Content uploaded by Alain Morin
Author content
All content in this area was uploaded by Alain Morin on Jul 09, 2019
Content may be subject to copyright.
Inner Speech
A Morin, Mount Royal University, Calgary, AB, Canada
ã2012 Elsevier Inc. All rights reserved.
Glossary
Ecological validity The degree to which the thoughts
recorded in a study reflect the thoughts that actually occur in
natural settings.
Inner speech The activity of talking to oneself in silence.
Left inferior frontal gyrus Also known as Broca’s area, the
brain region that gets reliably activated during inner speech
production.
Predicativeness The syntactically compressed, condensed,
abbreviated quality of inner speech. In inner speech, the
speaker tends to only articulate the predicate, not the
subject, for example, ‘hungry’ as opposed to ‘I am hungry.’
Private speech Self-directed speech emitted out loud by
children for self-regulatory purposes.
Rebound effect Rebound effect occurs when attempts at
suppressing unwanted thoughts actually lead to
experiencing them more often.
Scaffolding Scaffolding is adult assistance only on those
tasks that are beyond children’s skills.
Self-regulation Consists in altering one’s behavior, resisting
temptation, changing one’s mood, selecting a response from
various options, and filtering irrelevant information; also
includes setting immediate and distant goals, problem-
solving, planning, and decision-making.
Working memory System that keeps a small quantity of
information in an active state for a short period of time.
Zone of proximal development The difference between
what children can and can not do by themselves.
Overview
Definition
Inner speech is the activity of talking to oneself in silence. It is
often equated with the phonological loop portion of short-
term memory, which gets recruited when one needs to hold
auditory information in an active state for a short period of
time. However, inner speech does much more, as it plays a
central role in self-regulation (e.g., planning, problem-solving,
self-motivating), normal language functions such as reading
and writing, task-switching performance, remembering the
goals of action, rehearsing person-to-person communicative
encounters, and self-awareness, which includes mental time
travel into the past and the future. People report that around
one-fourth of their conscious waking life consists of inner
speech, indicating that it most certainly constitutes a significant
human mental activity.
Related Terms and Manifestations
Typical expressions used to designate inner speech are self-
talk, self-directed speech, subvocal, covert, or acommunica-
tive speech, auditory imagery, speech-for-self, propositional
thought, self-verbalizations, internal dialog or monolog, sub-
vocalizations, self-statements, and silent verbal thinking. Jean
Piaget used the term egocentric speech when discussing self-
talk produced aloud by children in social situations. He pro-
posed that egocentric speech served no function at all and
was the manifestation of children’s cognitive immaturity. Lev
Vygotsky insisted rather that egocentric (private) speech served
an important self-regulatory function. In Vygotsky’s view,
speech-for-self emitted out loud by children reflected intellec-
tual development, not egocentrism.
Young children engage in echolalia when they repeat
others’ words in an automatic manner for language learning
purposes or for the simple pleasure of using words. They emit
crib speech when producing soliloquies before falling asleep.
The idiom-embedded private speech refers to adults’ use of
private speech during public talks in order to restructure the
lecture or to self-regulate (e.g., OK, let’s now move on to the
next point). Inner speech writings represent hastily recorded
memos for self in notebooks, personal journals, on shopping
lists, etc. These messages may consist in single words or
phrases, or full paragraphs written in a very condensed and
often obscure style meaningful only to the self.
Theoretical Themes
Thought and Language
There exist two main views on the connection between thought
and language (which includes inner speech): language basi-
cally is thought and thought can exist without language. Plato,
an advocate of the first position, wrote that “When the mind is
thinking, it is simply talking to itself, asking questions and
answering them.” Following John Watson’s lead, behaviorists
also held that thought is identical to inner speech. Charles
Darwin seemed to agree with this theory when he proposed
that “A long and complex train of thought cannot be carried on
without the aid of words, whether spoken or silent, than a long
calculation without the use of figures or algebra.” Vygotsky
concurred and suggested that thought is not expressed in
words – it comes into existence through them. An alternative
formulation of this first position is the Sapir–Whorf hypothe-
sis, which suggests that the use of vocabulary that is unique
to one’s native language shapes one’s view of the world.
To illustrate, color terms within a language influence color
perception. Present work shows that using interrogative lan-
guage (e.g., Will I?) as opposed to declarative speech (I will)
increases motivation and goal-directed behavior.
The Wurzburg’s school of thought founded by Oswald
Kulpe in the late 1800s was a major proponent of the second
436
theory: pure thought can exist without language (or inner
speech); thought can be wordless. Experiments tentatively sup-
porting this position were conducted by Karl Buhler at the
beginning of the twentieth century. Participants listened to a
proverb and then were asked to press on a button once they
understood it. Most volunteers claimed that the button press-
ing was not accompanied by verbal thoughts or images, sug-
gesting that understanding (i.e., thought) precedes language.
None of these two extreme views are embraced by modern
cognitive scientists and linguists. Instead of asking if language
creates thought or if thought causes language, the question
being currently examined is: how does language affect thought
processes?
Vygotsky’s Position
Numerous theories of inner speech have been formulated.
Vygotsky’s theory, which emphasizes culture, language, and
internalization, arguably represents the most complete, origi-
nal, and coherent view available. In Vygotsky’s system, chil-
dren’s cognitive development is affected by culture in two
ways. First, children acquire most of their knowledge (the
contents of thought) through culture. In addition, not only
does culture teach children what to think but also how to
think. Intellectual growth emerges out of a dialectical process
in which problem-solving experiences are shared with parents,
teachers, siblings, peers, etc. Children can solve some problems
by themselves, yet other more challenging problems require
help from social agents. Vygotsky named the difference
between what children can and cannot do by themselves as
the zone of proximal development. He insisted that not
respecting this zone, either by helping children on tasks they
can complete on their own, or by not helping enough on
difficult tasks, impedes cognitive development. Ideally, people
interacting with children initially should guide most of the
problem-solving process and eventually transfer this responsi-
bility to the child. Language represents the core type of interac-
tion which allows social agents to convey information to
children. Gradually, children’s own language becomes their
principal tool of intellectual growth, first as speech-for-self
emitted aloud (private speech) to guide and control their
own actions and eventually as silent self-talk (inner speech).
Vygostky called this internalization – the process of using an
instrument of thought (inner speech) that was at first located
outside children (social speech). Much of Vygotsky’s theorizing
has lead to empirical predictions and most have received
support.
Methodology of Research
Like other internal cognitive processes, inner speech cannot be
directly observed but only indirectly inferred. Not surprisingly,
early efforts aimed at assessing inner speech were based on
introspection. Alfred Binet, for instance, would have his two
daughters work on various tasks and would ask them what
strategies they employed. He noted that the daughters often
reported thoughts such as “I told myself this ...”or“Isaid to
myself that ...”. From these observations, Binet deduced that
most thinking recruits inner speech.
Private Speech
Vygotsky held that studying the spontaneous production of
private speech in children was the most reliable way to tap
into inner speech. Although they both are not identical, pri-
vate speech, nonetheless, retains some characteristics that
inner speech is postulated to possess, and importantly, it
is objectively measurable and quantifiable. Indeed, private
speech has been extensively studied in natural settings (e.g.,
in the classroom) and in more controlled environments
in numerous situations (e.g., with others vs. alone). Private
speech is typically measured by calculating the total number
of raw utterances, the number of verbalizations per minute,
or the ratio of social to private speech. Private speech units
are coded and classified into different categories. Typical
categories are task-irrelevant private speech (e.g., word play,
emotional release, conversations with imaginary others),
task-relevant private speech (e.g., vocalizations about the
task or the child’s current or future task-related actions), and
partly internalized private speech consisting of inaudible
muttering, whispers, and silent lip movements. The frequency
and content of private speech are then correlated with behav-
iororperformance.
Only children’s private speech has been used as a window
into inner speech. Adults also engage in private speech, but
only when alone and thus much less frequently than children.
However, adults’ private writing has been investigated. Private
writing (also called introspective or inner speech writing)
refers to scribbling that people write on a piece of paper (e.g.,
a grocery list) or on the margins of a book, as well as in a
notebook (e.g., a writer’s thoughts). Private writing is consis-
tently condensed (like inner speech) and cryptic to others
because it is uniquely addressed to the self.
Production Methods
Production (or self-revelation) methods usually entail the exter-
nalization of ongoing inner speech as a task is being performed;
most are thus introspective, as opposed to retrospective. With
the think-out-loud technique, recordings of adults’ verbaliza-
tions are made as they are working on a given task; volunteers
are clearly encouraged to vocalize their thoughts. It is assumed
that the collected speech will mirror genuine inner speech, or at
any rate will supply a representative sample of it. To reduce the
likelihood of having participants produce a biased or unnatural
sample, instructions explicitly specify not to edit thoughts or to
worry about being understood. The videotape reconstruction
method consists in showing participants video recordings of
their behavior in precise situations (e.g., during task perfor-
mance) and asking them to recall (reconstruct) inner speech
content. A variation of this technique is the interview immedi-
ately following the execution of a given task, where volunteers
are probed for inner speech use during task completion. With
the thought listing technique, participants are asked to list their
subvocal activity once a task has been completed. The thought-
sampling method intends to collect a representative sample of
people’s inner speech in natural settings. Subjects wear a beep-
ing apparatus that produces audio signals at random intervals
throughout the day. Volunteers are asked to report the content
of their inner speech upon hearing the beep. All measurement
Inner Speech 437
techniques reviewed above include coding and classification of
inner speech units into different categories that are then corre-
lated with behavior or task performance.
Self-Reports
The most frequently used instrument to assess inner speech is
questionnaires made up of self-statements along a variety of
domains – for example, anxious versus nonanxious (This is
so stressful; I will make it), positive versus negative (I feel fine;
I just want to die), or social phobia (People don’t find me
interesting). Participants are invited to rate the frequency of
their self-talk using a Likert-scale. The main advantage of ques-
tionnaires is that they can easily and quickly be administered
to large groups of people. Conversely, because these scales
consist of premade self-verbalizations, they significantly restrict
the scope of natural inner speech that volunteers could other-
wise describe. Q-methodology is also used to estimate the con-
tent and/or functions of inner speech. Participants are invited to
sort out cards that depict inner speech units according to the
extent of agreement. Another approach, called the cued recall
technique, consists of eliciting self-reports of inner speech use
by presenting cards containing cue words to participants.
More ecologically valid methods exist. Researchers may use
an open-format question to gather spontaneously occurring
inner speech in people’s natural environments (e.g., When you
talk to yourself, what do you typically say to yourself?). Com-
pared to the random-sampling method, the open-ended format
approach suffers from being retrospective (and thus potentially
unreliable), but is not as limiting as the questionnaire. Slightly
less retrospective is the use of diaries (or first-person narratives),
which can also produce rich inner speech content.
Electromyographic Recordings
Alexander Sokolov wrote an entire book on the use of electro-
myographic recordings of movements of the lips and tongue to
assess inner speech incidence during problem-solving tasks.
Electromyography allows one to evaluate and record physio-
logical characteristics of muscles. This is done with an electro-
myograph that measures the electrical potential created by
muscle cells when these cells contract and are at rest. Move-
ments of the lips and tongue observed during overt speech are
also produced (with much less intensity) during covert speech.
These movements represent an objective external expression of
subvocal speech activity. Suction electrodes placed on the
tongue, sublingual horseshoe electrodes located under the
tongue, or surface electrodes attached to the lower lip are
used to record the target articulatory movements. It is thus
possible to infer inner speech activity (but not content) during
completion of various cognitive tasks such as mental calcula-
tion, silent reading, listening to speech, recall of verbal mate-
rial, and manipulation of graphic–visual information.
Articulatory Suppression
As seen above, articulatory movements accompany inner speech.
Logically, interfering with those movements (articulatory sup-
pression) will lead to inner speech interruption. In a typical
experiment, participants are invited to complete a task while
concurrently reciting verses or mentally counting backward
from 100. The forced articulation produced by the recitation or
counting blocks any other articulation that would be otherwise
required when spontaneously engaging in inner speech. Perfor-
mance deterioration implies that the task normally benefits from
inner speech use, although some argue that the deterioration
could be due merely to the extra attentional demands. Articula-
tory suppression clearly does not assess inner speech per se but
informs the researcher as to what types of cognitive work cannot
be accomplished without inner speech.
Brain-Imaging
Recent advances in brain-imaging techniques such as positron
emission tomography (PET) and functional magnetic resonance
imaging (fMRI) make it possible to record brain activity
of participants performing various cognitive tasks. Activation of
the left inferior frontal gyrus (LIFG) can be taken as evidence
of inner speech activity during task completion because the LIFG
is reliably recruited during the silent generation of words or
sentences (see Neuroanatomy section below). It will be possible
in the near future to identify which mental activitiesrely on inner
speech, as well as to dress a list of those which do not or partially
do so. To illustrate, preliminary work examining self-referential
activity suggests that focusing on one’s personality traits or
remembering personal past episodes engage inner speech,
whereas recognizing one’s face on a photograph does much
less so. The latest brain-imaging tool that has been developed is
repetitive transcranial magnetic stimulation (rTMS), which tem-
porarily interrupts normal cortical activity. Current experiments
are looking at the effects of applying rTSM on the LIFG. Brain-
imaging techniques do not allow access into the content of inner
speech – only its presence or absence is being objectified.
Summary
Each measurement technique described above presents specific
advantages and disadvantages. Ultimately, it is the type of prob-
lem being studied that should determine what method to select.
For instance, in a preliminary stage of an investigation, it would
be reasonable to employ an open-format or thought-sampling
approach to collect self-generated statements from depressed
patients. In a second part of the study, a validated questionnaire
could be created based on these vocalizations and administered
to larger clinical and nonclinical samples. Several assessment
techniques should be used to aim for convergent validity. For
example, in three independent studies on inner speech in anx-
ious people, one could use diaries (thus collecting self-generated
content), the videotape reconstruction procedure following
anxiety-producing social interactions, and a questionnaire
administered to anxious and nonanxious participants. Results
would then be integrated into one broader research report.
Development of Inner Speech
Social Origins
Vygotsky postulated that inner speech grows out of one’s social
environment. This hypothesis is supported by the fact that
there exist high positive correlations between rates of social
438 Inner Speech
interaction and private speech in children. Children immersed
in cognitively and linguistically stimulating environments
(situations more frequently observed in higher socioeco-
nomic status families) start using and internalizing private
speech faster than children from less fortunate backgrounds.
Indeed, children raised in environments characterized by low
verbal and social exchanges exhibit delays in private speech
development.
Development of Private Speech
Cross-sectional and longitudinal studies support Vygotsky’s
original view that the frequency of children’s private speech
follows an inverted-U relation with age. Private speech use
peaks at 3–4 years of age, decreases at 6–7 years of age, and
gradually fades out to be mostly internalized by age 10. As
private speech diminishes, one can observe more and more
frequent whispering and inaudible muttering – typical mani-
festations of partially internalized inner speech. Intellectually
gifted children tend to internalize private speech into inner
speech earlier than their average counterpart, and girls usually
show a faster private speech development than boys. Before
the age of 4, children are largely unaware of engaging in private
speech. The above trends seem to be universal and not sensitive
to cultural differences. One study examined private speech in
British and Saudi Arabian children and found no differences
in frequency of private speech.
Vygotsky suggested that fully internalized inner speech does
not reemerge as private speech. This assertion has recently been
challenged. Healthy adults have been observed to engage in
external speech when alone for self-regulatory purposes, as
well as for spatial navigation and search, concentration, and
emotional discharge and control. Ninety-six percent of all
adults report sometimes talking to themselves aloud.
Verbal Self-control
Alexander Luria conducted a set of experiments on verbal
mediation, aimed at determining the extent to which both
social and self-generated speeches effectively control children’s
behavior. Participants were instructed to press a rubber bulb
as they were told by the investigator to start, stop, or synchro-
nize presses with a flashing light and with his own speech. At
1.5–2.5 years, the instigating function of speech by the adult
(start) was successful but not the inhibiting function (stop).
Self-initiating and inhibiting functions of speech at that age
were lacking. At 3–4 years, both the initiating and inhibiting
functions of the experimenter’s speech and the initiating func-
tion of the child’s own verbal activity were objectified but not
the self-inhibiting function. The complete regulating function
of the child’s own speech was observed at 4.5 years.
The investigation of natural self-regulatory private speech in
children shows that it first follows behavior, then it accompa-
nies action, and finally it precedes it. Use of private speech for
self-regulatory purposes significantly increases between 3 and
4 years of age. Social agents who actively include children in a
collaborative effort to solve problems facilitate the develop-
ment of self-regulatory private speech. Joint work on common
problems helps children to internalize coping strategies and
produce individualized vocalizations adapted to solving pro-
blems encountered when alone.
The Private Speech Transition
A significant shift in children’s spontaneous use of private
speech occurs between the ages of 3 and 4. Three-year-old
children emit self-talk across a very large array of situa-
tions (e.g., goal-directed and unfocused activities), whereas
4-year-olds’ private speech tends to be observed almost exclu-
sively during self-selected but focused, sustained, goal-directed
activity. This indicates that 4-year-old children start engaging
in private speech in situations in which it is mostly required
for self-regulatory purposes.
Characteristics of Inner Speech
Fully internalized private speech transforms into inner speech,
and it takes a life of its own with distinctive characteristics
that differ from those of social speech. The semantic aspect of
speech becomes most prominent as its syntactic and phono-
logical dimensions move to the background. Unlike social
speech, inner speech is predicative: it is syntactically com-
pressed, condensed, and abbreviated. The context of speech
being always known to the self-talker, there is no need to
explicitly identify the subject of a thought. In inner speech,
sense prevails over meaning. The private significance of words
becomes more salient than their conventional meanings.
Agglutination refers to the creation of hybrid words that con-
stitute complex and uniquely personal concepts. The predica-
tive nature of inner speech explains why it is experienced as
a series of fragmented units – not as a smooth sequence of
fully developed verbal images. This actually accounts for the
observation that the rate of inner speech is significantly faster
than that of overt speech.
Inner speech often exhibits a dialogic quality because it
grows out of social speech, itself being mostly a dialog between
two persons. Inner speech is regularly described as including a
speaking self (the inner voice) and a self talked to (the inner
ear). Indeed, inner speech frequently adopts a format that
approximates social verbal interactions, with a series of alter-
nating lines, questions, and answers. Various acoustic qualities
of overt speech perception (e.g., sex, loudness, accent, and
language) are not reported during internal speech.
Neuroanatomy
Experiments employing PET and fMRI scans show that the
LIFG is more active during inner speech production elicited
by tasks such as silently articulating sentences or single words.
The LIFG is also recruited when subjects undertake working
memory tasks involving covert repetition of verbal material.
The LIFG corresponds to Broca’s area, which constitutes the
neurological foundation of both overt and inner speech gener-
ation. The LIFG is also referred to as the left ventrolateral
prefrontal cortex or left frontal operculum; it encompasses
Broadmann’s areas 44, 45, 46, and 47. Additional brain areas
are activated during inner speech use, including Wernicke’s
area, supplementary motor area, insula, and superior parietal
lobe on the left side, as well as right posterior cerebellar cortex.
Neuropsychological findings are consistent with brain-
imaging studies: accidental damage to the LIFG disrupts inner
Inner Speech 439
speech. rTMS applied to the LIFG interferes with internal speech.
rTMS applied to the motor cortex of the left hemisphere, which
is known to control mouth and tongue muscles, also inhibits
inner speech – an observation in line with the view that covert
speech necessitates articulation. The LIFG is not uniquely acti-
vated during inner speech production. Its activation can also be
recorded during cognitive control (the orchestration of thoughts
and actions in accordance with internal goals), working mem-
ory (temporary storage and manipulation of information),
selection among competing alternatives to guide response, and
interpretation of others’ behavior. Note that naturally occurring
inner speech, with its condensed and dialogic qualities, has
never been imaged; what has been measured so far is inner
speech defined as mental recitation.
Functions
Self-regulation
Overview
Fundamental human activities such as setting immediate and
distant goals, problem-solving, planning, and decision-making
are all part of a more global capacity called self-regulation. Pri-
vate speech in children and inner speech in adults have been
shown to be of critical importance for effective self-regulation.
Not surprisingly, the expression verbal self-guidance is
often used to designate self-regulation. Self-talk significantly
increases performance on tasks that entail the creation of com-
plex behavioral sequences and the concurrent judgment of
multiple behavioral choices. Four main groups of effective
self-verbalizations during problem-solving tasks have been
documented: (1) a clear definition of the problem; (2) an
efficient approach to the problem; (3) sustained concentration
on the problem; and (4) regular assessment of progress that
includes self-encouragement or strategic refocusing.
Private speech
Private speech use in children gradually increases as tasks
become more challenging. It is most often observed when
participants work on tasks that require executive control in
the absence of regulation offered by adults. Representative
tasks found in research reports are puzzles, memory tasks,
picture classification and discrimination, and sequencing
tasks. The following situations increase the likelihood of pri-
vate speech use for self-regulatory purposes: (1) completion of
goal-directed, academic, or problem-solving tasks as opposed
to free play or other activities; (2) working on difficult
yet achievable problem-solving tasks instead of easy tasks;
(3) being either alone or with peers in contrast to being with
a verbally regulating adult; and (4) working with an encourag-
ing as opposed to highly controlling adult. This is referred to as
scaffolding, which consists in adult assistance only on those
tasks that are beyond children’s skills.
That private speech use increases with task difficulty does
not automatically imply that it actually improves performance.
Two factors that need to be considered are task complexity and
concurrent versus future performance. Private speech will most
likely impede performance if the task is too difficult. It will
enhance performance if the task is within the child’s zone of
proximal development – that is, if it is within the child’s ability
range. Also, private speech use is more beneficial on the long
term (future) than on the short term (concurrent). Increased
performance following private speech use is delayed (dia-
chronic) instead of instantaneous (synchronic). Better perfor-
mance is typically observed during task sessions following the
time at which private speech was produced.
Self-talk use in sports
Numerous studies have examined the self-regulatory use of
self-talk in athletes while training and competing during sport-
ing activities such as water-polo, skating, golf, gymnastics, wres-
tling, and basketball. Self-verbalizations have been shown to
be superior to other mental strategies (e.g., mental imagery)
in enhancing sport performance. Dimensions of self-talk that
are usually measured or manipulated in studies are valence,
overtness, self-determination, self-instruction, self-motivation,
and frequency. Valence refers to the emotional content of self-
statements. While negative self-talk is associated with poorer
performance, positive self-talk does not significantly increase
it. Positive self-verbalizations are defined as those that help
to concentrate on the present, not on past errors or a remote
future. Negative self-statements constitute inappropriate, irra-
tional, counterproductive, or anxiety-producing thoughts.
Another quality of self-talk is overtness, as opposed to covert-
ness. It has been suggested that overt self-talk is more effective
at improving performance because it allowsthe athlete to repro-
duce situations where his or her coach verbally guides perfor-
mance aloud. Self-determined (i.e., self-selected) self-talk is
postulated to possess more motivational power than preformu-
lated self-talk assigned by a trainer. How-to-perform self-talk is
called self-instructional talk and is especially effective in practice
situations, whereas I-can-do-it talk represents motivational self-
talk and is more appropriate in competitive settings. Frequency
of self-talk often increases across various periods of a sporting
season and is linked to superior performance, although extreme
frequency (called paralysis by analysis) worsens it.
Language
Inner speech underlies basic language functions such as
reading, writing, speaking, and calculating. Deterioration of
inner speech caused by brain injury consistently produces
language disorders such as aphasia, agraphia, alexia, acalculia,
as well as reduced verbal short-term memory. Current studies
indicate that speakers monitor their own subvocal speech to
identify and correct phonological, lexical, or grammatical mis-
takes before they are spoken. Inner speech and crib speech are
often recruited when one spontaneously practices pronuncia-
tion and thinks about grammatical correctness. Significant lip
movements are observed during silent reading, and patients
with frontotemporal dementia and Gilles de la Tourette’s syn-
drome, who exhibit a loss of control over inner speech which
manifests itself in coprolalia (emitting unwanted vocalizations
in social situations), also cannot read in silence. These two
lines of evidence strongly suggest that inner speech and silent
reading are deeply associated.
Links between inner, oral, and written speech, reading, and
composing
As seen earlier, because inner speech is solely addressed to
the self, it is telegraphic, densely meaningful, fluid, dynamic,
440 Inner Speech
idiomatic, agglutinative, elliptic, and predicative. Written
speech, however, is produced for others and thus is fully
expanded, maximally elaborated, and orthographically explicit.
Composing refers to the process of putting ideas and words onto
paper; inner speech allows the writer to engage in mental
rehearsal, or planning, to determine what to write. Oral speech
lies in between innerand written speech – it is more spontaneous
and involuntary than written speech, it requires an interlocutor,
like inner speechit is dialogic, and it is made up of actual sounds.
Social speech develops first, followed by inner speech, then both
written speech and reading. But relationships between inner and
written speech are bidirectional. On the one hand, inner speech
serves as an internal rough draft for oral and written speech. On
the other hand, the very structure of written language forces an
analytic (metalinguistic) stance on inner speech. The sophisti-
cated, elaborated qualities of written speech encourage a less
spontaneous and involuntary use of inner speech. Not only
does inner speech constitutes the basis of written speech but
also underlies reading. In contrast with inner speech, writing
and reading require much more voluntary attention to the
phonological, syntactic, and semantic aspects of speech.
Language without inner speech?
One scientific article describes the case of a patient who
retained normal language functions despite being incapable
of inner speech. This evidence seems to contradict the claim
that language functions require inner speech. However, careful
neuropsychological assessment showed that the patient was
using mental imagery to compensate for the absence of inner
speech. That is, instead of verbalizing words to himself, the
patient could create clear images of words when talking to
himself – thus engaging in inner imaging as opposed to inner
speech. Apparently, this visual communication with oneself
effectively replaced inner speech and other language functions
such as reading and calculating.
Do deaf people talk to themselves?
Individuals born deaf never experience hearing sounds and
thus do not develop normal language skills, including inner
speech. Simply put, since deaf people would never have heard
spoken words, they cannot speak with or think in words. Can
they nonetheless engage in a form of wordless self-talk? The
answer is yes, with sign language. Use of sign language in deaf
individuals activates the LIFG – the brain area known to sus-
tain inner speech. As seen before, articulatory suppression
interferes with performance on tasks that benefit from inner
speech use in orally proficient individuals. A comparable dete-
rioration in task performance is obtained in sign users by
having signers hold objects firmly in their hands, which inhi-
bits self-sign language use. Also, deaf individuals tend to wave
hands when working on difficult tasks – the nonverbal equiva-
lent of private speech.
Memory
Working memory
Inner speech plays an important role in memory functions,
particularly in working memory. Working memory allows one
to keep a small quantity of information in an active state for a
short period of time. It is involved in numerous complex
cognitive abilities that include reasoning, decision-making,
problem-solving, and language understanding. Verbal and spa-
tial information are manipulated differently in autonomous
neuroanatomical systems. Each system is made up of three
functional parts with unique neural correlates. One compo-
nent stores information, another one rehearses – reactivates,
refreshes – it, and a third component (the executive one)
controls the overall processing of information in working
memory. Inner speech is synonymous with the rehearsal com-
ponent, at work, for example, when one mentally repeats a
phone number to assist later recall.
Autobiographical memory in bilinguals
Inner speech has also been shown to be involved in autobiog-
raphy, the memory of one’s own past. Phenomenological
inspection makes it clear that images are often used when
we store and recall personally relevant events. However, a
recent study suggests that autobiographical information is
also encoded and retrieved in words, or in narrative form.
A related issue is: in what inner language do bilinguals recall
their past? One cross-cultural study assessed inner speech of
autobiographical memory in Polish people who emigrated to
Denmark 30 years ago. Participants indicated retrieving per-
sonal memories in Polish for the decades prior to immigration
and in Danish after immigration. This observation thus sug-
gests that autobiographical material stored in one language is
more easily retrieved in the same language. All immigrants
spent 30 years in Denmark, but early immigrators (on average
24 years old at the time of immigration) indicated using more
Danish inner speech, while late immigrators (about 34 years
old at the time of immigration) reported more Polish inner
speech use.
Self-awareness
It is becoming increasingly evident that fully developed human
self-awareness depends on inner speech use. Self-awareness
refers to the ability to analyze the self and to build a self-
concept. It includes focus on private and public self-aspects
and consists in self-definition, self-recognition, self-evaluation,
self-esteem, mental time travel, and death awareness, to name
a few self-referential processes. Inner speech allows one to
identify and label various self-aspects – it makes it possible
for a person to represent internal states and reflect on them.
There is considerable empirical support for the notion of inner
speech participation in self-awareness. Multiple validated ques-
tionnaires measuring frequency of self-focus and use of inner
speech strongly correlate. Healthy participants claim to rely
significantly on inner speech while thinking about their future.
Brain-imaging studies of self-referential processing often report
activation of the LIFG and thus, most probably, inner speech
use. Brain injuries that produce a loss of inner speech are
associated with self-awareness deficits. The development of
theory-of-mind, which consists in thinking about others’ men-
tal states, is linked to private speech use in preschoolers.
Additional Functions
Inner speech does not uniquely play a role in self-regulation,
language, memory, and self-awareness. It is also involved in
Inner Speech 441
task-switching performance, the ability to switch back and
forth between two cognitive operations such as multiplying
and dividing numbers. Children learn to differentiate their
own voice from those of others by repeatedly hearing their
own voice through private speech use. People also talk to
themselves when mentally preparing an upcoming social
encounter. Inner speech is also activated when one focuses
attention on a given target, forms concepts, remembers the
goals of actions, or expresses emotions such as anger. Some
have proposed that praying engages inner speech.
Dysfunctional Inner Speech
Overview
Ruminative inner speech that focuses on real or imagined
negative self-aspects may amplify preexisting morbid conditions
such as test anxiety, bulimia, anorexia, lack of assertiveness,
insomnia, social anxiety, agoraphobia, sexual dysfunctions,
low self-esteem, and depression. Inner speech may also accom-
pany less severe and temporary negative states such as worry,
guilt, and shame. The content-specificity hypothesis suggests
that the aforementioned maladaptive conditions underlie nega-
tive self-talk associated with specific dysfunctional topics. To
illustrate, anorexia triggers inner speech revolving around
weight and physical appearances issues (e.g.,I must lose weight),
and low self-esteem is linked to self-worth cognition (e.g., I’m
insignificant). Accordingly, compulsive gamblers have irrational
thoughts about control of the game (e.g., I’m going to bet on
those rows again, this is a good game).
Asymmetry Between Positive and Negative Self-verbalizations
There exists an important and counterintuitive asymmetry
between positive and negative self-statements. Negative inner
speech has a significantly more dysfunctional influence than
positive self-talk on coping. Simply put, imagining the worst
through pessimistic inner speech (e.g., failing an exam) has
more negative impact than thinking positively (passing
an exam). This fact clearly goes against the fashionable belief
of positive thinking and suggests that eradicating negative
verbalizations might be more efficient than formulating posi-
tive ones. Healthy functioning is associated with a 1.7:1 ratio
of positive to negative self-verbalizations; maladaptive think-
ing is characterized by a 1:1 ratio. Studies that measure cogni-
tive change following psychotherapy confirm that negative
thoughts tend to decline while positive thoughts do not
increase. Negative verbalizations (e.g., disfigurement) increase
heart rate, but positive ones (e.g., harmony) have no signifi-
cant effect.
The Rebound Effect
Also perplexing is the observation that trying to suppress unde-
sirable thoughts in inner speech (e.g., I must stop thinking
about this) makes the thoughts more accessible (they are expe-
rienced more often) and leads to increased anxiety, depressed
mood, and lower self-esteem. This paradoxical phenomenon is
called the rebound effect.
Hearing Voices in Schizophrenia
The most plausible explanation for auditory verbal hallucina-
tions in schizophrenic patients is that they cannot properly
monitor their own self-generated inner speech. There is little
doubt that the voices are the result of patients’ own subvocal
activity. Indeed, the LIFG is active when schizophrenic patients
are experiencing verbal auditory hallucinations, suggesting
inner speech production. The precise underlying mechanism
is unknown, but one view implicates a verbal self-monitoring
deficit. Speech production in normal individuals generates a
corollary discharge that sends a message to the left temporal
lobe where verbal thoughts are identified. Basically, the mes-
sage tells the brain that it just produced speech and thus
the voice heard is one’s own inner speech. This communica-
tion between the frontal and temporal lobes is postulated to
be deficient in schizophrenic patients. Brain-imaging studies
support this hypothesis: patients with auditory hallucina-
tions show activity in the LIFG when engaging in inner speech
but fail to exhibit activity in the left temporal cortex. That
verbal hallucinations would be caused by an inner speech
dysfunction per se is improbable. Performance on short-term
memory tasks that require inner speech use is normal in
patients experiencing severe auditory hallucinations.
Hyperactivity
Initial explanations of hyperactivity in children involved the
notion of insufficient self-regulatory private (and inner) speech
causing poor self-control. Cognitive-behavioral psycholo-
gists designed various techniques to increase the use of verbal
self-guidance. A typical approach consisted of gradual stages
producing the internalization of self-regulatory speech –
for example, modeling, overt external guidance, overt self-
guidance, faded overt self-guidance, and covert self-guidance.
This method has recently been reevaluated, and most agree that
it is actually ineffective. First, the aforementioned approach
focuses on short-term results that apply to precise tasks; what
would be needed is a technique leading to long-term gains that
would generalize to broad academic and interpersonal beha-
viors. What used to be taught was self-control (i.e., copying
adults’ commands) instead of authentic self-regulation (i.e.,
self-generating flexible plans for action). Second, in point of
fact, hyperactive kids produce adequate self-regulatory private
speech. The nature of the deficit does not lie in lack of verbal
self-guidance – it is most likely neurobiological.
Conclusion
Inner speech constitutes a fundamental cognitive activity
not limited to working memory (the phonological loop) but
including self-regulation (initiating, shaping, guiding, and
controlling behavior), language functions, self-awareness
(e.g., autobiographical memory, mental time travel), emo-
tional release, task-switching, preparation for upcoming social
encounters, and more. Various methods exist to measure inner
speech, some easy to administer but limiting (e.g., question-
naires) and others more ecologically valid but time consuming
(e.g., thought sampling). The general agreement is that inner
442 Inner Speech
speech is social in origin, it is preceded by private speech, and
once internalized, it becomes mostly abbreviated and predica-
tive. Although various brain regions underlie inner speech
production, the LIFG seems to represent the most important
area. Inner speech resembles a double-edged sword – on the
one hand, it is associated with positive consequences such as
self-regulation, yet on the other hand, negative and ruminative
self-talk may lead to or maintain psychological disorders such
as anxiety and depression. Inner speech, nonetheless, remains
neglected compared to other important psychological con-
cepts. To illustrate, what people actually talk to themselves
about is largely unknown and current work is precisely aiming
at answering that and other exciting research questions.
See also: Agraphia and Alexia;Aphasia;Autobiographical
Remembering and the Self;Bilingualism and Multilingualism;
Language Development;Planning;Problem Solving;Psychology of
Reading;Schizophrenia;Social Cognition.
Further Reading
Berk LA (1992) Children’s private speech: An overview of theory and the status of
research. In: Diaz RM and Berk LE (eds.) Private Speech: From Social Interaction to
Self-regulation, pp. 17–53. Hillsdale, NJ: Lawrence Erlbaum Associates.
De Guerrero MCM (2005) Methodology of research on inner speech. In: De Guerrero M
(ed.) Inner Speech – L2: Thinking Words in a Second Language, Chapter 2, pp. 89–
118. New York: Springer.
Fuson KC (1979) The development of self-regulating aspects of speech: A review. In:
Zivin G (ed.) The Development of Self-regulation Through Private Speech,
pp. 135–218. New York: Wiley.
Hardy J (2005) Speaking clearly: A critical review of the self-talk literature.
Psychology of Sport and Exercise 7: 81–97.
Jones SR and Fernyhough C (2007) Neural correlates of inner speech and auditory
verbal hallucinations: A critical review and theoretical integration. Clinical
Psychology Review 27: 140–154.
Kendall PC and Hollon SD (1981) Assessing self-referent speech: Methods in
measurement of self-statements. In: Kendall PC and Hollon SD (eds.) Assessment
Strategies for Cognitive-Behavioral Interventions, pp. 85–118. New York:
Academic Press.
Larsen SF, Robert W, Schrauf RW, Fromholt P, and Rubin DC (2002) Inner speech
and bilingual autobiographical memory: A Polish–Danish cross-cultural study.
Memory 10: 45–54.
Levine DN, Calvanio R, and Popovics A (1982) Language in the absence of inner
speech. Neuropsychologia 20: 391–409.
Meichenbaum D (1977) Cognitive-Behavior Modification: An Integrative Approach.
New York: Plenum.
Morin A (2005) Possible links between self-awareness and inner speech: Theoretical
background, underlying mechanisms, and empirical evidence. Journal of
Consciousness Studies 12: 115–134.
Schwartz RM (1986) The internal dialogue: On the asymmetry between
positive and negative coping thoughts. Cognitive Therapy and Research
10: 591–605.
Sokolov AN (1972) Inner Speech and Thought. New York: Plenum.
Vygotsky LS (1943/1962) Thought and Language. Cambridge: MIT.
Winsler A (2009) Still talking to ourselves after all these years: A review of
current research on private speech. In: Winsler A, Fernyhough C, and
Montero I (eds.) Private Speech, Executive Functioning, and the
Development of Verbal Self-regulation, pp. 3–41. New York: Cambridge
University Press.
Zivin G (1979) Removing common confusions about egocentric speech, private speech,
and self-regulation. In: Zivin G (ed.) The Development of Self-regulation Through
Private Speech, pp. 13–50. New York: Wiley.
Inner Speech 443
