Consciousness and Cognition 10, 26–41 (2001)

doi:10.1006/ccog.2000.0471, available online at http://www.idealibrary.com on

Consciousness during Dreams

PierCarla Cicogna and Marino Bosinelli

Department of Psychology, University of Bologna, 40127 Bologna, Italy
E-mail: [email protected]

Two aspects of consciousness are first considered: consciousness as awareness (phenom-

enological meaning) and consciousness as strategic control (functional meaning). As to
awareness, three types can be distinguished: first, awareness as the phenomenal experiences
of objects and events; second, awareness as meta-awareness, i.e., the awareness of mental
life itself; third, awareness as self-awareness, i.e., the awareness of being oneself. While
phenomenal experience and self-awareness are usually present during dreaming (even if
many modifications are possible), meta-awareness is usually absent (apart from some par-
ticular experiences of self-reflectiveness) with the major exception of lucid dreaming. Con-
sciousness as strategic control may also be present in dreams. The functioning of conscious-
ness is then analyzed, following a cognitive model of dream production. In such a model,
the dream is supposed to be the product of the interaction of three components: (a) the
bottom-up activation of mnemonic elements coming from LTM systems, (b) interpretative
and elaborative top-down processes, and (c) monitoring of phenomenal experience. A feed-
back circulation is activated among the components, where the top-down interpretative
organization and the conscious monitoring of the oneiric scene elicitates other mnemonic
contents, according to the requirements of the dream plot. This dream productive activity
is submitted to unconscious and conscious processes.  2001 Academic Press
Key Words: consciousness; dream; self-awareness; dream generation processes.

Since the 1970s, the problem of consciousness, interpreted both as a subjective

experience and as a necessary condition for cognitive operations and motor perfor-
mance, has often been discussed by psychologists, neuroscientists, and philosophers.
There are, however, so many semantic variations of the words consciousness and
awareness as to make an exhaustive definition very difficult. For example, the ques-
tions concerning if the distinction between the concepts of consciousness and aware-
ness is really useful (Dennett, 1969, 1991; Jackendoff, 1987; Baars, 1988; Umiltà,
1988; Bosinelli & Cicogna, 1990; APS Forum, 1992, Natsoulas, 1992; Davies &
Humphreys, 1993, Hobson, 1999) or if consciousness can be considered a system of
control mechanisms (Johnson-Laird, 1983, 1988; Umiltà, 1988, 1997; Bosinelli,
1997). Finally, which hypothesis presents a stronger heuristic value: the hypothesis
of consciousness functioning, which assumes the existence of a centralized and super-
imposed agency (a central processor or an executive system or a supervisory system);
or the opposite, rival hypothesis, which excludes the existence of a privileged and
pyramidal control system or of a central processor (Dennett, 1969, 1991, 1995; John-
son-Laird, 1983, 1988; Norman & Shallice, 1985; Baddeley, 1986; Umiltà, 1988,

Address correspondence and reprint requests to PierCarla Cicogna, Department of Psychology, Univer-
sity of Bologna, Viale Berti Pichat, 5, 40127 Bologna, Italy.
26
1053-8100/01 $35.00
Copyright  2001 by Academic Press
All rights of reproduction in any form reserved.
 CONSCIOUSNESS DURING DREAMS 27

1997; Umiltà & Zorzi, 1995; Shallice, 1988, 1994; Crick & Koch, 1990; APS Forum,
1992; Farthing, 1992; Natsoulas, 1992)?
Many research studies and reflections about consciousness have also considered
the possibility of establishing whether there are one or more nervous structures surely
responsible for consciousness. Specific evidence in this connection is not yet avail-
able. In any case, the hypotheses relating to a distributed functioning of consciousness
rather than to a unitary center are prevalent (Umiltà & Zorzi, 1995; Umiltà, 1997).
Another hypothesis postulates that consciousness is a global function of the brain,
therefore excluding unitary or distributed locations (Dennett & Kinsbourne, 1992).
Probably, the most radical hypothesis about a precise and unitary seat of conscious-
ness is Gray’s hypothesis, which locates the contents of consciousness in the subicular
area of the septohippocampal system (subicular comparator) (Gray, 1995). It must,
however, be noted that Gray’s theory has aroused more disagreements than assents,
as it is based on his target article ‘‘The Contents of Consciousness: A Neuropsycho-
logical Conjecture’’ (Dennett, 1969, 1976, 1991, 1995; Diaz, 1995; Foss, 1995; In-
gvaldsen & Whiting, 1995; Kinsbourne, 1995; Merskey, 1995; Newman, 1995; Re-
eke, 1995; Revonsuo, 1995; Swerdlon, 1995; Umiltà & Zorzi, 1995; Velmans, 1995).
In our work we remain bound to a specific object of psychology, i.e., mental pro-
cesses (prescinding from neurophysiological correlations) and, we want to propose
a model of consciousness functioning which is based on the most important recent
findings of psychological research. We think it is possible to consider two aspects
of consciousness: first, consciousness as awareness (phenomenological meaning);
and, second, consciousness as strategic control (functional meaning) (Cicogna &
Bosinelli, 1994). As far as the phenomenological meaning is concerned, three types
of awareness can be distinguished.
The first type is awareness as the phenomenal experience of objects or of events:
awareness that/of (e.g., awareness that there is a bird on the branch/awareness of a
bird on the branch). It has many points of contact with Jackendoff ’s primary aware-
ness (1987) and Farthing’s primary consciousness, which is defined as ‘‘the direct
experience of percepts and feelings, and thoughts and memories arising in direct
response to them. It also includes spontaneously arising memories, thoughts and im-
ages, including dreams and day dreams. . . . In primary consciousness you are the
subject who does the thinking, feeling and action in regard, mainly, to external objects
and events’’ (Farthing, 1992, p. 12). According to Marcel (1988), the object or focus
of primary phenomenal experience may be considered external . In this case, the
phenomenal awareness implies indirect reports of external objects. For example, ‘‘I
can see an animal’’ and ‘‘I can hear some music,’’ where the object is phenomenally
there. On the other hand, the phenomenal experience may regard something concern-
ing the subject, and in this case the awareness implies direct reports of internal ob-
jects: ‘‘I feel thirsty’’ and ‘‘I have a pain in my leg,’’ where the object is phenome-
nally here.
The second type is awareness as meta-awareness. By meta-awareness we mean
all forms of reflective meta-cognition on primary awareness. In other words, meta-
awareness is the awareness of mental life itself, i.e., the awareness of having an
awareness of the first type (phenomenal experience that/of: ‘‘I am aware that I have
the awareness that/of’’). Meta-awareness has some analogies with reflective con-
28 CICOGNA AND BOSINELLI

sciousness (Farthing, 1992), which ‘‘consists of thoughts about one’s own conscious
experience per se . . . Your own conscious experiences, feelings and actions are the
objects of your thoughts.’’
The so-defined meta-awareness may be conceived either in a restricted or in an
extended way. In the case of oneiric processes, meta-awareness is expressed in vari-
ous ways and shades of meaning. When interpreted in its functioning as ‘‘reality
testing,’’ it probably corresponds to the most advanced level of what Purcell et al.
(1986) indicated as ‘‘self-reflectiveness,’’ which implies the possibility that the
dreamer consciously reflects on the fact that he/she is dreaming. Reality testing is
in fact the mental mechanism by which a person is able to decide whether an experi-
ence is generated by information from the world exterior to the brain or whether it
is purely a product of the mind, as in the dream. Meta-awareness often gives an
alternative of the no/yes type (‘‘I did not realize/I realized/that I was dreaming’’).
In other situations the difference between efficient and inefficient reality testing may
be notably vague (‘‘I am not sure that I was dreaming’’). Usually, the dreamer is
‘‘inside the scene,’’ taking part in the action/event/performance, or looking at this
action/event as an external observer; the subject is aware of being him- or herself,
even with some possible variations, which will be described afterward; but he or she
is unaware that this experience is derived only from mental activity. In fact, when
reality testing is inefficient, the material coming from the imaginative activity can
be interpreted as an object/event of external reality: hallucinations (for example, due
to neuropathological or psychopathological causes or drugs or sensory deprivation)
are produced in such a way. The loss of reality testing during the dream is due to
different circumstances, first among which include the lack of control sources, which
in turn depends on the closing of both the sensory input and the motor output.
The third type is awareness as self-awareness, i.e., the awareness of being oneself.
The relationship between meta-awareness and self-awareness can be synthesized as
follows. In Farthing’s model, self-awareness is a necessary condition for meta-aware-
ness. In fact, we must be aware of ourselves to be able to activate a meta-reflection
about the objects and about the modalities of our perceiving, feeling, and thinking.
It is important, however, to point out that the relationship does not work the other
way around. Self-awareness can be present also when meta-awareness is absent, as
it happens in the dream, where reality testing is almost always lost, while the phenom-
enal awareness of oneself is present. In conclusion, the statement ‘‘if meta-awareness
does exist, then self-awareness must be present’’ is true, while the statement ‘‘if self-
awareness does exist, then meta-awareness must be present’’ is false.
The phenomenology of self-awareness is highly complex. It seems to consist of
different components which are able to produce a range of puzzling operations, which
in turn can give rise to a host of different combinations and experiential modalities.
This is particularly evident in specific neuropsychopathological situations such as
confusion, delusions, body hallucinations, double self, identity troubles, and deper-
sonalization. Using the proposed consciousness/awareness model, it is possible to
distinguish many experiential modalities in the dream. The problem is quite simple
if one refers to the phenomenal experience of objects/events (first type of awareness)
because this particular world simulated in a hallucinatory way may differ slightly
from the world of the waking condition, even if the general ‘‘perceptive quality’’ of
 CONSCIOUSNESS DURING DREAMS 29

the dream very often shows some peculiar expressive features which are difficult to
define (see below) when the dream is remembered during waking.
As far the second type of awareness is concerned, it can be observed that meta-
awareness (reflective consciousness) is usually absent during sleep mentation. In an
already-classic work, Rechtschaffen (1978) tends to exclude the possibility of meta-
cognition during the oneiric experiences and describes the dream itself as endowed
with simple, nonreflective ‘‘single-mindedness and isolation.’’ On the other hand,
Purcell et al. (1986) have, instead, recorded the variations of self-reflectiveness men-
tioned before. As for meta-awareness as reality testing, its function is lost in more
than 90–95% of dreams. We give an example of a dream report in NREM sleep,
where no reference to reality testing efficiency is indicated, nor where any reference
to meta-awareness or to self-reflectiveness may be recognized:
Ex. 1: ‘‘I had arrived on a motor car. . . . I got out of the motor car, looking at the landscape. . . .’’
(Stage 2NREM report).1

Following Farthing’s model, phenomenal awareness during sleep can be considered

similar to ‘‘primary consciousness,’’ whereas ‘‘reflective consciousness’’ (meta-
awareness) is supposed to be absent or ineffective. Usually the subject can neither
think about his/her own thought producing the dream, nor preserve reality testing,
i.e., the dreamer is not aware that he/she is dreaming. When meta-awareness is fully
preserved, then the dreamer can realize that he/she is dreaming (lucid dream):
Ex. 2: ‘‘I went into my bedroom, lay on my bed and fell asleep and then, while I was dreaming
that I was sleeping, I dreamed that I was dreaming and that I was on the beach wearing a very
long and very large dress and very big shoes and I was walking very fast to meet Snow White.
Actually, there was Snow White on the beach and I was walking fast toward her. . . . While
I was dreaming that I was dreaming I realized that it was a dream, I had this strange feeling
as if I were very relaxed because actually I was asleep also in my dream and then I woke up
because in my dream I dreamed that I was awakening. . . .’’ (Stage REM)

This kind of experience of lucid dreaming is present in about 3% of the entries in

our Dream Data Bank. The partial and nonreflective functioning of consciousness
as meta-awareness could explain in part the so-called dream bizarreness during the
phenomenal experience, i.e., the representation of an experience which is factually
and/or logically improbable, odd, or impossible, but that is usually connected with
a subjective feeling, which attributes an aspect of regularity and normality to highly
irregular images or events:
Ex. 3: ‘‘I was in the country, talking with a friend of mine, also my girl-friend was there. . . .
At a certain point I could see my girl-friend inside that pool, and she was twisting and moaning,
and instantly she changed into a little dog, like a fox-terrier . . . and I was wondering why I
was unable to do the same. . . .’’

1
This example, as well as all the others which follow, belongs to a group of 800 laboratory dreams
(SO, Stage 2NREM, SWS, REM sleep) collected and scored during many years of work and then entered
into a databank (DDB, Dream Data Bank: Zito, Cicogna, & Cavallero, 1991). The general features of
the research, the instructions for the subjects, and the coding and scoring criteria often differed among
cases. For this reason it is not possible to present the reports as an internally compact and methodologi-
cally consistent statistical contribution. Thus, we use them to illustrate the different modalities of aware-
ness during dreams and the variability which exists among them.
30 CICOGNA AND BOSINELLI

In this case, the only mental reaction of the dreamer toward the dream’s bizarreness
is in regard to his being unable to produce personal identity transformations. This
phenomenon can be interpreted as absence of meta-awareness, whose correct func-
tioning should provide the dreamer with the ability to formulate judgments on the
basis of his knowledge of the world.
As far as the third type of consciousness as awareness is concerned, we believe
that self-awareness is present also in primary consciousness and consequently in the
dream, for it represents at least the ‘‘implicit ground’’ of the phenomenal experience:
The dreamer feels that he/she belongs to the scene in about 90% of the dream reports
in our Dream Data Bank records (moreover, in 3% of reports, a double representation
of the dreamer’s presence/image can be noted: on the one hand as a participant in
the action and on the other hand as a more or less external observer; see below).
As we have pointed out above, self-awareness corresponds to a conscious condition
where the subject’s attention is mainly oriented toward events which are external
to his/her reflective consciousness as well to his/her personal history. This type of
awareness is engaged only in the direct experience of perceptions, sensations, and
thoughts. However, the subject has the experience of the peripheral feedback of his/
her behavior, has the experience of various other sensations coming from the body,
and feels that he/she is the ‘‘origin’’ of perceptions and actions. This is the condition
of consciousness in which persons are during dreams. The oneiric world is hallucina-
tory and it gets information not from the real external world, but from the internal
world, which contains the dreamer’s knowledge (i.e., from long-term memory). In
the same way, the oneiric representation of oneself, in whatever form it may take
(global, partial, or bizarre), can receive information from self-knowledge only. The
latter becomes the necessary condition of both self-awareness and self-representation.
It is also necessary to take into account the fact that the oneiric hallucination is an
extremely specific process, involving the simulation of all the world elements that
we are able to perceive phenomenically (space, time, characters, actions, etc.). In
this way, oneiric and psychopathological hallucinations differ in that in the latter
usually only a single element is hallucinated and, moreover, this element is located
in a real space and in a real time.
The cognitive effort of reproducing the world in the dream is likely to require
awareness of oneself in the world. Implicitly or explicitly, the self is almost always
present in the dream. There is one clear exception—hypnagogic hallucinations: these
consist of static perceptive displays, like flashes, where there is no reference to either
participation or (at least usually) meta-awareness functioning:
Ex. 4: ‘‘There was a kind of white submarine, with two black large alphabetic letters, A and
C, engraved . . . nothing more’’ (Sleep onset)

Other exceptions may appear in mental experiences which do not take place during
sleep onset. In fact, many modifications are possible in the representation of oneself
in the oneiric scene.
In a previous research study (Bosinelli & Cicogna, 1990) a classification was at-
tempted concerning the patterns in which the experience of oneself may appear in
oneiric representations. For this purpose a nominal scale was prepared according to
which such experiences are placed in eight different categories.
 CONSCIOUSNESS DURING DREAMS 31

Category I. This concerns the absence of self-representation both as a physical

entity and as thinking subjectivity (for example, in the classic hypnagogic hallucina-
tions, cf. Ex. 4):

Ex. 5: ‘‘There was a character like Rambo, a character like that, heroic . . . going to the
Supermarket. When Rambo got out of the bus he went to the Coop and walked around inside
the Supermarket and . . . that’s all . . . I was not there.’’ (St.2-NREM)

Category II. This concerns awareness of one’s own thinking (i.e., presence of one-
self as purely a thinking agent). These are experiences where the dreamer is aware
of being a thinking subject and he or she is ‘‘a spectator’’ of his/her thought without
being able to realize that he/she is dreaming:

Ex. 6: ‘‘I was concerned about my examination. I was thinking of parts of the program I had
still to consider . . . I had the image of the open book, but it was only a thought with this
ground image, which was not inserted in any context.’’ (REM sleep)

Category III. This is static representation of oneself. (i.e., experiences which con-
cern exclusively one’s body; the total or partial self body image; more or less associ-
ated to proprioceptive, kinestesic, agreeable, or painful sensations). It is a type of
representation which is more complete than a simple noncorporealized thinking pres-
ence. On the other hand, in general, there is no ‘‘sequence’’ of which one may be
a spectator:

Ex. 7: ‘‘I was seeing my body lying on the bed, and it was completely white, better ‘‘beige.’’
I was able to see myself lying on that bed, I had to fall asleep. . . . I saw neither the room
nor the bed, only my body and the beige as a color.’’ (St. 2NREM)

Category IV. This is representation of oneself as passive observer of the oneiric

scene; there is the implicit or explicit awareness of the physical presence of oneself
observing the scene without taking part in it. The dreamer is inside the scene, but
only as an external observer. He/she is placed in a more or less defined space. Even
in the case in which that space is not defined at all, the dreamer’s experience cannot
exclude the awareness of ‘‘being there’’ and observing:

Ex. 8: ‘‘I was at a gasoline station and I was observing this scene: a child was mounting an
inflatable horse which had a motor inside. When the signal rang the horse bumped against a
pole. The child was my nephew.’’(Sleep Onset)

Category V. In most dreams, the dreamer actively participates in the event, and
the pluri-sensorial hallucination of oneself does not allow a clear distinction between
dreaming and waking self-awareness. (i.e., the precise awareness of oneself under
every aspect).

Ex. 9: ‘‘I was leaning against a chest of drawers, I was feeling unwell and there was my father
who kept on asking me what was the matter with me; I was suffering for the cat I had thrown
away for fun and had landed in a drawer of the commode from which he was unable to get
out. On the left side I saw my father and the cat’s half body protruding from the drawer, then
the whole chest of drawers. I was feeling as if I were in the real life.’’ (REM sleep)
32 CICOGNA AND BOSINELLI

Category VI. This includes situations in which the oneiric experience of oneself
is very peculiar; in some dreams the awareness is preserved only in respect to the
feeling of oneself, expressed by means of the identification with other characters or
even with objects:
Ex. 10: ‘‘There was a person, dressed like in the 7th–8th century, with a moustache and a
short beard, who was pondering on a book. He seemed engaged in a dialogue with himself,
but in reality I was pondering, it was an ideal dialogue with God, about problems regarding
theology or art criticism.’’ (REM sleep)

Ex. 11: ‘‘I was inside a gigantic photocopying machine. I knew I was inside this machine,
not as a physical human being but as an abstract entity, as a mind, so I couldn’t see myself.’’
(Slow Wave Sleep)

Category VII. This category includes double self-representation. Situations of this

type are rather frequent; they are represented in different ways. For example, when
the dreamer has both the role of the chief character and that of the observer or the
different protagonist roles (the pursuer and the pursued person). In all these situations
self-awareness is preserved, and it is referred from time to time to the character on
whom the action and the perception are focused; the physical characteristics which
produce the identity are also present in the representation of images; while the cloth-
ing, minimal physical features, abilities, and the social role can be changed:
Ex. 12: ‘‘I was in a sort of South-American country, two hundred years ago, I was mounting
a horse: with me there were two more persons riding a horse, and other people on foot. We
were pursuing a man, who was myself, we were pursuing him because he had some money,
and I was observing the whole scene’’ (REM)

Category VIII. The subjective experience is different during a lucid dream in which
a particular type of halving takes place because there is not a double self-representa-
tion, but a double awareness, where a very rapid and often unseizable passage from
the belief on the oneiric reality to the awareness of being dreaming takes place, i.e.,
the awareness of the oneiric quality of one’s own experience:
Ex. 13: ‘‘I remember a soccer match in which I was a player. I was in my bed and I was able
to see myself playing soccer, I was aware it was only my imagination, but I was unable to
control it, I was looking at myself playing but I had no corporal sensation, I realized it was
a dream.’’ (Slow Wave Sleep, stage 4)

When the dreamer can intentionally interfere in the oneiric plot, the self-awareness
condition is similar to the waking condition when a daydream or a reverie are pro-
duced, in this case, mainly utilizing analogic, visual-spatial representational codes,
and simulating reality by means of images and words.
Self-awareness during dreaming is limited to the response to the episodic situation
of the oneiric context; the mechanisms of self-regulation and judgment, which have
access to all the subject’s knowledge, seem to be nonfunctioning. In particular, as
we have already said, some specific selective judgment mechanisms do not seem to
be activated; such mechanisms allow the subject to reintegrate reality testing—the
subject realizes that the event he/she is experiencing is only a product of his/her
mind. On the other hand, it is likely that the inhibition of the judgment elements that
 CONSCIOUSNESS DURING DREAMS 33

are the constituent parts of meta-awareness contribute the hallucinatory characteris-

tics of the dream and, in conclusion, to its possibility of being realized.
So far, we have taken into consideration the phenomenological aspect of conscious-
ness in respect to the oneiric experience in general and to the oneiric experience of
oneself in particular. However, it is necessary to specify that when we refer to dream-
ing, both in terms of phenomenological experience and of dream production pro-
cesses, we refer to a type of mental activity which is present during REM and NREM
sleep. As is known, at present there is a mounting body of experimental evidence
in the psychophysiological, neuropsychological, and neuropharmacological areas,
which suggests that dreaming and REM sleep are dissociable events and supports
the hypothesis that oneiric activity is controlled by specific forebrain mechanisms
activated during different sleep stages.
The present work does not deal with this discussion in detail; we refer the reader
to the exhaustive review by Solms (1997). It is enough to point out that, on the basis
of mainly clinical evidence, Solms (1997, 1999) shows how the structures responsible
for REM sleep are not necessary for dream production. On the contrary, some specific
forebrain lesions may inhibit oneiric production also when the REM-inducing struc-
tures are not compromised. The activation of the neuro anatomic structures and of
the neurophysiological mechanisms involved in sleep may be differently modulated
according to the sleep stage (see, in the neuroimaging studies: Braun et al., 1997;
Hofle et al., 1997; Maquet et al., 1992, 1996, 1997); however, it is hypothesized that
the cognitive operations involved in dream generation be the same. In our opinion,
and in terms of mental processes, what may differ at the psychological level is the
‘‘engagement amount’’ of the cognitive system which is influenced by the modulation
of the physiological background. This may explain the occurrence of dreams of differ-
ent lengths and with different amounts of imagery, bizarreness, emotions, and so on
(Antrobus, 1991, 1999; Cavallero et al., 1992; Cicogna et al., 1991, 1998, 2000;
Foulkes, 1985, 1996). Our viewpoint also includes the hypothesis of a single dream
generator for both REM and NREM mental experiences. As far as this specific prob-
lem is concerned it is interesting to consider the contribution by Nielsen (1998) as
an attempt to overcome the persistent dichotomy between the theories of a single
dream generator (REM and NREM) and the theories of a double dream generator.
Nielsen formulates the hypothesis of the so-called ‘‘phantom REM’’: in short, many
mental experiences ascribed to NREM sleep might be due to a sort of intrusion or
expansion of REM sleep outside its usual boundaries. There is strong evidence in
support of this hypothesis (Nielsen, 1998). However, Nielsen himself points out that
it is difficult to apply this scheme to protocols collected during Slow Wave Sleep,
in particular during the first sleep cycle, preceding REM and far from sleep onset.
Also, many demonstrations of the existence of dream experiences during SWS (first
cycle of sleep) have been collected in our laboratory (Cavallero et al., 1992; Occhion-
ero et al., 1998).
As far as the functioning of consciousness (functional meaning) in dream produc-
tion is concerned, it is probably useful to start from a hypothetical cognitive model
of dream generation and to try to understand how and where different types and
aspects of unconscious and conscious processes can be introduced in the oneiric pro-
duction. Let us assume that a dream is the interaction of three components: (a)
34 CICOGNA AND BOSINELLI

bottom-up activation of mnemonic elements coming from long-term memory systems

and, to a lesser extent, of incorporated external–internal stimuli; (b) intepretative and
elaboration top-down processes; and (c) monitoring of phenomenal experience. A
feedback mechanism is activated among the components, starting with a bottom-up
process and then a top-down interpretative organization which, in its turn, may elic-
itate other mnemonic contents according to the requirements of dream’s plot planning
(Cicogna & Cavallero, 1993; Bosinelli, 1995). The simultaneous translation into im-
ages, language, and emotions, which constitutes the conscious phenomenological ex-
perience, may, in turn, influence dream development (that is, its plot planning)
through operations of control, selection, and mnestic activation.
According to the scheme presented here, the bottom-up mechanism provides the
activation of mnemonic elements from the long-term memory systems, with the ex-
ception of the procedural one, which is made inactive by the impossibility of the
subjects of supplying motor outputs. If we accept the LTM architectural model pro-
posed by Schacter and Tulving (1994), the mnemonic sources activated from the
bottom may belong to the Perceptual Representation System (PRS), to the semantic
system, and to the episodic system: The format and the consistency of the single
mnemonic units also depend on the system to which they belong. It may be that
perceptive traits or characteristics of objects or words are kept in the PRS at the
presemantic level; it may be that episodic information (bit of variable extension) is
connected to diurnal or remote traces of everyday life, and so on.
These activated memory elements are the object of a type of information pro-
cessing and interpretation by top-down mechanisms which reactivate the dreamer’s
conceptual knowledge and produce a kind of planning of the oneiric scene. The
scheme’s interpretative and organizational necessities imply the activation of sym-
bolic and conceptual knowledge, which must allow not only the interpretation but
also the integration of the piece of information that becomes spontaneously available.
As it develops, the unconscious planning simultaneously becomes the conscious syn-
tactic organization of the dream (i.e., translation into images and words).
Among the mnemonic contents, which are activated or reactivated during this cir-
cular process, the so-called cassettelike elements (Dennett, 1976) have particular im-
portance; they are preformed and well-structured fantasies that can be introduced as
such in the dream frame, with minimal modifications or no modifications at all. Also
Freud (1900) hypothesized that, during the ‘‘secondary elaboration’’ process, conve-
nient preformed fantasies might be utilized and inserted into the present dream.
The circular mechanism we have supposed to be active during dream production
can explain the various different aspects and levels assumed by the oneiric report.
The iterative feedback mechanism constructs successive drafts of the dream (more
than one of which can be simultaneously developed by this process) at an unconscious
level with only the end product, the dream itself, being accessible to awareness. The
top-down processes produce a first editorial draft of the dream experience, then the
elements of such a draft activate or inhibit other mnemonic elements, and so on. In
fact, we can imagine a kind of ‘‘stratification’’ of the mnemonic elements that pro-
duces the different subsequent editions of the tale. It is rather difficult, however, that
the traces of this kind of work be clearly present in the manifest dream report. It is
 CONSCIOUSNESS DURING DREAMS 35

evident that this model of dream production is also consistent with a general hypothe-
sis of consciousness functioning like the ‘‘multiple drafts model’’ by Dennett (1991).
This model implies that every kind of information (i.e., perception and thoughts) is
processed in the brain by parallel, multitrack processes. Consequently, every piece
of information is under continuous editorial revision. Following this view the concept
of a central agency or central supervisor inside the brain seems to be useless, for the
brain itself, as a whole, functions as a central control system (Dennett, 1991).
All the operations involved in dream generation are obviously unconscious. What
is the functional role of the conscious processes in the dream, as it is clear that the
dream itself is an experience of which the dreamer is aware? From the functional
point of view we assume that the only mental activities that can genuinely be classi-
fied as conscious in a dream refer to the monitoring of the outputs of unconscious
elaboration, the simultaneous interpretation of what is happening in the oneiric scene,
self-awareness, and the emotional experiences.
However, similarly to what takes place in waking thought, the conscious monitor-
ing of oneiric thought also allows control of the plot (representability, consistency,
judgement, etc.) and it is just in these operations that the cognitive functioning during
sleep shows the way it differs from waking. During dreaming the filter represented
by reflective awareness (meta-awareness), that is by self-regulation and judgement
mechanisms, is usually ineffective. As mentioned above, the functions ascribed to
the so-called primary awareness, which only records the peripheral feedback of the
hallucinated scene without integrating it with the whole of the dreamer’s knowledge,
including metacognitive knowledge, are operative. This can explain how phenomena
that occur very frequently in the oneiric experience, i.e., bizarreness, can be experi-
enced as plausible.
Furthermore, it is also possible that, even though working at a level of ‘‘low en-
gagement,’’ the control mechanisms are able to partially guide the development of
the scene through the activation of a retroactive process which can select and elicitate
new mnemonic information necessary to pursue the plot’s organization. As men-
tioned above, conscious monitoring occurs at the same time (i.e., in parallel) as un-
conscious planning, which has been produced following the processes outlined above
that activate and integrate the mnemonic set of information (see Fig. 1).
Therefore, the continuous feedback between bottom-up and top-down mnemonic
contents (conscious and unconscious) can also explain a peculiar characteristic of a
dream, i.e. the frequent, partial, or total newness and/or bizarreness of that mental
experience. In fact, this experience very seldom reproduces episodes or thoughts of
the dreamer’s real life, literally. The bizarreness may depend on many factors: the
results of a dream production characterized by masking mechanisms (Freud, 1900);
a supposed purely casual and chaotic brainstem origin of the mnemonic input, which
represents the first sources of dreams (randomness notion: Hobson et al., 1987, Hob-
son, 1988); the type of cognitive functioning in the dream that is the ability of the
mind–brain to produce novel objects and sequences, which may be incongruous or
bizarre from a waking viewpoint (Foulkes, 1985; Reinsel et al., 1992; Cicogna &
Cavallero, 1993; Antrobus, 1998). However, the mnemonic units can be combined
to form novel representations as a consequence of the joint activation of elements
36 CICOGNA AND BOSINELLI

FIG. 1. Model of dream generation.

or characteristics that normally do not occur simultaneously. We think that our dream
generation model may be usefully comparable with other models proposed by the
Hobson group, by Foulkes, by Seligman and Yellen, and by Antrobus.
Our model has some similarities with the one proposed by Hobson and Coll (1993,
1998), even though important differences must be indicated. For example, (a) Hobson
presumes a random origin of the activated mnemonic sources; the model presented
here is neutral about the random/nonrandom origin problem. (b) In the Hobson
model, there is no reference to a possible ‘‘circularity’’ of the process. (c) Finally,
Hobson, Stickgold, and Pace-Scott (1998), in their revised version of the activation-
synthesis model, the Activation-Input Model (AIM), maintain the persistent convic-
tion of an absolute predominance of REM sleep in the dream generation: The dream
should be considered as ‘‘an epiphenomenon of REM sleep’’ (Hobson, Stickgold, &
Pace-Scott, 1998, p. 12), ‘‘the special dream cognition . . . is strongly, if not exclu-
sively, correlated with REM physiology’’ (Hobson & Stickgold, 1994, p. 129). This
point contrasts with Solms’ mentioned analyses (1997, 1999) and with our assump-
tion that differences between REM and NREM mentation are mainly quantitative.
The model presented by Foulkes (1985) is the most compatible with our model
for its psychological-cognitive approach (HIP approach). The succession of stages
considered by Foulkes in the dream generation first includes the eliciting of a wide-
 CONSCIOUSNESS DURING DREAMS 37

spread activation in the mnemonic systems. Foulkes seems not to express himself in
a definite way on the randomized or nonrandomized ‘‘origin’’ of the memories activa-
tion, even though he seems more partial to the nonrandom origin: In fact, the excit-
atory and inhibitory processes intervening in this step of dream generation might
make one think of motivational selective components. The second mechanism consid-
ered by the Foulkes’ model consists of the organizational planning. The planner is
an active unconscious system continually engaged in deciding which units are to be
processed and how. The planner must therefore be selective and at the same time
creative. This synthesis of selectivity and creativity is the essential nucleus of the
oneiric production. In the transition from activation to planning Foulkes makes out
a difference which could mark the nonrandomness of the planning itself: ‘‘While
activation may be random, the system being activated is not’’ (Foulkes, 1985, p.
151). After the planning stage, there would be conscious interpretation of the uncon-
scious script and conclusive representation of the dream as an event. According to
the rules of serial information processing, conscious organization is the last step of
dream generation. Therefore this model states in a different way, compared with the
model proposed here, a parallel functioning of the unconscious and conscious pro-
cesses of the organizational planning of the dream. Furthermore, Foulkes (1999) in-
terprets the consciousness as a self-reflective act: this viewpoint differs from our
hypothesis about dream awareness as a nonreflective condition.
As for the model by Seligman and Yellen (1987), it draws on Hobson’s theories
and in part on the tonic-phasic model by Molinari and Foulkes (1969). In dream
generation three stages or components would be distinguished: periodical hallucina-
tions (or visual bursts), which appear during the REM stage; emotional episodes;
and, finally, synthesis or central integration. The two aspects which mostly distin-
guish their model from what is proposed in the present article are the probable casual
origin of visual bursts and the availability of the model almost exclusively for REM
sleep (likewise those of the Hobson group).
Our model is only partially compatible with the complex models proposed by An-
trobus in successive steps. All along Antrobus (1977, 1978) has pursued a neurocog-
nitive model (Antrobus, 1990, 1991; Antrobus & Bertini, 1992), in which he consis-
tently privileged the oneiro-productive role of cortical activation (expecially in REM,
yet without denying the existence of a NREM mentation) and of the variability of
perceptive thresholds. Step by step, his thought has evolved into the connectionist
model (Antrobus & Fookson, 1992; Antrobus, 1993, 1998). The first and most well
known of Antrobus’s connectionist models is called ASCIT (Across State Model of
Imagery and Thought). The PDP component of the model is clearly the most impor-
tant one, for it carries out a connection between the neurophysiologic models and the
cognitive models (see also Cavallero, 1991). In the last development of his reflection,
Antrobus clearly states that ‘‘Dreaming is a neurostate-dependent cognitive process,’’
completely different from waking mentation, by virtue of a sui generis cerebral pro-
duction ‘‘as a consequence of quasi-rhythmic changes in sub-cortical processes’’
(1998). Here his viewpoint is much different than ours, where the differences between
sleep and waking mentation are not so strongly stressed, although we are convinced
of their existence and of a possible post hoc distinction between the two types of
38 CICOGNA AND BOSINELLI

experiences. Second, the requirement of postulating a precise state dependence is

secondary in our model, which assumes a relative similarity of dreams recalled in
different stages of sleep.
Let us add some considerations which are not directly connected to the dream
generation model, yet more generally concern the functioning of cognition as far as
the dream as a final product is concerned. The conscious output of the dream is what
will be recalled by the dreamer (‘‘I dreamed . . .’’); it is the conscious phenomenal
input for long-term memory encoding and storage processes in the episodic system.
One may believe that the mnemonic processes of incidental learning are defective
during sleep; this phenomenon is particularly evident both in the so called ‘‘bad
recallers’’ and in some clinical cases, where subjects complain of the absence of
dreams (for example, after posterior and/or occipital lesions). An alternative explana-
tion of the lack of the dream recall should consist in a total absence of cognitive
work during sleep (a true interruption of cerebral functions) or in a lack of the mnestic
performance; deficits which should appear, inexplicably, only for night inputs.
As far as the relationship between consciousness functioning and oneiric activity
is concerned, one last interesting observation. When a dream is stored in the LTM,
at the moment of its retrieval it is immediately recognized as a dream. An automatic
access to a conscious higher level of consciousness (metareflexive) takes place: This
kind of functioning even allows a judgment on the subject’s representational condi-
tions (‘‘I was dreaming before; now I am recalling’’). Even in this case, as in waking
recalling episodes, a regular principle of encoding specificity might work, and it
makes the subject able to attribute such recall to an oneiric experience, as if the dream
episode had a marker, not yet clearly identified in its peculiar characteristics, which
permits the discrimination between the dream features and other episodes or thoughts.
Briefly, the subject is able to recall and to recognize both the dream episode content
and the cognitive condition in which he or she was during such episode. The recall-
control processes (central executive operations in working memory) enable us to
operate some post hoc distinctions on the qualities of waking and sleeping experi-
ences.

REFERENCES
Antrobus, J. (1977). The dream as metaphor: An information processing and learning model. Journal
of Mental Imagery, 2, 327–338.
Antrobus, J. (1978). Dreaming for cognition. In S. Arkin, J. Antrobus, & S. Ellman (Eds.), The mind
in sleep (pp. 569–581). Hillsdale, NJ: Erlbaum.
Antrobus, J. (1990). The neurocognition of sleep mentation: Rapid eye movements, visual imagery and
dreaming. In R. R. Bootzin, J. K. Kihlstrom, & D. L. Schacter (Eds.), Sleep and cognition (pp. 3–
24). Washington, DC: American Psychological Association.
Antrobus, J. (1991). Dreaming: Cognitive processes during cortical activation and high afferent thresh-
olds. Psychological Review, 98, 96–121.
Antrobus, J. (1993). Characteristics of dreams. In M. A. Carskadon (Ed.), Encyclopedia of sleep and
dreaming (pp. 98–101). New York: Macmillan.
Antrobus, J. (1998). Theories of dreaming (personal communication).
Antrobus, J., & Bertini, M. (1992). Introduction to the neuropsychology of sleep and dreaming. In J.
 CONSCIOUSNESS DURING DREAMS 39

Antrobus & M. Bertini (Eds.), The neuropsychology of sleep and dreaming (pp. 1–14). Hillsdale,
NJ: Erlbaum.
Antrobus, J., & Fookson, J. (1992). Parallel distributed processes and dream production. In S. J. Ellman &
J. Antrobus (Eds.), The mind in sleep—Psychology and psychophysiology (pp. 451–465). New
York: Wiley.
APS Forum (1992). Commentaries on ‘‘Is consciousness what psychologists actually examine? by
Thomas Natsoulas. American Journal of Psychology, 105, 591–597.
Baars, B. J. (1988). A cognitive theory of consciousness. Cambridge, UK: Cambridge Univ. Press.
Baddeley, A. D. (1986). Working memory. Oxford, UK: Claredon.
Bosinelli, M. (1995). Mind and consciousness during sleep. Behavioural Brain Research, 69, 195–201.
Bosinelli, M. (1997). Homunculus addio. Giornale Italiano di Psicologia XXIV, 1, 91–101.
Bosinelli, M., & Cicogna, P. (1990). Fenomenologia della consapevolezza di sé durante i vari stadi del
sonno. Rivista di Psicologia, 3, 61–78.
Braun, A. R., Balkin, T. J., Wesesten, N. J., et al. (1997). Regional cerebral blood flow throughout the
sleep-wake cycle. Brain, 120, 1173–1197.
Cavallero, C. (1991). Recenti sviluppi dello studio sperimentale sul sogno: l’approccio cognitivista. In
M. Bosinelli & P. Cicogna (Eds.), Sogni: figli di un cervello ozioso (pp. 354–371). Torino: Bollati
Boringhieri,
Cavallero, C., Cicogna, P., Natale, V., Occhionero, M., & Zito, A. (1992). SWS Dreaming. Sleep, 15(6),
562–566.
Cicogna, P., Cavallero, C., & Bosinelli, M. (1991). Cognitive aspects of mental activity during sleep.
The American Journal of Psychology, 104(3), 413–425.
Cicogna, P., & Cavallero, C. (1993). Coscienza e sogno. Rivista di Psicologia, 78, 27–33.
Cicogna, P., & Bosinelli, M. (1994). Models of consciousness and mental experiences during sleep.
Journal of Sleep Research, 3(Suppl. 1), 44.
Cicogna, P., Natale, V., Occhionero, M., & Bosinelli, M. (1998). Comparison between sleep onset and
morning awakening dreams. Sleep, 21(5), 462–470.
Cicogna, P., Natale, V., Occhionero, M., & Bosinelli, M. (2000). Slow Wave Sleep and REM mentation.
Sleep Research online, 3(2), 67–72.
Crick, F., & Koch, C. (1990). Towards a neurobiological theory of consciousness. Seminars in Neurosci-
ence, 2, 263–275.
Davies, M., & Humphreys, G. W. (Eds.) (1993). Consciousness. Oxford, UK: Blackwell.
Dennett, D. (1969). Content and consciousness. London: Routledge & Kegan Paul.
Dennett, D. (1976). Are dream(s) experiences? Philosophical Review, 85, 151–157.
Dennett, D. (1991). Consciousness explained. Boston: Little, Brown.
Dennett, D. (1995). Overworking the hippocampus. Behavioral and Brain Sciences, 18, 677–678.
Dennett, D., & Kinsbourne, M. (1992). Time and the observer: The where and when of consciousness
in the brain. Behavioral and Brain Science, 15, 183–247.
Diaz, J. L. (1995). Hunting for consciousness in the brain: What is (the name of ) the game. Behavioral
and Brain Sciences, 18(4), 679–80.
Farthing, W. G. (1992). The psychology of consciousness. Englewood Cliffs, NJ: Prentice-Hall.
Foss, J. (1995). On seeking the mythical fountain of consciousness. Behavioral and Brain Sciences,
18(4), 682.
Foulkes, D. (1985). Dreaming: A cognitive psychological analysis. Hillsdale, NJ: Erlbaum.
Foulkes, D. (1996). Sleep and dreams: Dream research: 1953–1993, Sleep, 19(8), 609–624.
Foulkes, D. (1999). Children’s dreaming and development of consciousness. Cambridge, MA: Harvard
Univ. Press.
Freud, S. (1900). Die Traumdeutung. Leipzig/Wien: Franz Deuticke.
40 CICOGNA AND BOSINELLI

Gray, J. A. (1995). The contents of consciousness: A neuropsychological conjecture. Behavioral and

Brain Sciences, 18(4), 659–676.
Hobson, A. (1988). The dreaming brain. New York: Basic Books.
Hobson, A. (1999). Consciousness. New York: Scientific American Library.
Hobson, A., Hoffman, S. A., Helfand, R., & Kostner, D. (1987). Dream bizarreness and the activation-
synthesis hypothesis. Human Neurobiology, 6, 157–164.
Hobson, A., & Stickgold, R. (1994). Dreaming: A neurocognitive approach. Consciousness and Cogni-
tion, 3, 1–15.
Hobson, A., Stickpold, R., & Pace-Schott, E. F. (1998). The neuropsychology of REM sleep dreaming.
Neuroreport, 16(9/3), R1–R14.
Hofle, N., Paus, N., Reutens, D., Fiset, P., Evans, A. C., & Jones, B. E. (1997). Regional cerebral blood
flow changes as a function of Delta and Spindle Activity during Slow Wave Sleep in humans. The
Journal of Neuroscience, 17(12), 4800–4808.
Ingvaldsen, R. P., & Whiting, H. T. A. (1995). Mind-your head! Behavioral and Brain Sciences, 18(4),
685–686.
Jackendoff, R. (1987). Consciousness and computational mind. Cambridge, MA: The MIT Press.
Johnson-Laird, P. (1983). Mental models. Cambridge, UK: Cambridge Univ. Press.
Johnson-Laird, P. (1988). A computational analysis of consciousness. In A. Marcel & E. Bisiach (Eds.),
Consciousness in contemporary science (pp. 357–368). Oxford: Clarendon.
Kahn, D., & Hobson, A. (1993). Self-organizing theory of dreaming. Dreaming, 3(3), 151.
Kinsbourne, M. (1995). Septohippocampal comparator: Consciousness generator or attention feedback
loop? Behavioral and Brain Sciences, 18(4), 687–688.
Maquet, P., Dive, D., Salmon, E., Sadzot, B., Franco, G., Poirrier, R., & Frank, G. (1992). Cerebral
glucose utilization during stage 2 sleep in man. Brain Research, 571, 149–153.
Maquet, P., Peters, J., Aerts, J., Delfiore, G., Degueldre, C., Luxen, A., & Franck, G. (1996). Functional
neuroanatomy of human rapid eye movement sleep and dreaming. Nature, 383, 163–166.
Maquet, P., Degueldre, C., Delfiore, G., Aerts, J. et al. (1997). Functional neuroanatomy of human slow
wave sleep. The Journal of Neuroscience, 17, 2807–2812.
Marcel, A. (1988). Phenomenal experience and functionalism. In A. Marcel & E. Bisiach (Eds.), Con-
sciousness in contemporary science (pp. 121–158). Oxford, UK: Clarendon.
Merskey, H. (1995). Comparators, functions, and experiences. Behavioral and Brain Sciences, 18(4),
689–690.
Molinari, S., & Foulkes, D. (1969). Tonic and phasic events during sleep: Psychological correlates and
implications. Perceptual & Motor Skills, 29, 343–368.
Natsoulas, T. (1992). Is consciousness what psychologists actually examine? American Journal of Psy-
chology, 105, 363–384.
Newman, J. (1995). Reticular-thalamic activation of the cortex generates conscious contents. Behavioral
and Brain Sciences, 18(4), 691–692.
Nielsen, T. A. (1998). Mentation during sleep: The NREM/REM distinction. In R. Lydic & H. A.
Baghdoyan (Eds.), Handbook of behavioral state control: Cellular and molecolar mechanisms
(pp. 101–128). Boca Raton, FL: CRC Press.
Norman, D., & Shallice, T. (1985). Attention to action. In T. Shallice (Ed.), Consciousness and self-
regulation. New York: Plenum.
Occhionero, M., Cicogna, P., Natale, V., Esposito, M. J., & Bosinelli, M. (1998). A comparision of
mental activity during slow wave and REM sleep. Journal of Sleep Research, 7(Suppl. 2), 190.
Purcell, S., Mullington, J., Moffit, A., Hoffman, R., & Pigeau, R. (1986). Dream self-reflectiveness as
a learned cognitive skill. Sleep, 9, 423–437.
Rechtschaffen, A. (1978). The single-mindedness and isolation of dreams. Sleep, I, 97–109.
Reeke G. N. (1995). Unitary consciousness requires distributed comparators and global mappings. Behav-
ioral and Brain Sciences, 18(4), 693–694.
 CONSCIOUSNESS DURING DREAMS 41

Reinsel, R., Antrobus, J., & Wollman, M. (1992). Bizarreness in dreams and working fantasy. In J.
Antrobus & M. Bertini (Eds.), The neuropsychology of sleep and dreaming (pp. 157–183). Hillsdale,
NJ: Erlbaum.
Revonsuo, A. (1995). Prospects for a cognitive neuroscience of consciousness. Behavioral and Brain
Sciences, 18(4), 694–695.
Schacter, D., & Tulving, E. (1994). Memory systems 1994. Cambridge, MA: MIT Press.
Seligman, M. E., & Yellen, A. (1987). What is a dream? Behavioral Research and Therapy, 25, 1–24.
Shallice, T. (1988). From neuropsychology to mental structure. Cambridge, UK: Cambridge Univ. Press.
Shallice, T. (1994). Multiple levels of control processes. In C. Umiltà & M. Moscovitch (Eds.), Attention
and performance (pp. 395–470). Cambridge, MA: MIT Press.
Solms, M. (1997). The neuropsychology of dreams: A clinico-anatomical study. Hillsdale, NJ: Erlbaum.
Solms, M. (1999). Dreaming and REM sleep are controlled by different brain mechanisms. Paper pre-
sented at the Annual Meeting of the American Academy for the Advancement of Science, Ana-
heim, Ca.
Swerdlon, N. R. (1995). Don’t leave the ‘‘un’’ of ‘‘consciousness.’’ Behavioral and Brain Sciences,
18(4), 699–700.
Umiltà, C. (1988). The control operations of consciousness. In A. Marcel & E. Bisiach (Eds.), Conscious-
ness in contemporary science (pp. 334–356). Oxford, UK: Clarendon.
Umiltà, C. (1997). La coscienza distribuita. Giornale Italiano di Psicologia XXIV, 1, 43–58.
Umiltà, C., & Zorzi (1995). Consciousness does not seem to be linked to a single neural mechanism.
Behavioral and Brain Sciences, 18, 701–702.
Velmans, M. (1995). The limits of neurophysiological models of consciousness. Behavioral and Brain
Sciences, 18(4), 702–703.
Zito, A., Cicogna, P., & Cavallero, C. (1991). DDB: Una banca dati per la ricerca sulla fenoemenologia
del sogno. Giornale Italiano di Psicologia, 5, 781–786.
Received April 8, 1999