Reflections on the Study of Dream Speech

Patricia A. Kilroe
California College of the Arts
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Dream speech is an understudied area of dream research worthy of attention for

This document is copyrighted by the American Psychological Association or one of its allied publishers.

its potential to shed light on the nature of the interactions between the dream-self and
dream-others, the patterns of discourse that occur among dream characters, and the
structure and content of dream speech itself. The history of the study of dream speech
is surveyed. Investigation of the structure and content of dream speech points to
interesting similarities and differences in waking, imagined, and dreamed speech.
Dream speech data support recent evidence that higher-order cognitive activity is a
feature of dreaming no less than of waking thought. The study of dream speech offers
a window on understanding dream structure and content more broadly.

Keywords: dream speech, dream characters, patterns of dream speech interaction, dream
speech structure and content, imagined speech

Dream speech may be said to occur when a dreamer reports having “said” or
“heard” speech during a dream. Speech the dreamer reports hearing may or may
not be directed at the dreamer, that is, the dreamer may witness a dream character
“speaking” to another dream character, or speech may be heard by the dreamer
that is directed at no one in particular. Both written language and the dreamer’s
thoughts belong with dream speech under the same umbrella of “evidence of
language in dream content,” but this article focuses primarily on what dreamers
perceive as speech in its auditory form: monologues, dialogues, general conversa-
tion, and disembodied voices.
The National Sleep Foundation (2015) classifies somniloquy, better known as
sleep talking, as a sleep disorder, though one that causes no physical harm. It is
generally not grouped together with dream speech, yet it
may occur during both the REM (rapid eye movement) and non-REM sleep phases. When
it happens during REM sleep . . . it’s caused by ‘motor breakthrough’ of dream speech: One’s
mouth and vocal cords, usually inactive when we’re sleeping, briefly get switched on, and
words spoken by one’s character in a dream are spoken out loud. (Wolchover, 2012)

Arkin (1981) found that sleep talking during early night non-REM (NREM)
sleep often made little or no sense, while sleep talking during REM sleep and
later-night NREM sleep showed more comprehensible, meaningful utterances that

This article was published Online First March 24, 2016.

Correspondence concerning this article should be addressed to Patricia A. Kilroe, Writing and
Literature Program, California College of the Arts, 5212 Broadway, Oakland, CA 94618. E-mail:
[email protected]

142
Dreaming © 2016 American Psychological Association
2016, Vol. 26, No. 2, 142–157 1053-0797/16/$12.00 http://dx.doi.org/10.1037/drm0000016
 Dream Speech 143

fit appropriately with the reported activity of the dream. To illustrate the dream
speech–sleep talking overlap, I offer an example of a dream I awoke from because
someone—me—was talking in the room:
I’m in a small old-fashioned farmhouse. Others are there. We’re in the dining room, around
a round table. There is conversation in the room. I wake up hearing myself talking. I’m also
aware that my lips are moving as I’m lying in bed.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

While the question of the extent to which sleep-talking represents instances of

This document is copyrighted by the American Psychological Association or one of its allied publishers.

dream speech is an interesting one, the focus in the present article is on the
characteristics of dream speech proper. (For a comprehensive work on sleep
talking, see Arkin, 1981.) This understudied area of dream research is worthy of
attention for its potential to shed light on the types of characters who engage in
dream speech, the patterns of dream speech interaction, and the structure and
content of dream speech.

The Study of Dream Speech

Although they are unlikely to be authentic dream reports, texts from ancient
history suggest early recognition that dream speech occurred, especially in the form
of messages from the divine. Mesopotamian, Egyptian, and Greek dream records
all include visitations from the gods. According to Lincoln (2003, p. 6), a dream
announcing the birth of a hero or prophet was a particular class of dreams in
ancient times:
Now Satui (father of the great Egyptian magician Senosiris) went to sleep and dreamed a
dream. Someone spoke to him saying, “Thy wife has conceived, and the child she will bear,
will be called Senosiris, and many are the miracles that will be done by him in the land of
Egypt.”

Not dissimilar in pattern is the dream of Joseph that appears in the New
Testament:
But when he had considered this, behold, an angel of the Lord appeared to him in a dream,
saying, “Joseph, son of David, do not be afraid to take Mary as your wife; for that which has
been conceived in her is of the Holy Spirit. And she will bear a Son; and you shall call His
name Jesus, for it is He who will save His people from their sins.” (New American Standard
Bible, 1972, Matthew 1:20 –21)

And God speaks directly to Jacob in the Old Testament account of Jacob’s
ladder:
And [Jacob] had a dream, and behold, a ladder was set on the earth with its top reaching to
heaven; and behold, the angels of God were ascending and descending on it. And behold, the
Lord stood above it and said, “I am the Lord, the God of your father Abraham and the God
of Isaac; the land on which you lie, I will it give to you and to your descendants. Your
descendants shall also be like the dust of the earth, and you shall spread out to the west and
to the east and to the north and to the south; and in you and in your descendants shall all the
families of the earth be blessed. And behold, I am with you, and will keep you wherever you
go, and will bring you back to this land; for I will not leave you until I have done what I have
promised you.” (New American Standard Bible, 1972, Genesis 28:10 –16)

Dream reports of this type are plentiful among ancient sources. So while
Aristotle in the 4th century BCE rejected the notion that the gods sent messages to
 144 Kilroe

dreamers, and later thinkers followed suit, the point stands that, whether prophetic,
divine, or otherwise characterized, the experience of speech in dreams was
culturally acknowledged among ancient populations. Perhaps no confirmation of
this is more convincing than the following exchange between Socrates and
Theaetetus, reported by Plato:
Socrates: There’s a question you must often have heard people ask—the question what
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

evidence we could offer if we were asked whether in the present instance, at this moment,
we are asleep and dreaming all our thoughts, or awake and talking to each other in real life.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Theaetetus: . . . There is nothing to prevent us from thinking when we are asleep that we are
having the very same discussion that we have just had. And when we dream that we are
telling the story of a dream, there is an extraordinary likeness between the two experiences.
(Plato, 1992, 157b-c)

As the founder of modern dream theory, Freud did attempt an explanation for
dream speech, yet paradoxically his explanation may have steered his followers
away from the topic. While he acknowledged that “speeches and conversations,
whether reasonable or unreasonable” were not uncommon in dreams, he insisted
that “the dream-work cannot actually create speeches.” Instead, he proposed that
the dream extracted “from the dream-thoughts fragments of speeches which have
really been made or heard,” choosing and assembling these fragments arbitrarily
(Freud, 1900/1953, p. 454). Freud did concede that speech and its meaning may vary
somewhat in the dream from what the dreamer heard during waking:
The text of the speech is either retained unaltered or expressed with some slight displace-
ment. A speech in a dream is often put together from various recollected speeches, the text
remaining the same but being given, if possible several meanings, or one different from the
original one. (Freud, 1900/1953, p. 339)

Yet in the 1909 edition of The Interpretation of Dreams, Freud admitted “a

single exception to this rule” in the form of his patient known as “Rat Man,” a
highly intelligent but obsessional neurotic for whom, according to Freud, the
“spoken words which occurred in his dreams were not derived from remarks which
he had heard or made himself” but instead “contained the undistorted text of his
obsessional thoughts.” Freud did not dwell on this exception to his claim regarding
the nature of dream speech, however, and it remained undeveloped.
Why was Freud unwilling to grant a productive role to speech in dreams? An
answer may not be easy to come by, for his writings on this point are contradictory,
as we see from the Rat Man example just cited. One possibility is the effect on
Freud of an 1861 book by German psychologist and philosopher Karl Scherner, Das
Leben des Traums (The Life of the Dream). In addition to being influenced by
Scherner’s ideas about dream symbolism, Freud may have been convinced by
Scherner’s assertion that the dreaming mind is “without the power of conceptual
speech” (quoted by Massey, 1990), and formed his theory around this claim. Joseph
Delboeuf also influenced Freud’s thinking on dream speech, as Freud noted in The
Interpretation of Dreams (Freud, 1900/1953, p. 217, n. 2): “Only one writer on the
subject seems to have recognized the source of spoken phrases occurring in dreams,
namely Delboeuf (1885, 226), who compares them to clichés.”
Another possibility may be found in Freud’s primary process model of
dreaming (for a thorough discussion of which, see Robbins, 2004). In the primary
 Dream Speech 145

process model, the dream is a nonrational, preverbal, sensory-perceptual experi-

ence, absent of reflective thought. Such a model has no place for the logic of
speech—a secondary process—and what the dreamer reports as speech is therefore
designated as regressive mimicry in order to fit the primary process model. But as
with the Rat Man example cited above, Freud seemed to allow exceptions to his
own rule about dream speeches being no more than repeated speech from waking:
“Other sorts of speeches,” he wrote, which are not felt by the dreamer “as having
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

been heard or spoken (that is, which have no acoustic or motor accompaniments in
This document is copyrighted by the American Psychological Association or one of its allied publishers.

the dream), are merely thoughts such as occur in our waking thought-activity and
are often carried over unmodified into our dreams” (Freud, 1900/1953, p. 455). But
while a role for verbal thinking during dreaming is granted here, precisely what
kinds of speeches have “no acoustic or motor accompaniments” is not addressed
beyond the assertion that “it is easy as a rule to make the distinction with certainty”
between “anything in a dream [that] has the character of direct speech” and that
which is “merely thought” (Freud, 1900/1953, p. 216). Freud’s lack of a fully
developed account of speeches in dreams notwithstanding, his facile solution to the
question of dream speech together with his dominance of the field of dream
research may be one reason that interest in dream speech was relatively lacking for
much of the 20th century.
A notable exception to this neglect, however, is the work of a contemporary of
Freud’s, German psychiatrist Emil Kraepelin. Kraepelin, best known for his work
on the classification of mental illnesses, was in fact the coiner of the term “dream
speech” (German Traumsprache). He published a monograph in 1906 entitled Über
Sprachstörungen im Traume (On Speech Disorders in Dreams). The monograph
contains 286 examples of dream speech, most of which were Kraepelin’s own.
Curious about a parallel between dreaming and psychosis (as was Freud), Kraepe-
lin’s purpose in studying dream speech was to understand the abnormal speech of
schizophrenic patients, so his corpus is one of ill-formed utterances, including errors
of word choice, grammar, and sentence meaning.
While work on dream speech among followers of Freud was minimal (but see,
e.g., Isakower (1954), for treatment of dream speech as evidence of the superego
within the psychoanalytic framework), consideration of the structure of language as
a model for understanding the formation of dreams did receive attention. Following
Freud but also strongly influenced by Ferdinand de Saussure’s groundbreaking
work on structural linguistics, Lacan (1966/2006, p. 413) famously proclaimed that
“the unconscious is structured like a language.” By this he meant that the
unconscious is not a morass of formless chaos but rather, like language, operates
through a complex network of signs and rule-governed processes.
Subsequent work by cognitive psychologists, most notably David Foulkes,
investigated dreaming as a cognitive activity, recognizing the importance of
language in that activity, for example, “Dreams are phenomena with many
organizational and structural parallels to language” (Foulkes, 1978, p. 153) and
“however ‘visual’ dreaming may seem, it may be planned and regulated by the
human speech production system” (Foulkes, 1985, p. 174). More recent studies
support the continuity theory of dreaming, finding that higher-order cognitive skills
demonstrated in waking—which necessarily include language use—are also shown
in dreaming: “[T]he same structures and processes involved in the construction of
waking experience are involved in the construction of dreaming experience”
 146 Kilroe

(Kahan, 2011, p. 113; see also Cavallero & Foulkes, 1993; Foulkes, 1999; Kahan &
LaBerge, 1994, 1996, 2011).
Strauch and Meier (1996) studied the overlap between dream speech and
thinking, carrying forward a distinction that Freud, as previously noted, only
cryptically asserted “easy as a rule” to distinguish:
There is a good deal of talk in dreams, but it is not always clear whether it is an auditory
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

phenomenon. Dreamers often find it difficult to decide whether they actually heard the
words, or whether they just knew somehow that there had been talk. (pp. 81– 82)
This document is copyrighted by the American Psychological Association or one of its allied publishers.

Analysis of their corpus of dreams showed that, while 48.5% of dream

impressions reported by dreamers were sensory impressions, thoughts, knowledge,
and memories made up 42.7%, and emotions accounted for only 8.9% of
impressions. Strauch and Meier suggested that “thought processes during dreams
. . . should be regarded as an essential part of the dream, as dreamers do not just
observe an event, but react to it” (p. 84).
Enough experimental work on dreaming as cognition had been done by the
close of the 20th century that William Domhoff could confidently repudiate Freud’s
position on dream speech:
Freud believed that all significant speeches in dreams can be traced to memories of speeches
heard or sentences read, but the analysis of hundreds of speech acts in dreams collected in
sleep laboratories shows they are usually new constructions appropriate to the unfolding
dream context, not reproductions. (Domhoff, 2000)

It can be said with sufficient confidence, then, that speech as a productive

cognitive activity does occur during dreaming, and that the speech of dreams is not
merely a verbatim repetition of speech from the waking state. But what of the
nature of dream speech? The general characteristics of dream speech have received
some, if not sufficient, attention both in and out of the sleep laboratory, as the
following highlights make clear.
Calvin Hall’s (1951) content analysis of 2,668 actions from 1,000 dreams
showed that among activities in dreams, movement such as walking or running was
the most common (34% of dreams), and talking was the next most common activity
(11% of dreams).
In a laboratory study, Shimizu and Inoue (1986) found “a positive relationship
between dreamed speech and phasic discharges of speech muscles during sleep” (p.
213).
Based on a sample of 500 REM dreams from 44 subjects over 161 nights in a
sleep laboratory, Meier (1993) scored dream activities in accordance with the
categories of the Hall and Van de Castle (1966) content analysis system. Verbal
activity was found in two thirds (63.6%) of the dreams. Meier (1993) found that
30.3% of the 2,437 activities among the 500 lab dreams were scored as verbal,
compared to 23.9% of 4,832 activities among the Hall and Van de Castle (1966)
home dreams (p. 61). The predominant subject of the sample dream conversations
was everyday life, including food, dates, and clothing. Relationships, life in general,
and gossip occupied the second level of importance, and the dream conversations
generally did not “express worries about the future or anxiety about the present
state and the course of the world.” The dream characters who engaged in
conversation with the dreamers were most often known to the dreamer, including
 Dream Speech 147

friends, colleagues, relatives, and acquaintances. Meier (1993, p. 61) concluded

from her study that verbal activity “is the principal means of social interaction in
dreams” and that dream speech “is mostly syntactically correct . . . , and it is mainly
appropriate to the overall context of the dream scenario” (p. 67).
More recently, Barrett (2009) reported on an analysis of 62 sample dream
reports containing dream speech, collected from college students. Eighty-two
percent of them (51 dreams) were found to contain linguistically well-formed, if
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

short, simple sentences (“Don’t let him get away. He wants to run.”); 5% (3
This document is copyrighted by the American Psychological Association or one of its allied publishers.

dreams) contained a neologism (“Lola was the guloff and Jeannie was his wife.”);
8% (5 dreams) had odd word choices, some of which might qualify as a mixed
metaphor (“You are a pebble that thinks it’s a comet.”); and 5% (3 dreams)
contained a phrase or fragment (“My son . . .”). An additional 19 dream reports
claimed the occurrence of dream speech but without directly quoted language (“I
definitely heard exact words out loud.”).
The most extensive work on dream speech to date, however, has been done by
Frank Heynick, as reported in numerous publications beginning in the 1980s.
Heynick’s (1993) book, Language and Its Disturbances in Dreams: The Pioneering
Work of Freud and Kraepelin Updated, presents a thorough investigation of dream
speech in the work of Freud and Kraepelin. In addition, Heynick (1993, and
elsewhere) discussed the results of two of Heynick’s own experiments conducted in
the 1980s, in which sample dreams were elicited from members of the general
population in the Netherlands. In the first of these experiments, subjects awakened
by an alarm clock at home recorded utterances recalled from any dream in progress
at the time of awakening. In the second experiment, subjects sleeping at home were
awakened randomly by telephone and reported any dream scenario and its
utterances in progress directly to the caller-experimenter. Heynick found that up to
50% of the collected data contained dialogue or monologue as part of the reported
dream. In the first experiment, 60% of the dream utterances were produced by the
dreamer, and 40% by another dream character, most often directed to the dreamer;
the numbers were reversed in the second experiment (40% uttered by the dreamer,
60% by another dream character). Heynick also found that the great majority of
these utterances were well-formed, that is, did not show deviation from ordinary
spoken language in the waking state.
These various studies are significant for confirming that language is produc-
tively generated during dreaming, as well as for identifying the relative well-
formedness of dream utterances and which dream characters produce them. Dream
speech merits additional investigation, however. What follows is a preliminary
consideration of three topics that may yield findings of relevance to an increased
understanding of verbal language as a feature of dream content.

Dream Speech Happens: Characters, Interaction Patterns,

Structure, and Content

The several perspectives on dream speech to be surveyed here include (a) the
types of dream characters who engage in dream speech, (b) patterns of dream
speech interaction, and (c) the structure and content of dream speech. Each
perspective offers a basis of comparison between the speech of dreaming and that
 148 Kilroe

of waking, and each is presented here in an exploratory way, without claim to

exhaustivity.

Types of Dream Characters Who Engage in Dream Speech

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

In his analysis of one thousand dreams, Hall (1951) found that the most
This document is copyrighted by the American Psychological Association or one of its allied publishers.

common characters beyond the dreaming self were strangers (43% of dreams),
friends and acquaintances (37% of dreams), and family members, including in-laws
(19% of dreams).
We can ask, then, other than the dream-self, who typically “speaks” in a
dream? The possibilities are multiple and overlap with many, though not all, forms
of waking interaction. My own survey of 500 dreams chosen randomly from a
variety of sources suggests that the following dream-character types are among the
most common speakers:

(A-1) People from the dreamer’s current waking life

(A-2) People from the dreamer’s past with whom there has been no recent
contact

(A-3) The deceased (usually known to the dreamer)

(A-4) Celebrities, royalty, and other media figures not personally known to the
dreamer

(A-5) Characters who bear a resemblance to someone known to the dreamer,

or who represent a blend of people and/or animals known to the dreamer

(A-6) Completely novel characters (“strangers”)

(A-7) Animals (known or unknown to the dreamer in waking life)

(A-8) Spirits or felt presences

(A-9) Unseen narrators/voiceovers

In ordinary waking experience, we of course interact with people from our

current life (A-1); less commonly, we may interact unexpectedly with people from
our past (A-2), but we do not converse with the deceased (A-3). We also do not
usually speak to celebrities (A-4), although through various media we may hear
them speak to others. Neither do we encounter people who are a blend of two or
more people (A-5), although it is certainly not unusual to meet someone who
reminds us of someone we know. Strangers (A-6), of course, are everywhere,
although interacting with them as if we know them well is generally atypical.
Animals (A-7) cannot speak human language in waking life except through the
magic of film or television (although the tradition of talking animals in literature
has a long history, e.g., Aesop’s Fables). Reports of visits from spiritual entities
(A-8) are infrequent in modern society, and an unseen narrator or “voiceover” may
be heard through a loudspeaker or other media. None of the forms of linguistic
 Dream Speech 149

interaction mentioned above are difficult to find in dream reports, but for space
considerations only the patterns identified as unusual or impossible are illustrated
here (A-3, -4, -7, and -8).
Characters representing the deceased (A-3). In the following report the
dreamer, a semiretired man in his 60s, converses with his father, who had been
deceased for a decade at the time of the dream:
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

I am involved in a project with some woman that needed to get done, like a theater project.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

My dad [deceased] arrives (he was expected) with my stepmother. He looks thin and sick. I
go to hug him. I ask him, “How are you doing?” He says, “It’s very serious.” (Meaning his
condition; he was having to attend to his own health.) Then I explain that I need to go to this
meeting for an hour, but will be back in an hour. I feel guilty because what I had to do was
not so serious, not like his condition.

Characters representing celebrities, royalty, and other media or public figures

not personally known to the dreamer (A-4). The following report features two
celebrities and includes both direct (quoted) and indirect (reported) speech. The
dreamer is a middle-aged woman, a writer and editor:
I walk into a social event with one or more others. . . . Then I’m in another room and am
helping someone famous edit a paper. . . . The celebrity, Woody Allen (though younger than
in waking life), is scrolling through his paper on a laptop set up on a long folding table; he
leans into the screen, elbows on the table, chin on the back of rounded hands. I’m trying to
give helpful suggestions but he doesn’t want my input. He ignores it, continuing to look at
the screen, but I keep trying to say things until he finally says, “Yeah, I’ll get my wife Louise
to help me with it, she’s really good at editing.” Then I enter a pool room at the end of a set
of party and editing rooms. I’m trying to get past a family with kids so I can get into the small
whirlpool before they do. I walk on hands and knees, in puddles. I do get into the pool. Soon
a man is there, Bruce Willis. He approaches me with intense interest. I tell him playfully that
he has no boundaries. He says, “And you do?” I say “Yes, I’m sensitive to other people
around, like the family there, who would not want to witness sexual activity here.”

Animal characters (A-7). In the following report, the dreamer, an 11-year-

old boy and member of the Ottawa tribe of North America, converses with an
animal on the fifth night of fasting for a vision of his future:
I dreamt that I was alongside a lake and had not had anything to eat for some time. I was
wandering in search of food for quite a time when I saw a big bird (majg). This bird came
over to where I was staying and spoke to me, telling me that I was lost and that a party was
out searching for me; and that they really intended to shoot me instead of rescuing me. Then
the bird flew out into the lake and brought me a fish to eat and told me that I would have
good luck in hunting and fishing; that I would live to a good old age; and that I would never
be wounded by a shot-gun or rifle. This bird who had blessed me was the kind that one rarely
has a chance of shooting. From that time on the majg (loon) was my guardian spirit. (cited
by Lincoln, 2003, p. 272)

Spirit, presence, or unseen speaker (A-8). Robert Van de Castle reported a

dream by a young boy who had had an accident, and it was predicted by his doctors
that he would never walk again. His grandmother reported that the boy told her the
following dream:
A very large man was at the end of his bed with a very bright circle going around and around
his head. “He told me not to be afraid, Na Na. I will be walking by next summer.”

A year later, the boy was indeed walking (Van de Castle, 2014, p. 29).
While each of these examples is unique, the number of examples is easily
multiplied, despite the unreal nature of the interactions from the perspective of
 150 Kilroe

waking reality. So while some of the types of figures we interact with in dreams
closely resemble those we might interact with in the waking world, others are
unique to the dream world and yet are not uncommon in that world.

Patterns of Dream Speech Interaction

This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

The next list features common forms of dream speech interaction, some of
which resemble waking interactions.

(B-1) A dream character (human, animal, spirit) speaks or sings lyrics to the
dreamer.

(B-2) The dreamer speaks or sings lyrics to another dream character or

characters.

(B-3) The dreamer or another dream character speaks or sings aloud to no one
in particular.

(B-4) The dreamer participates in general conversational interaction with one

or more others.

(B-5) The dreamer witnesses conversation among others without participating.

(B-6) The dreamer hears a disembodied narrator or voiceover.

(B-7) The dreamer experiences verbal thoughts.

(B-8) The dreamer writes or sees written words (in a language familiar or
unfamiliar to the dreamer).

The examples below illustrate some of the interactions less likely to overlap
with waking forms of interactions, including singing as a means of communication
(B-1), speaking to no one in particular (B-3), and voiceover (B-6). Examples of the
dreamer’s verbal thoughts (B-7) are included both here and in the next section,
which also contains an example of written language (B-8).
(B-1, -3, -7) A dream character sings lyrics; a dream character speaks/sings to
no one in particular; the dreamer experiences verbal thoughts.
The following dream report, entitled “Sing Out: The Qualities of a Queen,” is
one of my own. It illustrates not only a character singing lyrics, but also the dreamer
and another character speaking aloud to no one in particular, as well as a
dream-thought:
I’m in an auditorium, similar in design to the San Francisco opera house. I’m almost at the
top, and mostly to the left. Someone is speaking about candidates who are next in line for
queen. I ask a question out loud: “What are the qualities a queen must have? That prompts
a song from the MC (who’s a little lower down than me and more in the center, on a dark
concrete landing or walkway); he’s a singer ready to perform in these situations (that is, when
the right question is asked). I don’t think it’s John Prine, though his name comes to mind. He
has an acoustic guitar and begins singing, by way of answering my question. It’s a song I
know (“. . . for crying out loud, the love I have for you, was never in doubt . . .”).
 Dream Speech 151

Upon awakening I realized the lyrics are to the song “Babylon” by David
Gray. I thought there might be a pun on “babble on,” which convinced me to
choose dream speech for a conference talk over another topic I was considering at
the time.
The dreamer hears a disembodied narrator or voiceover (B-6). The dreamer,
a man in his 30s, reported waking up giggling at the narrative of the voiceover at the
end of this dream:
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

It started with me touring a run-down building, then deciding to repair a brick staircase,
which led across a polluted river. I wound up walking across with two others, looking for a
place to swim, but the water was so dirty and brown and full of the occasional bit of floating
debris that we just couldn’t imagine it. The river, though, had a trail running along it, and the
trail ran behind a series of upper-middle income suburban homes, which lined the river and
which had swimming pools in their back yard. What we needed to do was find a pool to sneak
into. Which we did and we swam. At some point we suspected the home owner had spotted
us and called the police and even as we clambered out of the pool lights approached in the
distance—for our late afternoon adventure had taken us into the evening. We ran along the
trail and back into something that resembled a trailer park, one of which was my home.
There, the voiceover which was my own (and yet not my voice), described in a Mark-Twain-
esque fashion: “The town was like an old cat, getting fatter and more neurotic with each
passing day.”

The Structure and Content of Dream Speech

In waking life, linguistic expression on all conceivable subjects is at least

hypothetically possible, although there is an assumption that talk will be situation-
ally appropriate. (A chef on a cooking show, e.g., will be expected to talk about
food, not car repair.) What preoccupies us in waking may also preoccupy us during
dreaming, so it is difficult to categorize the range of topics that dream speech may
express. But does the language of dream speech expression differ from that of
waking experience? If so, how? Consider the following aspects of dream speech
structure and content.

Patterns of linguistic structure in dream speech:

(C-1) Well-formed, logical statements of everyday content

(C-2) Structurally well-formed statements that are illogical or nonsensical in

meaning

(C-3) Neologisms and ill-formed structures, including fragments

Patterns of semantic content:

(C-4) Messages of advice, comfort, or instruction, either sought or unexpected

Well-formed, logical statements of everyday content (C-1). My survey of 500

dreams suggests that grammatically well-formed statements constitute the most
common form of dream speech (approximately 90% of the sample); about 60% of
the utterances dealt with ordinary topics. The following example from a woman in
her 50s contains both dream speech and a report of the dreamer’s thoughts, both of
which are quite ordinary:
 152 Kilroe

I am with a group of 4 –5 people, R. among them. We are walking up in front of a large ranch
house where we’re to give conference talks. R. is at the head of the group, I am second, and
the others are next to and behind me. One of them asks R. “Are you okay?” I look up,
thinking of course he’s okay, but I see he has tears in his eyes and is upset about something.
I wonder if I caused him to feel that way. I ask him, “What’s wrong?” To both the first
person’s and my question, he gives a neutral, noncommittal response, but I see that the tears
are about to overflow from his eyes.

Other than the fact that academic conferences are not commonly held in ranch
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

houses, the situation and the dream speech in this report bear such a resemblance
This document is copyrighted by the American Psychological Association or one of its allied publishers.

to ordinary experience that the dream report could credibly pass for a waking
narrative. The example of ordinary speech reinforces the findings of Meier (1993)
that the topics of much dream speech resemble those of ordinary waking life, as
well the findings of Barrett (2009) and Heynick (1993) that most dream utterances
are grammatically well-formed.
Structurally well-formed sentences that are illogical or nonsensical in meaning
(C-2). Freud’s famous “Irma Injection” dream report includes the following
statement: “M. said: ‘There’s no doubt it’s an infection, but no matter; dysentery
will supervene and the toxin will be eliminated.’” By Freud’s own analysis
(subsequently much deconstructed, though not relevant here), the statement is
nonsensical. In the summer of 1895 Freud had been providing psychoanalytic
treatment to a young family friend, “Irma.” The treatment had been only partially
successful, and Irma had been unwilling to accept Freud’s proposed solution, so
they agreed to suspend further treatment for the summer. On the day preceding the
dream, Freud’s friend “Otto” visited and reported that Irma was “better, but not
quite well.” Freud took this as a reproof of his methods, and that evening he wrote
out Irma’s case history in order to justify himself to the respected “Dr. M.” The
dream he had that night includes examination and discussion of patches and scabs
visible when Irma opens her mouth, which is what prompts Dr. M.’s nonsensical
statement about dysentery. Freud found meaning in the statement by applying his
then-fledgling method of association, but the point is that the dream statement itself
is grammatically correct but illogical when held up to the light of day-world
knowledge. (See below for the full report of “Irma’s Injection.”)
Neologisms and ill-formed structures, including fragments (C-3). A neolo-
gism is the creation of a new word by any of a variety of linguistic processes.
Neologisms produced in dreams were formally identified as such at least as early as
Freud and Kraepelin. Freud found many in his own and his patients’ dreams. One
of the best known is his dream of the following sentence: “It’s written in a positively
norekdal style.” (Freud, 1900/1953, p. 331) Freud’s initial sense of the meaning of
norekdal was that it was a parody of two German superlatives, kolossal and
pyramidal, but further analysis led him to unpack the word and discover it as a
blend of “Nora” and “Ekdal,” characters in Ibsen’s plays A Doll’s House and The
Wild Duck, respectively. Other than Freud, the most attention paid to neologisms
in dream speech to date has been by Kraepelin (1906) and Heynick (1993).
Perhaps it is ironically fitting, given his rejection of a productive role for speech
in dreams, to review the full text of Freud’s “Irma’s Injection” dream. Not only is
it the “specimen dream” Freud analyzed in detail to support his dream theory, but
it is also rich in dream speech.
 Dream Speech 153

A large hall—numerous guests, whom we were receiving.—Among them was Irma. I at once
took her to one side, as though to answer her letter and to reproach her for not having
accepted my “solution” yet. I said to her: “If you still get pains, it’s really only your fault.”
She replied: “If you only knew what pains I’ve got now in my throat and stomach and
abdomen—it’s choking me.”—I was alarmed and looked at her. She looked pale and puffy.
I thought to myself that after all I must be missing some organic trouble. I took her to the
window and looked down her throat, and she showed signs of recalcitrance, like women with
artificial dentures. I thought to myself that there was really no need for her to do that.—She
then opened her mouth properly and on the right I found a big white patch; at another place I
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

saw extensive whitish gray scabs upon some remarkable curly structures which were
This document is copyrighted by the American Psychological Association or one of its allied publishers.

evidently modeled on the turbinal bones of the nose.—I at once called in Dr. M., and he
repeated the examination and confirmed it. . . . Dr. M. looked quite different from usual; he
was very pale, he walked with a limp and his chin was clean-shaven. . . . My friend Otto was
now standing beside her as well, and my friend Leopold was percussing her through her
bodice and saying: “She has a dull area low down on the left.” He also indicated that a
portion of the skin on her left shoulder was infiltrated. (I noticed this, just as he did, in spite
of her dress.) . . . M. said: “There’s no doubt it’s an infection, but no matter; dysentery will
supervene and the toxin will be eliminated.” . . . We were directly aware, too, of the origin
of the infection. Not long before, when she was feeling unwell, my friend Otto had given her
an injection of a preparation of propyl, propyls . . . propionic acid . . . trimethylamin (and I
saw before me the formula for this printed in heavy type). . . . Injections of this sort ought
not to be given so thoughtlessly. . . . And probably the syringe had not been clean. (Freud,
1900/1953, pp. 139 –140)

In this single dream report, we have examples of the dreamer speaking to

another dream character, a dream character speaking to the dreamer, the dreamer’s
verbal thoughts, the dreamer participating in a general conversational interaction,
a grammatically well-formed but illogical statement, and the dreamer seeing written
language. As for whether or not any of the language in the dream is a repetition of
something that had been said or heard by Freud, Freud comments only that his
statement to Irma, “If you still get pains, it’s your own fault” is the kind of thing he
might have said to her and “may actually have done so” (Freud, 1900/1953, p. 141).
But it is a tentative explanation, and none of the other language of the dream is
accounted for in this way. We can take note, moreover, that by Freud’s own
analysis, both “Irma” and “Dr. M.” each represent a blend of two (and for Irma,
three) individuals known to Freud (cf. dream character Type A-5 above).
Semantic content: messages of advice, comfort, or instruction, either sought or
unexpected (C-4). The examples cited earlier of dream reports from ancient
literature all have the quality of being either oracular predictions of the future,
comfort regarding the dreamer’s waking distress, or advice to the dreamer. The
report in the previous section from Van de Castle about the boy who was reassured
by a man with a halo that he would walk again is evidence that dreams of this nature
continue to occur. But words of comfort and wisdom do not necessarily come from
an explicitly divine presence, as the following examples show.
In an interview with Clare Johnson (2014), Patricia Garfield reported a dream
she had almost a decade before, after her husband Zal passed away; the dream is
entitled “The Great Baptism.” The expression of comfort and wisdom comes from
a character identified as a “wise woman” and takes both written and spoken form:
I am with Zal at a gathering of Jewish people, listening to the words of a wise woman. She
speaks of a coin (the coin of life?). I see it, large, with a deckled edge. On one side is printed
the word ALWAYS. . . . On the other side of the coin is another word beginning with “A,”
perhaps AFTER or maybe AGONY—something connected with pain. The gist of her talk
is that the two sides are indivisibly linked, the great love and the afterward, the love and the
pain, but somehow it’s all OK: life is whole. . . .
 154 Kilroe

In the next example, reported by Deloges (2014), the dreamer, a medical

doctor and member of the Innu tribe of Quebec, receives an instructional message
from a deceased relative. The dream subsequently inspired the dreamer to change
the course of his life.
In this vivid dream, he saw his grandfather telling him to come back to Quebec and do
something useful for the people, suggesting that he organize a great walk through which First
Nations People would recover the pride of their origin. All people, First Nations and others,
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

would be welcomed to walk together in this intergenerational and multicultural trek.

This document is copyrighted by the American Psychological Association or one of its allied publishers.

In the final example of this type, the dreamer, who was struggling with the loss
of a loved one at the time of the dream, receives a spoken message from a friend
and in turn offers the friend some written advice:
Hosana (a weaver and friend of the dreamer) is at her loom, doing something on it, tying
knots at the top. I stand behind her and the loom. I see the loom is strung from low left to
high right [instead of the ordinary up and down and across]. I ask Hosana, “What is that?”
She replies, “It’s my own invention.” I ask, “What does it do?” She says, “It changes the
direction.” I give Hosana a small packet that says, “Invest in yourself.”

It is noteworthy that the name of the dreamed-about friend is “Hosana.” The

resemblance of this name to the Biblical exclamation “Hosanna,” used to express
worshipful praise or joy (from the Hebrew expression meaning “Save us, please!”),
was identified by the dreamer during group dream work, although the duplication
of what the White Knight said to Alice in the chapter of Through the Looking-Glass
bearing a title with the same words—“It’s my own invention”—was not.
While it is common for individuals to be comforted or advised by others in
waking experience, the examples above suggest that in dreams the provider of that
comfort and advice may appear as extraordinary or otherworldly, and the advice
itself may be life-changing or enigmatic.
In this section, several patterns of structure and one pattern of content in
dream speech have been identified, including the logic or illogic of statements, the
relative well-formedness of utterances, and apparent messages. The patterns
support findings from the research discussed in the previous section, that in general
dream speech is syntactically well-formed and lexically appropriate; the general
well-formedness of dream speech in turn supports the findings that higher-order
cognitive skills are active during dreaming. Moreover, errors of word choice and
sentence structure in dream speech only reinforce the observation that dream
speech resembles waking speech, where linguistic errors also occur.

Dream Speech, Waking Speech, and Consciously Imagined Speech

To what extent do dream speech and speech in the waking state resemble each
other? In waking, we cannot generally predict with certainty what others are
thinking or will say aloud, or how they will react to what we say. On the other hand,
as speakers we are capable of planning what we are going to say in thought before
we speak. And although we listen to our conversation partners, we are not unlike
actors in a play, preoccupied with our own part in a conversation—the points and
responses we intend to make. We engage in turn-taking, repair errors of discourse,
 Dream Speech 155

and follow other cultural conventions of interaction. Our utterances may vary from
brief fragments to long, complex statements.
In consciously imagined speech, by contrast, such as in daydreams or, for that
matter, the narratives of fiction, we impute speech to others, whether they are
known to us or are fantasy figures. When it involves people we know, the content
of this talk is not necessarily what they have said or even might be expected to say.
Imagined speech is just that, imagined, yet it draws from the content, structure, and
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

stylistic rules used in waking speech, as its general purpose would seem to be to
This document is copyrighted by the American Psychological Association or one of its allied publishers.

rehash, rehearse, or fantasize interactions with others. In this respect, dream speech
can be characterized as unconscious imagined speech. As examples here have
shown, dream speech can resemble ordinary waking speech in both form and
content; unlike waking interactions, however, yet similarly to imagined interactions,
dreamed interactions result from the dreamer-as-scriptwriter. Even if their speech
is consistent with what their waking counterparts might express, dream characters
say what we imagine them to say, albeit unconsciously.
In both structure and content, much of dream speech may pass for waking
speech, although generally in shorter and simpler utterance forms. Even the
oddities of dream speech such as neologisms and nonsense statements occur in
waking discourse, either as unintentional errors or as intentional products of the
creative use of language. But while the structure of dream speech may for the most
part be borrowed from that of waking speech, and while the topics of dream speech
may bear a strong resemblance to those of waking speech, interactions in dream
speech, like those of imagined speech, are products of mental activity alone. (Yet
curiously, speech errors are less likely to occur in consciously imagined speech than
in either waking speech or dream speech.) And if we go so far as to grant, with
Freud, that some dream utterances do echo utterances heard by the dreamer while
in the waking state, we must also acknowledge that some utterances spoken in the
waking state also echo utterances heard in the waking state.

Conclusion

Possible reasons why dream speech has been understudied by modern dream
researchers may include the fact that Freud and his followers, who dominated
dream research for much of the 20th century, held to the idea that dream speech is
mimicry of speech uttered in waking reality. Another possibility is the persistent
assumption, even among dream researchers, that dreaming is predominantly visual
in nature. Still another is the reluctance of dream-research scientists to study
content such as spoken messages in dreams because these continue to be associated
with religion and mysticism, and “messages” are necessarily viewed as antithetical
to any theory of dreams that holds dream content to be meaningless.
Yet dream speech data support work by Foulkes, Kahan and Laberge, and
others demonstrating that higher-order cognitive functions are active during
dreaming, and further study of dream speech has the potential to extend our
understanding of cognition during dreaming. Questions remain, however. Precisely
how are the topics, words and sentences of dream speech selected? What
neurolinguistic processes generating dream speech result in ordinary well-formed
utterances on the one hand, but nonsense statements, ill-formed utterances, or
 156 Kilroe

neologisms on the other? What is the relationship, if any, between the speech of
dream content and the inner speech that is at work during dreaming in the
construction of visual puns, conceptual metaphors, and possibly the fabric of the
dream narrative itself (for which see, e.g., Foulkes, 1985; Kilroe, 2000, 2013; Lakoff,
1997)? Further investigation of dream speech will determine its characteristics in
greater detail and will increase our understanding of dream content and dream
formation more broadly.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
This document is copyrighted by the American Psychological Association or one of its allied publishers.

References

Arkin, A. M. (1981). Sleep talking: Psychology and psychophysiology. Hillsdale, NJ: Erlbaum.
Barrett, D. (2009, June). Speech and language use in dreams. Paper presented at the 26th annual
International Association for the Study of Dreams Conference, Chicago, IL.
Cavallero, C., & Foulkes, D. (Eds.). (1993). Dreaming as cognition. New York, NY: Harvester
Wheatsheaf.
Delboeuf, I. (1885). Le sommeil et les rêves [Sleep and dreams]. Paris, France: Alcan.
Deloges, C. (2014). Traditional and modern day dreaming among the Innu of Quebec, Canada.
Dreamtime, 31, 7–8.
Domhoff, G. W. (2000). The misinterpretation of dreams. American Scientist, 88. Retrieved from
http://www.americanscientist.org/bookshelf/pub/the-misinterpretation-of-dreams
Foulkes, D. (1978). A grammar of dreams. New York, NY: Basic Books.
Foulkes, D. (1985). Dreaming: A cognitive–psychological analysis. Hillsdale, NJ: Erlbaum.
Foulkes, D. (1999). Children’s dreaming and the development of consciousness. Cambridge, MA:
Harvard University Press.
Freud, S. (1953). The interpretation of dreams (J. Strachey, Trans. & Ed.). New York, NY: Avon Books.
(Original work published 1900)
Hall, C. (1951). What people dream about. Scientific American, 184, 60 –63. http://dx.doi.org/10.1038/
scientificamerican0551-60
Hall, C., & Van de Castle, R. (1966). The content analysis of dreams. New York, NY: Appleton-Century-
Crofts.
Heynick, F. (1993). Language and its disturbances in dreams: The pioneering work of Freud and
Kraepelin updated. New York, NY: Wiley.
Isakower, O. (1954). Spoken words in dreams; a preliminary communication. Psychoanalytic Quarterly,
23, 1–6.
Johnson, C. (2014). Amazing dreamers. Dreamtime, 31, 27–28.
Kahan, T. L. (2011). Possible worlds, possible selves: Dreaming and the liberation of consciousness. In
S. Kakar (Ed.), On dreams and dreaming (pp. 109 –206). New Delhi, India: Viking.
Kahan, T. L., & LaBerge, S. (1994). Lucid dreaming as metacognition: Implications for cognitive science.
Consciousness and Cognition, 3, 246 –264. http://dx.doi.org/10.1006/ccog.1994.1014
Kahan, T. L., & LaBerge, S. (1996). Cognition and metacognition in dreaming and waking: Comparisons
of first and third-person ratings. Dreaming, 6, 235–249. http://dx.doi.org/10.1037/h0094459
Kahan, T. L., & LaBerge, S. P. (2011). Dreaming and waking: Similarities and differences revisited.
Consciousness and Cognition, 20, 494 –514. http://dx.doi.org/10.1016/j.concog.2010.09.002
Kilroe, P. (2000). The dream pun: What is a play on words without words? Dreaming, 10, 193–209.
http://dx.doi.org/10.1023/A:1009428923348
Kilroe, P. (2013). Inner speech in dreaming: A dialogic perspective. Dreaming, 23, 233–244.
http://dx.doi.org/10.1037/a0033654
Kraepelin, E. (1906). Über Sprachstörungen im Traume. Leizpig, Germany: Engelmann.
Lacan, J. (2006). The instance of the letter in the unconscious, or reason since Freud. In B. Fink (Trans.),
Ecrits: A selection (pp. 412–444). New York, NY: Norton. (Original work published 1966)
Lakoff, G. (1997). How unconscious metaphorical thought shapes dreams. In D. J. Stein (Ed.), Cognitive
science and psychoanalysis (pp. 89 –120). New York, NY: American Psychoanalytic Association.
Lincoln, J. L. (2003). The dream in Native American and other primitive cultures. Mineola, NY: Dover.
(Original work published 1935)
Massey, I. (1990). Freud before Freud: K. A. Scherner. Centennial Review, 34, 567–576.
Meier, B. (1993). Speech and thinking in dreams. In C. Cavallero & D. Foulkes (Eds.), Dreaming as
cognition (pp. 58 –76). New York, NY: Harvester Wheatsheaf.
National Sleep Foundation. (2015). Sleep talking. Retrieved from http://sleepfoundation.org/sleep-
disorders-problems/sleep-talking
 Dream Speech 157

New American Standard Bible. (1972). (text ed.). Carol Stream, IL: Creation House.
Plato. (1992). Theaetetus (B. Williams & M. Burnyeat, Eds., M. J. Levett, Trans.). Indianapolis, IN:
Hackett.
Robbins, M. (2004). Another look at dreaming: Disentangling Freud’s primary and secondary process
theories. Journal of the American Psychoanalytic Association, 52, 355–384. http://dx.doi.org/10.1177/
00030651040520021201
Scherner, K. (1861). Das Leben des Traums. Berlin, Germany: H. Schindler.
Shimizu, A., & Inoue, T. (1986). Dreamed speech and speech muscle activity. Psychophysiology, 23,
210 –214. http://dx.doi.org/10.1111/j.1469-8986.1986.tb00620.x
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Strauch, I., & Meier, B. (1996). In search of dreams: Results of experimental dream research. Albany, NY:
This document is copyrighted by the American Psychological Association or one of its allied publishers.

SUNY Press.
Van de Castle, R. (2014). Dreams and subtle energy: Light, sound, angels. Dreamtime, 31, 29 –30.
Wolchover, N. (2012, March 26). Why do people talk in their sleep? Live Science. Retrieved from
http://www.livescience.com/33794-people-talk-sleep.html

Correction to Vogelsang et al. (2016)

In the article “The Continuity Between Waking-Life Musical Activities

and Music Dreams” by Lukas Vogelsang, Sena Anold, Jannik Schormann,
Silja Wübbelmann, and Michael Schredl (Dreaming. Advance online
publication, March 24, 2016. http://dx.doi.org/10.1037/drm0000018), four
errors appear in text due to production errors. In the introduction, second
paragraph, the third sentence should read: Paul McCartney reported that
he woke up with a lovely tune in his head and first he didn’t believe it was
his own creation. In the Discussion section, second paragraph, the last
phrase of the first sentence should read: but much higher compared with
the sample of Schredl et al. (2015). In the Discussion section, the last
phrase of the third paragraph should read: were much higher and, thus,
detecting a substantial relationship was more likely. In the Discussion
section, seventh paragraph, the second to the last sentence should read:
This relationship has been shown for athletes. All versions of this article
have been corrected.

DOI: http://dx.doi.org/10.1037/drm0000029