces are indications that ancient threat
of the aggressions in their dreams, and when they are in- avoidance programs still remain with us (Marks & Nesse
volved they are more often the victim than the aggressor 1994). Dreaming simulates and rehearses these ancestral
(Domhoff 1996; Hall & Van de Castle 1966). threat-avoidance programs in order to maintain their effi-
884 BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6Revonsuo: Reinterpretation of dreams
ciency, because the costs of a single failure to respond ap-
propriately when the danger is real may be fatally high,
while the costs of repeated threat simulations during sleep
are remarkably low.
Our present-day encounters with unfamiliar males in the
waking life are not predominantly aggressive. In the ances-
tral human environment, however, intergroup aggression
and the violent competition over access to valuable re-
sources and territories is likely to have been a common oc-
currence. Since intergroup warfare and violence was and
still is almost exclusively practiced by males (Wrangham &
Peterson 1996; see also Campbell 1999), encountering
male strangers is likely to have been a potentially threaten-
ing situation in the ancestral environment, comparable to
the threats presented by dangerous animals. Indications Figure 1. Percentage of animal dreams in relationship to age
that unfamiliar males often present a mortal threat to off- (modified from Van de Castle 1970).
spring come from other primates where infanticide by ge-
netically unrelated males is common (Hrdy 1977). Fur-
thermore, human infants universally develop stranger fear dreams. Dogs, horses, and cats accounted for 28% of ani-
at about six months of age, and even in the modern world mals in children’s dreams but 38% in college students’
are much more likely to be killed or abused by genetically dreams (Van de Castle 1983). Thus, the proportion of do-
unrelated adults than by close kin (Daly & Wilson 1988). mestic animals increases and that of wild animals decreases
Thus, although an overwhelming majority of our current with age.
waking-life encounters with animals and male strangers are Due to the methods of collecting the dream reports, the
not particularly aggressive or threatening, dream content studies mentioned above may have included a somewhat bi-
still reflects the ancestral conditions in which such encoun- ased sample of dreams.5 However, also in the laboratory
ters were potentially life-threatening. Dreams are biased study of Foulkes (1982b), animals were the major charac-
toward simulating threats that were common in our ances- ters in the dream reports of children 3–5 and 5–7 years of
tral environment. age, appearing in 30–45% of the reports. Also the decrease
in the number of animal characters with increasing age was
3.4.2.2. Children’s dreams. If dreams are naturally biased confirmed. Strauch (1996) reports results from both home
toward simulating ancestral threats, then we should expect and laboratory REM dreams in Swiss children 9–11 years
that the traces of these biases are strongest early in life, of age. Both types of dreams involved more animals than
when the brain has not yet had the chance to adjust the bi- young adults’ dreams, again confirming the decrease of
ases in order to better fit the actual environment. This dream animals with increasing age. Home dreams con-
seems to be the case when it comes to the appearance of tained animals about twice as often as laboratory dreams,
animals and aggressions in children’s dreams. One of the which was explained by dream report length: home dreams
most prominent differences between child and adult dreams were longer and included more characters. Girls’ dreams
is the much larger number of animal characters in children’s contained more animals than boys’ dreams. In the REM
dreams. Hall and Domhoff (1963; 1964) analyzed about dreams, 102 animals were found. In girls’ dreams, tame an-
500 dream reports from children aged 2–12 years; Hall imals and pets prevailed (63%) over wild native or exotic an-
later increased the sample to 600 dreams and Domhoff imals (37%), whereas in boys’ dreams, wild animals were
(1996) reports the results from this larger sample. Animal much more common than tame ones (61 vs. 39%). Taken
characters make up about 25–30% of all characters in the together, on the average one out of two animals encoun-
dreams of children 2–6 years of age, and about 15% in 7– tered in the children’s dream world is an untamed wild an-
12 years of age, whereas the normative finding for adult imal. For boys around 10 years of age this is the most com-
dreams is about 5% (Domhoff 1996). mon type of dream animal.
Van de Castle (1983) also reports studies of children’s Hall and Domhoff (1963) showed that children also have
dreams. The 741 dream reports (one from each child) were a higher rate of aggression in their dreams than adults. The
written down by schoolteachers or directly reported by the greatest amount of aggression occurs in the dreams of chil-
pupils themselves. The general trend toward a decrease in dren 2–12 years of age. According to Domhoff (1996),
the frequency of animal dreams as a function of age is much of this larger amount of aggression is with animals
clearly manifested. Two-year averages in the percentage of and the child is usually the victim of an attack by the ani-
animal dreams for children 4–16 years old were 39.4% for mal. In Strauch’s (1996) data of combined REM and home
4–6 years olds, and 35.5, 33.6, 29.8, 21.9, and 13.7% for the dreams, about 30% of all the animals appearing in 10-year-
next consecutive two-year age groups. In an earlier study on old children’s dreams were in the role of aggressors, com-
a smaller sample of dreams, Van de Castle (1970) reported pared to 10% in adults’ dreams.
closely similar figures (Fig. 1). Levine (1991) studied the representation of conflicts in
Surprisingly, in their dreams children often encountered the dreams of 77 children who were about 10 years of age
animals that were seldom or never encountered in the wak- and came from three different cultures: Bedouin, Israeli,
ing world. Wild or frightening animals (e.g., snakes, bears, and Irish. Conflictual dreams accounted for about two-
monsters, lions, spiders, gorillas, tigers, wolves, insects) com- thirds of the reports and were reported about twice as of-
prised nearly 40% of all animal characters in children’s ten as nonconflictual dreams in all three cultures. The
dreams in this study, but less than 20% in college students’ Bedouin children, who were living in a traditional semi-
BEHAVIORAL AND BRAIN SCIENCES (2000) 23:6 885You can also read
