You are on page 1of 138

The Core Deficit in

Autism and Disorders

of Relating and
The Journal Of Interdisciplinary Council on
Developmental Developmental and Learning Disorders
and Learning

Special Edition Simon Baron-Cohen, Ph.D., et. al.

Christopher Gillberg, M.D.
Volume 5 Stanley I. Greenspan, M.D.
Number 1 Nancy Minshew, M.D.
2001 Peter Tanguay, M.D., et. al.
Andrew W. Zimmerman, M.D. and
Barry Gordon, M.D., Ph.D.
Journal of Developmental and Learning Disorders

The goal of The Journal of Developmental and Learning Disorders is to improve the identifica-
tion, prevention and treatment of disorders that interfere with adaptive developmental and
learning processes.
Many disciplines work with these disorders, including, for example, child development,
psychology, psychiatry (child and adult), social work, education (including special and early
childhood), neurology, pediatrics, speech language pathology, occupational and physical ther-
apy, nursing and others. Those who work with developmental and learning disorders often
have opportunities to communicate with colleagues within their own disciplines, but less often
with those in other disciplines who share the same concerns and interests. Yet, each of the dis-
ciplines is discovering its overlap with other fields, as well as new ways to identify, prevent and
treat developmental problems. There is also enormous growth in research on the development
of the nervous system and the various physical, psychological, and cognitive factors that influ-
ence it. Many of these discoveries are informing the development of new diagnostic and inter-
vention approaches.
The Journal will present research, clinical and case studies from various disciplines work-
ing with developmental and learning disorders. It will also report on new information, events,
services, training, and networking activities taking place throughout the country and the world,
providing periodic information updates and overviews. The Journal will cover a range of top-
ics, including overviews of important and growing areas, new research, clinical observations,
intervention studies, discussions of clinical practice, and educational interventions.
As of the fall 2001 the Journal has begun a new policy and will be published in hard copy
once a year. Articles that have been peer reviewed and accepted for publication in our Journal
will be posted on our website at WWW.ICDL.COM. Each fall we will publish the Journal in
hard copy containing all of the articles for that year.

Call for Papers

The Journal of Developmental and Learning Disorders, an official publication of the

Interdisciplinary Council on Developmental and Learning Disorders, encourages original
submissions. Papers will be peer reviewed. Please submit five copies, double spaced and a disk
in Word format. Tables, charts and pictures should be submitted in camera ready format. The
title page should include the author’s full name, street and e-mail address, telephone and fax
numbers, affiliation, and a 150 word summary of contents (abstract). Reference format is
American Psychological Association style.
Send submissions to: Ms. J. Raphael, MSW, 3213 Midfield Road, Baltimore,
Maryland 21208. Phone or fax: (410) 486-1251. E-mail: JO@ICDL.COM.
For subscription information: (301) 656-2667 or WWW.ICDL.COM.

Volume 5 2001 Number 1


The Affect Diathesis Hypothesis: The Role of Emotions in the Core Deficit in 1
Autism and in the Development of Intelligence and Social Skills—
Stanley I. Greenspan, M.D.
Studies of Theory of Mind: Are Intuitive Physics and Intuitive Psychology 47
Independent?—Simon Baron-Cohen, Sally Wheelwright, Amanda Spong,
Victoria Scahill and John Lawson
Asperger Syndrome and High Functioning Autism: Shared Deficits or 81
Different Disorders?—Christopher Gillberg, M.D., Ph.D.
A System of Classification for Autism Spectrum Disorder—Peter E. Tanguay, 95
M.D., Julia Robertson, M.D., Ann Derrick, A.R.N.P., M.S., C.S.
The Core Deficit in Autism and Autistic Spectrum Disorders— 107
Nancy J. Minshew, M.D.
Neural Mechanisms in Autism—Andrew W. Zimmerman, MD and 119
Barry Gordon, M.D., Ph.D.

Bethesda, Maryland Copyright © 2001
Journal of Developmental and Learning Disorders
Stanley I. Greenspan, M.D.
George Washington University

Associate Editor
Serena Wieder, Ph.D.

Assistant Editor
Georgia DeGangi, Ph.D.
Treatment and Learning Center

Administrative Editor
Jo Raphael, M.S.W.

Editorial Board

Margaret Bauman, M.D. Toby Long, Ph.D., P.T.

Harvard University Georgetown University
Harry Chugani, M.D. Stephen W. Porges, Ph.D.
Wayne State University University of Maryland
Leon Cytryn, M.D. Barry Prizant, Ph.D., C.C.C.-S.L.P.
George Washington University Brown University
Sima Gerber, Ph.D. Ricki G. Robinson, M.D., Ph. D.
Queens College University of Southern California
Arnold P. Gold, M.D. Rebecca Shahmoon Shanok, M.S.W., Ph.D.
Columbia University Child Development Center, New York
Myron Hofer, M.D. Milton Shore, Ph.D.
Columbia University Catholic University
Pnina Klein, Ph.D. Richard Solomon, M.D.
Bar-Ilan University, Israel University of Michigan
Pat Lindamood, M.S., C.C.C.-S.L.P.
Lindamood-Bell Learning Processes
The Role of Emotions in the Core Deficit in
Autism and in the Development of Intelligence
and Social Skills

Stanley I. Greenspan, M.D.

Abstract. In this paper we will explore the role of affect in the core deficit in autism and
in the development of intelligence and social skills. We discuss how children with autistic spec-
trum disorders may uniquely, for biological reasons, miss a critical developmental capacity, the
ability to connect affect or intent to motor planning and sequencing capacities as well as sym-
bol formation and, therefore, have a difficult time engaging in the long reciprocal chains of
affective interaction so necessary for creative and abstract thinking and high-level social skills
(Affect Diathesis Hypothesis). We will also discuss how these same affective interactions
underlie intelligence and social development. Additionally, we explain that to improve assess-
ments and interventions for children with a variety of challenges including autistic spectrum
disorders, it is imperative to appreciate the role of affective interchanges in disordered and
healthy development. Finally, we explain that to fully operationalize the role of affective inter-
action and the Affect Diathesis Hypothesis for the assessment and intervention process, we have
formulated the Developmental, Individual-Difference, Relationship-Based model (DIR)
(Greenspan, 1992, 1997b; Greenspan & Wieder, 1997, 1998, 1999).

The Affect Diathesis Hypothesis


There is a growing appreciation of the role of emotional interactions in human

development. It’s been long acknowledged that affective exchanges influence such
basic capacities as the formation of relationships, self esteem, and impulse control.
Recent studies suggest that emotional interaction in infancy and early childhood
also influences cognitive and language capacities. For example, higher quality care,
including emotionally sensitive caregiving, is associated with stronger cognitive,
language, and emotional and social development (NICHD, 1998, 1999, 2000;
Vandell & Wolfe, 2000; Peisner-Feinberg, et al., 1999). Risk factors that undermine


caregiver and family emotional functioning are associated with compromises to

intellectual functioning (Sameroff, Seifer, Barocas, Zax, & Greenspan, 1986;
Sameroff, Seifer, Baldwin, & Baldwin, 1993). A large empirical base, neurological
research looking at brain lesions interfering with emotional regulation, explo-
rations of the types of thinking that are part of skillful social interactions (i.e., emo-
tional intelligence), and concepts of multiple intelligences have further increased
interest in the role of emotions (Shonkoff & Phillips, 2000; Damasio, 1994;
Goleman, 1995; Gardner, 1983).
In spite of greater interest in the role of emotions in human development, how-
ever, there has not been sufficient understanding of how emotions exert their influ-
ence. How do emotions and emotional interactions affect intelligence and its related
cognitive and language capacities as well as many complex social and self-regulation
capacities? What are the psychological mechanisms of action by which affects work
on these different aspects of the mind?
In The Growth of the Mind (Greenspan, 1997), we presented a theory of the process
through which emotional interactions influence intelligence. This theory suggested
that affective interactions emerge earlier than the sensorimotor schemes postulated
by Piaget (1962) and that they are the most primary probes we use to understand,
conceptualize, and “double code” our experiences with the world. It also suggested
that most types of abstract thinking are based on reflections on these personal affect-
ive experiences.
In this article, we further develop this theory regarding how emotional processes
influence various aspects of the mind. The Affect Diathesis Hypothesis examines
the critical role of affective interactions in self-regulation, communication, language,
creating meanings, and constructing a sense of reality. It also examines how various
types of deficits in the expectable diathesis (i.e., spread) of affects during early devel-
opment contributes to our understanding of autistic phenomena and other develop-
mental and emotional problems.

Part 1: The Affect Diathesis Hypothesis and Autistic

Spectrum Disorders

Among the many symptoms of autism, language, cognitive, and social deficits
are prominent. Recent studies suggest that these compromises can be conceptual-
ized as a series of functional developmental deficits. The pattern of these deficits
provide clues regarding a core psychological mechanism that may express the
neurological differences characterizing autistic spectrum disorders. When children
with autism are compared to children without autism, and level of intelligence, as
measured with IQ tests, is controlled for, there are a number of autism-specific
functional developmental problems. These include deficits in the ability for empa-
thy and seeing the world from another person’s perspective in both physical and
emotional contexts (theory of mind) (Baron-Cohen, 1994); higher-level abstract
thinking, including making inferences (Minshew & Goldstein, 1998); and shared

attention, including social referencing and problem-solving (Mundy, Sigman, &

Kasari, 1990). In addition, deficits in the capacities for affective reciprocity
(Baranek, 1999; Dawson & Galpert, 1990; Lewy & Dawson, 1992; Osterling &
Dawson, 1994; Tanguay, 1999; Tanguay, Robertson, & Derrick, 1998) and func-
tional (pragmatic) language (Wetherby & Prizant, 1993) also appear specific to
Do these functional developmental deficits stem from a common pathway? In
clinical work with infants and children with biological and environmental challenges
and without challenges, we have found that the capacities for empathy, psychological
mindedness, abstract thinking, social problem-solving, functional language, and
affective reciprocity all stem from the infant’s ability to connect affect or intent to
motor planning capacities and emerging symbols (Greenspan, 1979, 1989, 1997a).
Relative deficits in this core capacity leads to problems in higher-level emotional and
intellectual processes. The core psychological deficit in autism may, therefore,
involve an inability to connect affect (i.e., intent) to motor planning and sequencing
capacities and symbol formation.
A child’s capacity to connect affect to motor planning and emerging symbols
becomes relatively apparent between 9 and 18 months of age as the infant shifts from
simple patterns of engagement and reciprocity to complex chains of affective reci-
procity that involve problem-solving interactions. Consider a 14-month-old child
who takes his father by the hand and pulls him to the toy area, points to the shelf,
and motions for a toy. As Dad picks him up, and he reaches for and gets the toy, he
nods, smiles, and bubbles with pleasure. For this complex, problem-solving social
interaction to occur, the infant needs to have an emotional desire or wish (i.e., intent
or affective interest) that indicates what he wants. The infant then needs to con-
nect his desire or affective interest to an action plan (i.e., a plan to get his toy). The
direction-giving affects and the action plan together enable the child to create a pat-
tern of meaningful, social, problem-solving interactions. Without this connection
between affect and action plans, complex interactive problem-solving patterns are
not possible. Action plans without affective direction or meaning tend to become
repetitive (perseverative), aimless, or self-stimulatory, which is what is observed
when there is a deficit in this core capacity.
As the ability to form symbols emerges, the child needs to connect her inner
affects (intent) to symbols to create meaningful ideas, such as those involved in func-
tional language, imagination, and creative and logical thought. The meaningful use
of symbols usually emerges from earlier and continuing meaningful (affect-mediated)
problem-solving interactions that enable a toddler to understand the patterns in her
world and eventually use symbols to convey these patterns in thought and dialogue.
Without affective connections, symbols like action plans are used in a repetitive
(perseverative) manner (e.g., scripting, echolalia).
The capacity to connect affect to action plans and symbols is likely part of a larger
transformation of affect. The infant goes from global and/or catastrophic affective
patterns (in the early months of life) to reciprocal ones. The capacity for engaging in
a continuous flow of reciprocal affective interactions enables the child to modulate

mood and behavior, functional preverbal and verbal communication, and thinking.
It also enables more flexible scanning of the environment because the child gets feed-
back from what he sees and, based on that feedback, explores further. There is, there-
fore, more integrated visual-spatial and motor functioning because intense global
affects push for discharge and vigilant or overly focused or highly distractible visual-
motor patterns, whereas long chains of reciprocal interaction support back-and-forth
exploration of the environment and, therefore, flexible, broad, integrated perceptual
In facilitating back-and-forth interaction with the environment, the capacity
for reciprocal interaction also facilitates associative learning. Associative learning
(building up a reservoir of related experiences, thoughts, feelings, and behaviors
which give range and depth to one’s personality, inner life, and adaptive responses)
is necessary for healthy mental growth. Its absence leads to rigid, mechanical
feelings, thinking, and behavior patterns, as are often seen in autistic spectrum
Reciprocal, affective interactions and affectively-guided problem-solving inter-
actions and symbols, as can be seen, are necessary for all the unique capacities that
distinguish individuals with autism from individuals without autism, as outlined
earlier. For example, long chains of social reciprocity depend on affect guiding
interactive social behavior. Shared attention, which includes social referencing and
shared problem-solving, also depends on affect guiding interactive social behavior.
Empathy and theory of mind capacities depend on the ability to understand both
one’s own affects or feelings and another person’s affects or feelings and to project
oneself into the other person’s mindset. This complex emotional and cognitive task
begins with the ability to exchange affect signals with another person and, through
these exchanges, emotionally sense one’s own intent and the other person’s intent
through a sense of “self” in interaction with another. Similarly, higher-level abstract
thinking skills, such as making inferences, depend on the ability to generate new
ideas from one’s own affective experiences and then reflect on and categorize them
(Greenspan, 1997a).
In observations of infants and toddlers heading into autistic patterns and in taking
careful histories of older children with autism, we noted that children with autistic
spectrum patterns did not fully make the transition from simple patterns of engage-
ment and interaction into complex affect-mediated, social problem-solving interac-
tions. They, by and large, did not progress to a continuous flow of back-and-forth,
affective, problem-solving interchanges (i.e., a continuous flow of circles of commu-
nication). Even affectionate children who were repeating a few words or memorizing
numbers and letters, and who went on to evidence autistic patterns, did not master,
for the most part, this early capacity to engage in a continuous flow of affect-
mediated, gestural interactions. They were also unable to develop empathy and
creative and abstract thinking unless they were involved in an intervention that
focused on facilitating affect-mediated interactions.
In a review of the functional developmental profiles of 200 children with autistic
spectrum disorders, we observed that most of the children shared this unique

processing deficit. Approximately two-thirds of the children who developed autistic

spectrum disorders had this unique type of biologically based processing deficit that
involved the connection of affect or intent to motor planning and sequencing capac-
ities as well as to emerging symbolic capacities (Greenspan & Wieder, 1997). At the
same time, however, the children differed with regard to other processing deficits
involving their auditory, motor planning, visual-spatial, and sensory modulation abil-
ities. These differences accounted for the different types and degrees of language,
motor, and cognitive impairments that accompany the fundamental deficit in engag-
ing in fully purposeful and meaningful social and intellectual interactions.
The hypothesis that explores the connection between affect and different pro-
cessing capacities (the core deficit) is called the affect diathesis hypothesis. In this
hypothesis, as indicated previously, a child uses his affect to provide intent (i.e., direc-
tion) for his actions and meaning for his words. Typically, during the second year of
life, a child begins to use his affect to guide intentional problem-solving behavior
and, later on, meaningful use of symbols and language. Through many affective
problem-solving interactions, the child develops complex social skills and higher-
level emotional and intellectual capacities.
Because this unique processing deficit occurs early in life, it can undermine the
toddler’s capacity to engage in expectable learning interactions essential for many
critical emotional and cognitive skills. For example, she may have more difficulty
eliciting ordinary expectable interactions from her parents and the people in her
immediate environment. She may perplex, confuse, frustrate, and undermine pur-
poseful, interactive communication with even very competent parents. Without
appropriate interaction, she may not be able to comprehend the rules of complex
social interactions or to develop a sense of self. These may include implicit social
functions and social “rules,” and developing friendships and a sense of humor, which
are learned at an especially rapid rate between 12 and 24 months of age (Bell, 1970;
Emde, Biringen, Clyman, & Oppenheim, 1991; Greenspan, 1979, 1997a; Kagan,
1981; Piaget 1981/1954; Werner & Kaplan, 1963; Winnicott, 1931). By the time a
child with processing difficulties receives professional attention, her challenging
interaction patterns with her caregivers have, therefore, excluded her from important
learning interactions and may be intensifying her difficulties. The loss of engagement
and intentional, interactive relatedness to key caregivers may cause her to withdraw
more idiosyncratically into her own world and become even more aimless and/or
repetitive. What later looks like a primary biological deficit may, therefore, be part
of a dynamic process through which the child’s lack of affective reciprocal interac-
tions has intensified specific, early, biologically-based processing problems and
derailed the learning of critical social and intellectual skills.
The capacity for long chains of affective reciprocity may have early roots which
can potentially lead to earlier opportunities for identifying risk patterns. A precursor
of the capacity to connect affect to motor planning and symbol formation in the
second year of life may be the capacity observed initially in the early months of life
for connecting motor actions to interactive, affective rhythms conveyed through
facial expressions, vocalizations, or other gestures. For example, babies will move

rhythmically to the rhythm of their mother’s voice (Condon & Sander, 1974;
Condon, 1975). We are currently conducting studies to see if children at risk for autis-
tic spectrum disorders evidence a deficit in their earlier, rhythmic affective-motor
patterns as well. The Affect Diathesis Hypothesis may, therefore, involve a number
of developmental levels. This possibility has important implications for early identi-
fication and preventive strategies.
Biologically based processing (regulatory) difficulties often contribute to, but are
not always decisive in, determining relationship and communication difficulties.
When problems are perceived early, appropriate professional help can, to varying
degrees, teach children and caregivers how to work with the processing (regulatory)
dysfunction, including helping the toddler connect affect to emerging action plans
and associated relationship and communication patterns. Many children can become
capable of forming warm relationships and, to varying degrees, climb the develop-
mental ladder leading to language and thinking capacities.
There are many children who do not evidence autism but have developmental
problems in which intentionality or purposeful action is difficult in its own right (e.g.,
severe motor problems). Such problems result in less practice in using intentional
behavior and participating in intentional interactions. These children may, therefore,
either have difficulty forming or may secondarily lose their ability to connect intent
or affect to motor planning because they are unable to exercise this critical function
(i.e., their impaired motor skills make purposeful action difficult). In these circum-
stances, creating purposeful interactions around any motor skill (e.g., head or tongue
movements) may strengthen the affect-motor connection and reduce aimless, repeti-
tive behavior, thereby facilitating problem-solving and thinking. Recent MRI studies
suggest that practicing and improving motor skills may enhance the developmental
plasticity of neuronal connections (Zimmerman & Gordon, 2000).
Also, as indicated, in our review of 200 cases of autistic spectrum disorders,
although many children shared a primary deficit (i.e., connecting affect to processing
capacities), their differences in the levels and strengths of developmental functioning
in other processing capacities or “component parts” tended to determine symptoms
and splinter skills, such as whether a child lined up toys (which requires some motor
planning) or just banged them, or scripted TV shows (which requires some auditory
memory) or was silent. It was also found that children with relatively stronger com-
ponent parts tended to make rapid progress once they were helped to connect affect
or intent to their other processing capacities. Children with weaker component parts
tended to make more gradual progress and required specific therapies, such as
speech and occupational therapy in an intensive manner, in order to improve the
component part directly, as well as work with the affect-component part connection.
These observations are consistent with recent neuroscience studies suggesting
that different processing capacities may compete for cortical access, depending on
functional use (Zimmerman & Gordon, 2000). They are also consistent with neuro-
psychological studies of individuals with autism, but without mental retardation, that
show that “within affected domains, impairments consistently involved the most
complex tasks dependent on higher-order abilities” (i.e., concept formation, complex

memory, complex language, and complex motor abilities) (Minshew & Goldstein,
1998). Higher-level capacities tend to depend more on “meanings” which, in turn,
depend on affective interactions with the world. Furthermore, these observations are
also consistent with work on the shifts to a more complex central nervous system
organization, including hemispheric connections that occur at the end of the first
year of life and early part of the second year, just as the ability to engage in affect-
mediated chains of social problem solving are on the ascendancy (Benson & Zaidel,
1985; Courchesne, et al., 1994; Dawson, Warrenburg, & Fuller, 1982; Sperry, 1985).
Interestingly, many children with autistic spectrum disorders use peripheral vision,
rather than central vision, to scan their environment (i.e., they don’t look directly at
caregivers but seem to look from the side). The neuroanatomy of the visual tracks is
such that peripheral vision only requires one hemisphere, the left or right one, to
function. Central vision, however, requires both hemispheres to function together
(because some of the pathways cross over and others do not). It would be reasonable
to explore the hypotheses that problems in integrating the two hemispheres, which
facilitates long chains of reciprocal, affective interaction and integrated central vision,
may contribute to autistic spectrum patterns. Work showing that the limbic system
and hippocampus is developing, including forming cortical connections, at around
11/2 years of age, is also consistent with these clinical observations on the increasingly
purposeful and meaningful use of actions and ideas in the second year of life
(Bauman, 2000).

Part II: Transformation of Affects in the First Three

Years of Life

Historically, emotions or affects have been viewed in a number of ways: as out-

lets for extreme passion, as physiologic reactions, as subjective states of feeling, as
interpersonal social cues (Young, 1943; Greenspan, 1997b). Beginning nearly six
decades ago, the importance of emotions for aspects of learning was documented
by psychoanalytic observers such as René Spitz and John Bowlby, who described
the effects of emotional deprivation, and Heinz Hartmann and David Rappaport,
who explored clinical and theoretical relationships (Spitz, 1945; Bowlby, 1952;
Hartmann, 1939; Rappaport, 1960). Sibylle Escalona (1968) and Lois Murphy
(1974) further explored affective development and described individual differences
in infants and their relationship to psychopathology. Building on this work,
Pediatrician T. Berry Brazelton systematized the observation of the infant’s social
and emotional repertoire (Brazelton & Cramer, 1990). In Intelligence and Adaptation
(1979) and other works (Greenspan, 1989, 1992; Greenspan & Greenspan, 1985;
Greenspan & Lourie, 1981; Greenspan & Wieder, 1998, 1999), Greenspan pre-
sented developmental observations and a model to integrate cognitive and affec-
tive aspects of the developing mind. In The Growth of the Mind (1997b), Greenspan
showed how emotions create, organize, and orchestrate many of the mind’s most
important functions, including intelligence and emotional health. He further

showed that intellect, academic abilities, sense of self, consciousness, and morality
have common origins in our earliest and ongoing emotional experiences and that
emotions are the architects of a vast array of cognitive operations throughout the
life span.
During the formative years there is a sensitive interaction between genetic pro-
clivities and environmental experience. Experience appears to adapt the infant’s
biology to his or her environment (Hofer, 1988, 1995; Rakic, Bourgeois, & Goldman-
Rakic, 1994; Greenough & Black, 1992; Weiler, Hawrylak, and Greenough, 1995;
Holloway, 1966; Turner & Greenough, 1983; Turner & Greenough, 1985; Thinus-
Blanc, 1981; Wiesel & Hubel, 1963; Singer, 1986; Hein & Diamond, 1983; Schanberg
& Field, 1987). In this process, however, not all experiences are the same. As
described in the prior section, children seem to require certain types of experiences
involving a series of specific types of emotional interactions geared to their particu-
lar developmental needs.
The difficulty in connecting affect to motor planning and symbols discussed in
the last section is only one part of a larger set of transformations of affect that
depend on specific types of emotional interactions. To more fully understand the
importance of affect in autism, and the development of intellectual and social skills,
it may prove useful to explore a number of affective transformations during the first
three years of life.
In the first year, affects become more complex. There is a transition from simple
affective states like hunger and arousal to, by 8 months, complex affect states like sur-
prise, fear and caution, joy and happiness, and enthusiasm and curiosity. As the child
progresses, affects become more differentiated. Eventually, affects organize recipro-
cal interactions and problem-solving. Then they become symbolized. Eventually, it
becomes possible to reflect on them. The transformations affects undergo can be
described in terms of six core early organizations that give the organism its desire to
act and underlie intelligence and emotional health (Greenspan, 1997b).
First, to attend to the outside world, and eventually to have joint attention or
shared attention, requires affective interest in the world outside one’s own body—in
sights, sounds, and movements. Obviously, parents who provide pleasurable sights
and sounds to a new baby will entice the baby into focusing on the world. Babies
who are visually hypersensitive or auditorially hypersensitive will need a different
type of enticement—a more soothing type—to take an interest in that outside world.
Babies who are underreactive will require more animated interactions. Consider
infants with low muscle tone who are underreactive and who grow up in institutional
care. They often lose weight and some don’t survive.
An early and continuing component of shared attention involves attending to the
world outside one’s own body with rhythmic, affectively-mediated perceptual motor
patterns. For example, in the early months of life, babies can be observed to move
their arms and legs in rhythm to their mother’s voices (Condon & Sander, 1974;
Condon, 1975). Soon they begin integrating what they hear and see (Spelke, 1976).
By four to five months, one can readily observe synchronous movement in
rhythm with mother’s affective communication via her voice, facial expressions, or

body movements. As development proceeds, reciprocal gestural, vocal, and verbal

communication generally occurs in an interactive rhythm. A consequence of this
may be the observation that it’s harder to remember or understand verbal phrases
presented in a monotone than in an affective rhythm.
The second functional developmental capacity is engagement. For an infant to
engage with a caregiver requires joy and pleasure in that relationship. When that’s
not present, children can withdraw and become self-absorbed. For children who
have processing problems, it may be much harder to pull them into that joyful rela-
tionship. But in clinical work, we have not encountered many children evidencing
autistic spectrum disorders who could not be pulled into greater degrees of relat-
ing through therapeutic work that works with their processing differences and rela-
tionships at the same time. Engagement and relating appears to be a very flexible
capacity. While language and certain cognitive functions may improve slowly for
some children, the capacity for warmth and relatedness seems to progress more
The third functional developmental capacity is two-way purposeful communica-
tion. Two-way communication and affective reciprocity, which Peter Tanguay and
Simon Baron-Cohen discuss (in this issue) obviously requires affect to provide the
“intent.” When an infant reaches for Daddy to take the rattle off his head or hand it
back to him, or gets into a back-and-forth smiling game, one clearly sees affect
(intent) guiding the interaction (i.e., the infant wants that rattle). Piaget thought
means-ends relationships occurred at nine months with motor behavior (i.e., an
infant reaching and pulling a string to ring a bell). But Piaget missed the role of affect.
The baby’s affective probe occurs much earlier than the motor probe. Causal affec-
tive behavior occurs earlier than causal large muscle motor behavior. First we see a
smile begetting a smile, a frown begetting a frown. Later on, we see the baby reach
for and give back objects. At this stage as well, the affect diathesis is occurring, now
transforming relating into two-way, affective communication (rather than just joyful
interest in the caregiver).
The fourth level of transformation occurs between 10 months and 18 months. It
involves the development of a range of new capacities, all related to the toddler’s
ability to engage in longer sequences of affective reciprocal interactions with clear
intent or problem-solving goals and the ability to perceive and interact in these larger
patterns. This transformation enables the toddler to form a more integrated sense of
self, integrate affective polarities, social problem-solve, and broaden visual-spatial
and auditory processing abilities.
As indicated earlier, during this stage we often start seeing differences with some
of the children who will be diagnosed with autistic spectrum disorders. Many of the
children have relative degrees of mastery of the first three stages, but the fourth stage
is more difficult. Now, the infant has to connect his affect to his motor planning and
sequencing, as well as his emerging symbol formation. In order to be able to take a
parent by the hand and walk him to the toy area and point to the toy that he wants,
the infant has to have both a constant inner signal of intent or affect and a motor plan
which is connected to it. If either one of those are problematic the infant may not be

able to implement complex, problem-solving, social behavior. As a child begins to

imitate words (“Mom,” “Dad,” “go,” “door”), those symbols or words have to be
invested with affect to have meaning. The word “juice” only has meaning to the
degree the child can invest the words with the many affective experiences that come
to mean “juice,” including the pleasure and image of drinking the juice.
Even grammar, which Chomsky and other linguists have assumed is innate,
depends on affect and affective interactions to become functional. Children with
autistic spectrum disorders frequently verbalize nouns in a repetitious way (“Door,
door.”). If the intervention can get them affectively interactive, however, they can
often learn to use proper grammar. For example, a child is opening and closing a
door. We get stuck behind the door. If they push us away, they are becoming pur-
poseful. Purposeful, affectively-mediated behavior creates a foundation for the pur-
poseful and meaningful use of words. Soon the same child is saying “go” while
pushing us away. We may then say, “Where go? Where go?” We might further say,
“Should we go away or stay? Away or stay?” The child may say, “Go away, go away.”
Now the child is using correct grammar. Noam Chomsky and his colleagues
(Chomsky, 1966, 1980; Pinker, 1994) were mistaken when they thought that gram-
mar was largely innate and that only experience in a very general sense was required
to turn on the language switch. Grammar requires very specific types of affective
experience. Affective reciprocity is needed to create purposeful action and then
related purposeful symbols or words. The affect, by providing intent, enables the
components of language to align (e.g., “open door” versus “door, door, door.”). Many
investigators may have missed the importance of affective reciprocity because it
occurs routinely with most infants and toddlers and their caregivers.
At the fourth level of transformation many complex social problem-solving inter-
actions, in addition to involving vocalizations and words, involve visual-spatial pat-
terns. Johnny can find his mother in the other room because between 12 months and
18 months, he has learned to construct a visual-spatial road map of the house. This
road map allows him to do what Margaret Mahler described as “separation-individ-
uation” (Mahler, Pine, & Bergman, 1975). He explores independently, but then
comes back to his base of security for “refueling.” Mahler, however, did not realize,
he could use his visual-spatial processing and auditory processing to refuel from afar.
He can have his security blanket from across the room, either by looking at Mom
and communicating with looks or other gestures or by hearing her voice from
another room and figuring out where she is based on her voice. If his visual-spatial
or auditory processing is weak, or the affect system is not investing his visual-spatial
or auditory processing, however, this emerging capacity will not properly form.
Simon Baron-Cohen (in this issue) describes the folking of cognition and divides
children into the physicists (interested in the mechanics of things) and the novelist
group (interested in people and emotional interactions) in terms of relative cognitive
and social strengths and weaknesses. What happens, however, when a child has
strong visual-spatial and mechanical skills and invests that system with a lot of pleas-
urable affective experience with primary caregivers? Do they become both a physi-
cist and a poet (i.e., a physicist with a warm heart, big smile, and feel for people)?

One possibility is yes. We have observed that children who are able to use strong
visual-spatial systems in reciprocal affective interactions often use physical
metaphors to deal with emotions. They’re not the kind of people who are afraid of
affect or feeling. Quite to the contrary, they talk about their feelings, but often use
physical images to communicate (“I feel as angry as a four megaton nuclear blast!”).
The affect diathesis hypothesis suggests that affect invests not simply the capac-
ity for complex interactions to give meaning to sounds, words, and behaviors, but
also invests processing capacities, such as motor planning and visual spatial process-
ing. It gives meaning and functional range to these capacities. Individuals can then
use different processing capacities in a variety of ways—not just to solve an equation,
but also for social, personal problem-solving (e.g., figuring out how to work the
crowd politically).
It may not appear obvious how to affectively interact through the visual-spatial
system. But, playing chase games, hide-and-go-seek games, and treasure hunts are
good vehicles for bringing visual-spatial capacities into an affective context, as are
discussions involving how angry or loving we feel on a 1–10 scale. This is exactly
what we do with some children who need more work “affecting” their visual-spatial
and motor planning systems. The stage for these types of games are often set earlier
in the third and early fourth stages, with peek-a-boo and other types of play that com-
bine pleasurable affect, affective expectation, and visual-spatial processing.
In this fourth stage, the child is also beginning to integrate affective polarities.
Early on, infants tend to have extreme affect states—all happy or gleeful or all sad—
but by 18 to 19 months we see children begin to shift affect states more readily and
actually integrate affect states such as happiness and sadness, anger and closeness.
They can be angry and then seem to want forgiveness and make up. When playing
with a 13-month-old child, it feels like if he were angry and had a gun, he very well
might pull the trigger. With the 18-month-old, it feels like he integrates his caring and
anger. He might look mad and feel connected and warm at the same time. One can
often feel the quality of these affect states when playing with infants and toddlers at
different ages.
Complex affective patterns emerging during this fourth transformation lead to
patterns of affective expectations. A toddler purposefully behaves cooperatively,
hams it up and acts funny, or becomes mischievous, all in relation to specific inter-
nal states. A child also anticipates what Daddy’s going to do next when he comes
home. He looks at his face and if it’s angry, he hides behind the door. If he looks like
he’s feeling warm and will be nice, the toddler will run up and flirt with him. The
toddler can anticipate the “other” as well as the “self” in terms of affects and behav-
iors. This capacity sets the stage for many of the “theory of mind” tasks that Simon
Baron-Cohen describes (this issue).
At the fifth level, transformations involve the affect system investing ideas. For
example, in pretend play, affects or desires drive the theme (dolls hugging or kissing)
as well as functional language (“I’m hungry,” “I’m angry,” “Give me that.” “Look! I
want to show you something.”). Functional language, whether it’s on a need basis
(Give me juice.”), or at a collaborative “show you this or that,” or sharing opinions

“I didn’t like that” basis, is very different from simply labeling objects or pictures.
Here is also where IQ tests fall down. IQ tests don’t discriminate well enough
between the different uses of ideas and language, such as between pragmatic lan-
guage or creative and abstract thinking and simply using language to label objects or
for rote, memory-based problems.
Many researchers use IQ tests to create comparable groups. For example, chil-
dren with autism and without autism are matched on IQ tests and then given theory
of mind tasks or other cognitive tasks. The goal is to see if the children with and with-
out autism differ on theory of mind tasks when they have similar IQs. If they differ,
it would suggest that theory of mind tasks distinguish autistic from non-autistic chil-
dren. But, there is a problem with this approach. Even though the groups have sim-
ilar IQs, they may not be the same in terms of their intelligence. Their intelligence
patterns may be quite different. For example, they may have very different pragmatic
language levels and very different abstract reasoning skills. In these studies, we must
control for pragmatic language and abstract thinking (i.e., true intelligence), not IQ,
if we want to get at the essence root of how certain cognitive capacities, such as those
involved in theory of mind tasks, are related to autistic spectrum disorders. In other
words, theory of mind capacities may relate more to the level of pragmatic language
and abstract thinking than to whether a child does or does not have autism. Only
proper control groups can tease out this distinction.
Research on theory of mind tasks and other cognitive or perceptual capacities,
such as the ability to discriminate facial expressions to be cognitively rather than
affectively oriented. For example, asking a child to figure out how another child is
feeling in the middle of a power struggle, in heated debate, or in situations that
might lead to disappointment or sadness, would get at his ability to understand
someone else’s true emotions in an emotional context. Figuring out what someone
else will see in a room, on the other hand, or identifying pictures of facial expres-
sions are more cognitive and perceptual tasks. These areas of research have led to
cognitively oriented interventions. These interventions do not work enough at the
level of affective interchanges. The importance of engaging in long chains of recip-
rocal affect cueing in order to establish a sense of self and a sense of other (through
these back-and-forth affect signals) is often overlooked in many of these interven-
tions. The cognitive procedures utilized in such interventions are many steps
removed from the affective interchanges that are necessary to establish the com-
promised capacities.
For example, there has been a great deal of interest in children with autistic spec-
trum disorder evidencing difficulty in discriminating facial expressions of different
emotions. Based on this research, children who have trouble interpreting the emo-
tional expressions of others are “taught” about emotional expressions by looking at
pictures of people with different facial expressions or through identifying emotional
expressions of others in structured exercises. This conscious cognitive appreciation
of a picture is, however, not what’s missing. What’s missing is the intuitive, almost
automatic sense of another person’s affect. This is the capacity one uses in appreci-
ating a friend’s subtle, emotional state or in working the crowd at a cocktail party.

In other words, the understanding of the other person’s emotions is experienced very
rapidly through a personal, visceral, emotional reaction. In fact, we can often
respond to the person’s affect before it even consciously registers. Thus, we flirt back,
look puzzled or grimace in anger as part of our intuitive, affective response. Once we
have experienced, at the intuitive level, the other person’s emotional signal, we can
also reflect on it in a conscious and deliberate manner. We may say to ourselves,
“they look sad” or happy or angry. In making these determinations, however, we are
relying on our own affective response, not simply on the other person’s facial expres-
sion. Also, as indicated, in the ordinary course of events, such as working the crowd
at a cocktail party or negotiating peer relationships on the playground, there are
many affect signals being exchanged in a brief period of time. If a child or adult con-
sciously tries to figure out each separate one, they will be doomed to failure and con-
fusion. Therefore, the only way to help a child with problems reading affect signals
is to provide him or her extra practice in experiencing and reading those signals (i.e.,
in social situations involving lots of reciprocal, affective interactions, initially with
one-on-one caregiver, child, and peer play and gradually in more complex situa-
tions). The “practice” needs to involve the personal inner experiences of someone
else’s affect, as well as one’s own, in a series of reciprocal interactions. Similarly, chil-
dren who have theory of mind problems are often provided with cognitive exercises
involving figuring out other people’s perspectives, rather than working at the primary
level of affective signaling, which is often compromised and at the core of these chil-
dren’s problems.
Some may believe that children with autism or Asperger’s Syndrome are not able
to learn to feel their own and someone else’s affect and, therefore, can only learn to
read facial expressions through pictures or perform theory of mind tasks in a con-
scious, deliberate manner. We have found this assumption not to be correct. With a
program focusing on relating and affect cueing, the majority of children made
progress in this capacity (Greenspan & Wieder, 1997).
In general, the missing piece in many intervention programs is a lack of under-
standing of the developmental steps involved in acquiring certain cognitive, social,
and emotional skills. By understanding these steps, which often involve transforma-
tions of affect, intervention strategies can help the child master the critical founda-
tions for cognitive and social skills.
At the sixth level of transformation, a child builds bridges between affectively
meaningful ideas. Establishing reality-testing, a symbolic sense of self, and moving
back and forth between fantasy to reality depends on reaching this next level. For
example, critical to establishing reality-testing (which is the basis for later abstract
thinking) is an affective “me” intending to do something with an affective “other.”
There has to be an interaction involving affect between the “me” and the “other” to
establish a psychological boundary (i.e., an affective sense of what’s “me” and an
affective sense of what’s “outside me”). That boundary doesn’t come out of reading
books or out of doing puzzles. It comes from interactions involving the exchange of
affective gestures and symbols. It comes out of interactions such as “I want this.” “No,
you can’t have it,” or “Yes, you can.” In addition, these interactions must be part of

a continuous flow of back-and-forth affective gestures. Islands of affective interactions

followed by self absorption leads to an “in and out” affective probe or rhythm with
the external world (reality). A stable sense of reality requires a continuous interactive
relationship to the significant “others” in our lives. Abstract and inferential thinking
grows from a solid reality boundary.
A stable reality boundary also allows empathy to develop. A child can project a
“me” into a “you” and figure out how “you’re” feeling to the degree that a child has
established a separate sense of “me.”
In this context, abstract thinking, empathy and theory of mind tasks are exten-
sions of the stage of building bridges between affective ideas. In addition to affective
interactions, the child’s individual differences in processing capacities (i.e., visual-
spatial processing, motor planning, auditory processing) will contribute to these
advanced mental capacities and they have to be taken into account in considering
these capacities.
For example, consider the theory of mind task Simon Baron-Cohen (this issue)
has described where the child is looking into the basket and the other child has
to describe what he’s doing. In a task like that, if we show one child a diagram,
he’ll get it and tell you exactly what the other child is doing. But if we describe
it in words to that child (not visually, but auditorially). “Gee, a child is looking in,
what is he seeing?” The child won’t get it. Another child, however (also with an
autistic spectrum disorder), may be just the opposite—stronger with the verbal
than the visual. The processing capacity is, therefore, a very important component
of a child’s ability for, and pattern of, abstract thinking and empathy. Also, if a
child has sequencing problems and there are five steps in the problem, the child
may get lost simply because of the sequencing challenge, not because of an inabil-
ity to project himself into someone else’s shoes. Therefore, higher levels of
abstract thinking, including theory of mind tasks, may occur in certain processing
modalities and not other modalities and with regard to certain affect realms and
not others.
In summary, we’ve described the affect transformation that occurs in each of
the six functional developmental capacities. Affect is responsible for helping
the child go from simple interest in the world all the way up through social prob-
lem-solving (and procedural knowledge). It enables a child to progress through
procedural knowledge up to symbolic knowledge. It gives meaning to what the
child hears and how he processes visual-spatial information and sequences motor

The importance of genes is often cited to minimize the role of affective interactions and the environ-
ment, but the role of genes is more complex than is often acknowledged. Consider a very interesting study
on schizophrenia. Investigators looked at identical twins that shared a placenta and identical twins that
did not share a placenta. The concordance rates of the identical twins that shared a placenta was very
high in keeping with the literature on the genetic basis of schizophrenia. The concordance rates of iden-
tical twins that did not share a placenta were very low close to the rates for dizygotic twins (Davis, Phelps,
& Bracha, 1995). Therefore, unless we control for the sharing of a placenta, we won’t know how much
the genes contribute and how much the intrauterine environment contributes. This article was published
in Schizophrenia Bulletin, but has not been sufficiently used in behavioral genetics research.

Distinguishing the Capacity to Engage from the Capacity

for Exchanging Affective Signals

In discussing the different stages of development, it’s important to emphasize the

difference in the level of affective transformation involved in patterns of simple engage-
ment and relating versus reciprocal affective signaling. Many children with autistic
spectrum disorder are capable of deeply engaging and forming patterns of warmth,
trust, and dependency with a great deal of pleasure and joy. Some of these children will
have varying degrees of difficulty, however, in developing ongoing, reciprocal, affec-
tive interchanges. Even if they are warmly and deeply engaged, it’s especially difficult
for some children to develop a continuous flow of reciprocal, affective interactions.
There are many reasons for this difficulty, including biologically based process-
ing difficulties involving motor planning, visual-spatial processing, auditory process-
ing, or sensory modulation. Motor planning problems, for example, make it hard for
the child to sequence and, therefore, engage in a multi-step, affective interactions.
Visual-spatial processing difficulties make it hard for the child to construct larger spa-
tial patterns and, therefore, picture and negotiate a multi-step, affective sequence
leading to a goal. Sensory reactivity difficulties will also make it hard for children to
participate in long chains of reciprocal, affective interactions. They become over-
whelmed with catastrophic affects, with short bursts of intense reactions rather than
modulated, long chains of interaction. The distinction between the ability to
engage and the ability to engage in long chains of affective, reciprocal interactions is
especially important for children with autistic spectrum disorders. We hypothesized
that children with autistic spectrum disorders have a biologically based deficit in the
capacity to connect affect to motor planning and sequencing and, therefore, are
unable to enter into long chains of reciprocal, affective interaction. They also often
have motor planning, visual-spatial, language processing, and sensory reactivity
problems further intensifying this basic difficulty.
Yet, many children with autistic spectrum disorders have, or are capable of rela-
tively quickly forming, patterns of engagement that involve a few circles of back-and-
forth affective interchange. They are capable of engaging with pleasure, warmth, and
joy. They’re, therefore, capable of the earlier levels of affect transformation, involv-
ing basic engagement, even though they have difficulty with forming reciprocal affec-
tive interchanges. The basic capacity to love, experience intimacy and deep
dependency is relatively stronger than generally acknowledged. It is the capacity for
reciprocal, affective gesturing (i.e., the ability to negotiate within a loving relation-
ship) that’s more problematic.
Many of the children we’ve worked with start off with the capacity for deep pleas-
ure. They enjoy cuddling, being held, and show joyful smiles when their caregivers
engage them in a warm pattern of relating. Other children appear to be more
avoidant, self-absorbed, or affectively constricted in terms of showing joy and pleas-
ure. Often these children can be helped to enjoy fundamental relating in a deep and
satisfying manner, once we figure out their sensory processing and motor profiles.
For example, some of the children are very sensory over reactive and, therefore,

uncomfortable with closeness involving touch or high- or low-pitched sounds. When

the sensory environment is tailored to their unique profiles, these children begin evi-
dencing enormous pleasure in relating. Children who have difficulty with motor
planning become more deeply engaged when their caregivers learn how to position
themselves in a way that makes the motor planning challenge simpler. In other
words, the child who only has to reach out and hug or is helped to be in pleasur-
able sensory contact so he can feel where his caregiver is, finds it much easier to
reach his goal of closeness with mother and father than the child who has to sustain
focus and interest through four or five independent actions in order to reach mother
or father.
In our review of 200 cases (Greenspan & Wieder, 1997) of children who received
a comprehensive, intensive program of intervention, we found that the first gain the
vast majority of children made was in the capacity to engage with warmth and pleas-
ure. The capacity for engaging with warmth and pleasure occurred before gains in
language skills, cognitive skills, or motor skills. Engaging appeared to be the aspect
of functioning that was the most quickly responsive to the intervention efforts.
The observations that the capacity to engage responds most quickly and that
many children diagnosed with autistic spectrum disorders can show a lot of joy and
pleasure even before an intervention program has begun suggests a possible mis-
conception about autistic disorders. Often, the capacities for affective reciprocity and
affective engagement are believed to be part of the same process and children with
autism are, therefore, viewed as less capable of love and intimacy than others. But,
affective interactions are fundamentally different from the capacity to form a basic
relationship characterized by pleasure and joy and, ultimately, a sense of trust and
Many of the children we’ve been following for over five years (some for over 10
years) evidence patterns of closeness and warmth that are both joyful and deeper, in
many respects, than typical children their age are able to show. We have observed
that many evidence a great deal of warmth and closeness for their age group. It’s not
surprising that these children can evidence intimacy because their parents have been
spending extraordinary amounts of time with them in a warm and nourishing way
that is sensitive to their processing profiles. Children without challenges whose par-
ents are very available also form deep, satisfying patterns of closeness, warmth, and
As indicated, historically, however, it has often been thought that children with
autistic spectrum disorders have a biologically-based deficit in their ability for
warmth and closeness (i.e., the capacity for a deep sense of love). It’s been thought
that this is part of their difficulty with forming patterns of social interaction. It’s been
believed that they experience the type “autistic aloneness” that Kanner described in
his classic descriptions (Kanner, 1943). Even though many studies have refined
Kanner’s original observations and different degrees of social relating have been
described and incorporated into diagnostic criteria for children with autistic spectrum
disorders, nonetheless, the perception persists (and continues to a partial degree even
in the most recent diagnostic criteria) that children with autistic spectrum disorders

are less able to engage with depth and warmth and with a deep, abiding sense of love
than other children who don’t have these developmental challenges.
Our clinical observations would suggest that this common belief is simply incor-
rect. If we distinguish a child’s capacity for deep, joyful relating from the capacity for
affective, reciprocal interchanges, it is possible to observe that many children with
autistic spectrum disorders are capable of the full range of warmth, love, and close-
ness. This intimacy is relatively easy to observe in families who focus a great deal on
promoting relaxed intimacy by observing spontaneous relating for long periods of
time (hours, not minutes) and attending to all the subtle ways the children have of
showing their intimacy. In our review of 200 cases, over half the children evidenced
a deep rich capacity for intimacy and over 90% showed a continuing growth in this
pattern (Greenspan & Wieder, 1997).
One may expectedly raise the question: why do many children with autistic spec-
trum disorders appear to spend so much time avoiding relationships or with con-
stricted affect or in states of self-absorption? Why don’t many show a great deal of
obvious pleasure and joy in relating to others? The answer to these two important
questions is that children with autistic spectrum disorders who are capable of enor-
mous warmth, joy, and deep relating can withdraw from relationships if the rela-
tionships are experienced as aversive or painful or simply not pleasurable. This often
happens, not because the children are incapable experiencing joy or because the par-
ents are not extraordinarily loving,but because caregivers are not sufficiently helped
to figure out the unique sensory processing profile of the child so that they offer pat-
terns of relating in a way that is pleasurable and deeply satisfying.
Some of the children have sensory processing patterns that are not so challeng-
ing and their caregivers find it relatively easy to approach and entice them into
warm, nurturing patterns of relating. Other children, however, have complicated sen-
sory patterns where caregivers require assistance. If this assistance is not forthcom-
ing, children can pull away from the very relationships they might otherwise enjoy
and seek. In addition, if the children are not helped to progress into reciprocal affec-
tive interactions, it’s hard for them to negotiate intimacy and, as indicated, they are
more likely to experience intense, catastrophic affects.
This distinction between the child’s capacity to engage and the child’s capacity
for affective reciprocal interchanges is being emphasized because it clarifies the mis-
perception about children with autistic spectrum disorders mentioned above, that is,
that this group of children somehow loves less deeply or less profoundly than other
children. This misperception, as is well known, can easily become the basis for a self-
fulfilling prophecy if caregivers are discouraged from trying to find ways to draw
their children into deeper patterns of intimacy. Simply not encouraging caregivers to
explore this capacity can be a way of it becoming a prophecy come true.
There are, in fact, two misperceptions that need to be avoided. One is that the
children who have a more challenging time learning to relate because of their bio-
logical differences, (such as sensitivities to sound or touch or language problems), are
unable to learn to relate warmly through the availability of special interactions geared
to their unique developmental profiles. The other misperception, equally worrisome,

is that caregivers who construct special patterns of care that woo their child into rela-
tionships are somehow contributing to the causes of their children’s problems. What
we’re saying here is that many children, especially those with autistic spectrum dis-
orders, have biological differences that express themselves in the way the child
processes sensations and organizes and plans responses. These biologically based
processing differences can make the expectable milestones of learning to relate and
communicate very challenging and, in some cases, possibly even impossible given
current knowledge. The caregiving environment, however, while therefore not the
cause of the child’s biological challenges can be a vital part of the process that helps
the child master, to varying degrees, aspects of his or her biological challenges. We
have found that the caregiving environment can be especially vital in helping chil-
dren with autistic spectrum disorders learn to engage with greater degrees of warmth
and intimacy if their caregiving overtures are tailored to the child’s developmental
profile. Therefore, while not being the cause of the problem, the caregiving environ-
ment can be an important component of a comprehensive approach to intervention.
Perhaps the best way to conceptualize the challenges to forming relationships for chil-
dren with autistic spectrum disorders is as follows. Due to their unique sensory pro-
cessing profiles, the negotiation of a deep sense of intimacy is a complex and subtle
process. The capacity appears to be there, but needs to be met with caregiving over-
tures that are sensitive to the child’s unique processing patterns. It is very easy for chil-
dren with challenging processing profiles to respond negatively to even simple
environmental challenges. Children without these processing challenges might easily
have a more flexible capacity to engage others even under difficult conditions. The
children with complex processing profiles are, therefore, extremely sensitive to the
subtleties in their environments and can easily regress or form patterns of avoidance
and self-absorption. At the same time, however, they can be drawn into wonderfully,
deep, and satisfying patterns of closeness and intimacy. Parents are in a unique posi-
tion because of their long-term relationship to the child to foster this intimacy and
engage in long sequences of pleasurable affective interactions. Children’s capacity for
intimacy, coupled with their sensitivities, in fact, suggest a need for nurturing, inti-
mate care that is especially deep, flexible, and persistent.

Part III: The Role of Emotions in the Development of

Intelligence and Social Skills

If each of the transformations just described occurs, affects give rise to higher
and higher levels of intelligence and emotional health. The emotional interactions
described in the prior section, therefore, are not simply responsible for early social,
cognitive, and language capacities, but for higher level intellectual and social capa-
cities as well. A continuous flow of emotional interactions between children and
caregivers is especially important for the development of the highest levels of
human thinking involving self-reflection, making inferences, creating new prob-
lem-solving strategies, having empathy and insight, and regulating mood and

behavior. This section explores the connection between affect and the capacity to
regulate behaviors and construct the highest levels of intelligence and social inter-
actions. In doing so, it shows why interference in the development of affective
interactions, as seen in autism, is associated with a large range of cognitive, behav-
ioral, and social deficits.
Our awareness of the importance of affect for higher level cognitive and social
skills was heightened when we observed and talked with children with strong self-
awareness and reflective thinking skills, and realized that most of them also showed
positive self-esteem, demonstrated a capacity for moral judgment, were analytical in
their reasoning, and did well in school and with their peers. We sought to understand
what helped them become this way, and, therefore, spoke with them and with other
children who had opposite personal characteristics. We learned that what we com-
monly label as intelligence, social skills and morality was based on the child’s ability
to use his affects to think, which, in turn, is supported by the six types of emotional
interactions, described in the prior section, that are negotiated during the early years
of life.
For example, when we asked a group of eight-year-olds abstract questions, such
as what they thought about justice or fairness, their comments were very revealing.
Some of the children responded with a rote listing of people who behaved “fairly,”
such as a particular parent or teacher or television character. However, others gave
far more reflective answers, along the lines of “Well, when I hit my brother after he
hit me it was unfair for me to be punished, but when I hit him first it was fair for me
to get a punishment. If I bump into him by accident, it’s not fair to be punished, but
if I do it on purpose it is fair.”
Not surprisingly, when we looked at the two groups of children more closely,
those who gave us the rote list tended to be the ones who were experiencing more
problems in their relationships and in their schoolwork. The children who gave us
more creative and reflective responses tended to do better in these social and intel-
lectual areas.
We then took a second look at the more reflective responses and discovered that
they had two components. This was true whether our test question focused on fair-
ness or any other abstract quality, such as honesty, friendship or freedom. The first
component was that the children’s responses always started off with a personal anec-
dote, an account of lived emotional experience. The second component was that the chil-
dren put these experiences with abstract concepts into some sort of analytic
framework and context.
When we later asked this same question of adolescents, they were able to list
more categories (five different types of fairness, for example) and supplied an even
more worldly-wise analytical framework. But in every instance, and at every age, two
components were evident in the more sophisticated replies: lived emotional experi-
ences and a framework, or context. The children who didn’t have a lot of lived emo-
tional experiences—due to either nurturing or biological challenges that interfered
with interaction, such as difficulties with language, for example—tended to be the
ones who responded with concrete lists, rather than anecdotes. Interestingly, we

found that the most effective way to teach such children analytic reasoning or think-
ing was by creating opportunities to have more “lived emotional experiences” and to
reason about them. We observed that even children with severe developmental prob-
lems, including autistic patterns, could become more creative and reflective when
they were exposed to more one-on-one interactions with their caregivers.
It thus became increasingly clear that certain types of emotional experiences
seemed to be necessary for generating abstract ideas or thinking. This concept
had been overlooked, because for many hundreds of years it has been a given in
Western culture that emotions are separate from intelligence. We have commonly
assumed that emotions are experienced as bodily reactions or passions that lead us
to do irrational things; and more recently we have also viewed them as the cues that
enable us to function socially. But in these conceptions of emotions, they are viewed
as fundamentally different from intelligence, which has been considered to be the
logical part of our minds that helps us be rational and make sense of the world.
Our new observations suggest that emotional interactions play a far more critical
role in intellectual functioning. They can help us go beyond Howard Gardner’s
important idea of separate, multiple intelligences (1983), or Antonio Damasio’s
research on the brain which suggests that emotions are important for judgment but
somehow separate from academic capacities or overall intelligence (1994). Even Jean
Piaget, the pioneering cognitive psychologist, overlooked this vital connection.
Piaget observed that when an eight-month-old is accustomed to pulling a string that
is attached to a bell and to hearing the bell ring in response, he will eventually stop
pulling the string if it is detached from the bell. To Piaget, this sort of behavior
revealed that the child is a causal thinker, because he pulls the string only if it leads
to his hearing the sound of the ringing bell. Although Piaget’s observations were
accurate, he did not realize that this was not the child’s first opportunity to learn
about causality. A baby’s first lesson in causality occurs many months earlier, when
he pulls on his mother’s or father’s heartstrings with a smile that brings a responsive
smile of delight or some other joyful expression to his parent’s face. The child then
applies that emotional lesson to the physical world of pulling strings, banging objects
and the like.
At each succeeding stage of development, we have found that emotional interac-
tions like a little baby’s smile leading to a hug enable the child to understand how the
world works, and eventually to think, solve problems and master academic chal-
lenges. Emotions are actually the internal architects, conductors, or organizers of our
minds. They tell us how and what to think, what to say and when to say it and what
to do. We “know” things through our emotional interactions and then apply that
knowledge to the cognitive world.
Consider how a young child first learns how to say “Hi!” as he greets other people.
A toddler doesn’t memorize lists of appropriate people to say hello to. He merely
connects the greeting with a warm friendly feeling in his gut that leads him to reach
out to other people’s welcoming faces with a verbalized “Hi!” If he looks at them and
has a different emotional feeling inside, one of wariness, he’s more likely to turn his
head or hide behind your legs.

It is his feeling, or affect, that triggers the child’s decision over whether or not to
greet a stranger. We encourage this kind of “discrimination” because we don’t want
our children to say “Hi!” to a menacing stranger in a back alley. We want them to say
hello to nice people like Grandma. If the child won’t say “Hi” to his grandmother,
it’s because he’s not experiencing a warm feeling inside. If we can teach Grandma
how to evoke that warm, fuzzy feeling inside her grandchild, she’ll eventually get a
friendly “Hi!” back in response. Similarly, if a child learns to greet those people who
make him feel warm inside, he will quickly say “Hi!” to a friendly teacher or to a new
playmate. He carries his emotions inside him, helping him to generalize from known
situations to new ones, as well as to discriminate, or decide when and what to say.
Not only thinking grows out of early emotional interactions; so does a moral
sense of right and wrong. The ability to understand another person’s feelings and care
about how he or she feels can only arise out of a series of nurturing interactions. We
can only feel empathy if someone has been empathetic and caring with us or else we
would not know what the feeling of empathy felt like. Even something as purely aca-
demic and cognitive as math and concepts of quantity is based on early emotional
experiences. “A lot” to a three-year-old is more than he wants; “a little” is less than
he expects. Later on, numbers can systematize this feel for quantity. Children with
math blocks can be helped to become solid students by going back to the early emo-
tional roots of learning about quantity. Similarly, concepts of time and space are
learned by the emotional experiences of waiting for Mom, or of looking for her and
finding her in another room.
Words also derive their meaning from emotional interactions, as our example of
fairness illustrates. A word like “justice” acquires content and meaning with each new
emotional experience of fairness and unfairness. Comprehension of a word like
“apple” is based on numerous emotional experiences involved in eating one, throw-
ing one and giving one to your teacher, in addition to its obvious physical character-
istics of redness and roundness. As we showed earlier, even our use of grammar,
which the noted linguist Noam Chomsky and others believe is largely innate and
only needs some very general types of social stimulation to get going, is based in part
on very specific early emotional interactions.
To provide the optimal quality and quantity of the types of emotional interactions
requires, however, not simply caring loving adults, but caring loving adults who are
consistent in the child’s life and who have time to spend with the child each day.
Ongoing empathetic caring is especially important for qualities of the human mind
that can only be learned through ongoing experience—compassion through compas-
sion, and intimacy by experiencing intimacy. These capacities are different from
motor, language or cognitive skills. They require consistent care giving of one or a
few stable caregivers who are there for years in the child’s life. For example, children
can learn altruistic behaviors in an impersonal way by copying or learning rules (i.e.,
to do “the right thing”), but truly caring for another human being only comes through
experiencing that feeling of being cared for oneself. It is a self-evident truth that to
feel an emotion we have to experience that emotion in an ongoing relationship.
We can’t experience emotions that we have never had, and we can’t experience the

consistency and intimacy of ongoing love unless we’ve have had that experience with
someone in our lives. For some it may be a grandmother or an aunt, or it may even
be a neighbor, but it must be there. There are no shortcuts to life’s most important
Therefore, family patterns that foster healthy relationships are essential for
healthy emotional and intellectual growth. In one study we found that families with
four or more risk factors interfering with relationships were 20 times more likely to
have marginal IQ scores and behavior problems at age four. This pattern continued
and was again documented when the children were 13 years old (Sameroff, Seifer,
Barocas, Zax, & Greenspan, 1986; Sameroff, Seifer, Baldwin, & Baldwin, 1993).

Emotional Interactions and the Regulation of Behavior,

Feelings, Moods, Problem-Solving, and Thinking

The basic feature of caring, caregiver-infant relationships, back-and-forth affec-

tive interactions within which we read and respond to the baby’s signals and the baby
reads and responds to our affective overtures, is responsible for a surprisingly large
number of vital mental capacities.
For example, the child eyes the red rattle and parent holds it up and the child
reaches for it, takes it, the parent holds his hand out and the child puts it back in the
parents hand, there is a big smile, the parent smiles back, the child makes some
sounds, the parent makes some sounds back. We have an interactive dialogue
involving sounds and gestures and emotional expressions, such as big smiles, all
happening in a rapid back-and-forth set of exchanges. As described earlier, this
rapid back-and-forth set of exchanges (i.e., “opening and closing circles of commu-
nication” or “reciprocal affective interactions”) is the beginning of learning to think
purposefully or causally, “I can make something happen.” It also teaches babies
how to take initiative (you do something and it makes something happen, your smile
gets a smile from mommy or daddy). The infant is beginning to have a sense of
purpose and will and, very importantly, a sense of “self” (it’s “me” making some-
thing happen, “me” getting that smile or getting that little red rattle with reach-
ing out “my” hand). The sense of self, will, purpose, assertiveness and the
beginning of causal logical thinking all occur through long chains of reciprocal
affective interactions.
But something else also occurs as a product of these back-and-forth interactions
that has not been sufficiently described before. Through these reciprocal affective
interactions the child learns to control or modulate his behavior and his feelings. We
all want children who are well regulated or modulated, that is, who can be active and
explorative some of the time, concentrate and be thoughtful and cautious other
times, joyful yet other times. We all want children who can regulate their emotions
in a way that is appropriate to the situation and also regulate their behavior in a way
that is appropriate to the situation. We all admire adults who are able to do this. But
how do we learn to control or regulate our behavior and moods?

What is the difference between a child who can regulate their mood, emotions
and behaviors and a child who can’t and where the slightest frustration feels cata-
strophic, as though the world were going to end, or where anger is enormous and
explosive, or sadness seems like it will go on for ever and ever? What is the differ-
ence between these extreme and almost catastrophic emotional reactions and the
ones that are finely regulated and responsive to the situation at hand? (Sometimes
there is sadness, sometimes joy, sometimes even anger.)
The difference is in the degree to which the child learns and masters this impor-
tant capacity for rapid back-and-forth exchanges of emotions and gestures through
these reciprocal interactions.
When a child is capable of rapid back-and-forth interactions with his care-
giver, he is able to negotiate, in a sense, how he feels. If he is annoyed, he can make
a look of annoyance or a sound or hand gesture. Mother may come back with a
gesture indicating “I understand” or “OK, I’ll get the food more quickly,” or indicat-
ing “Can’t you wait just one more minute?” Whatever the response is, if it is respon-
sive to his signal he is getting some immediate feedback that can modulate his
response. The anger may be modulated by the notion that mother is going to do
something, even if she can’t do it immediately. Just the sound of her voice signals to
him that she is getting that milk bottle ready and it’s coming real soon. Or his push-
ing the bottle away with an angry glance at her and her putting her hand out to take
it from him is another way he knows that his anger is being responded to. Similarly,
when his big smile gets a smile back he knows his joy is being responded to. He gets
a sense that he can regulate his emotions through regulating the responses he gets
from environments. He gets the quick sense that he and his caregivers are in this pat-
tern where they are regulating one another, where the back and forth is a finely-
tuned nuance system involving lots of feedback.
We now have a fine-tuned system rather than a global or extreme one. The child
doesn’t have to have an extreme tantrum to register his annoyance, he can do it with
just a little glance and a little annoyed look. Even if mother doesn’t agree with him
or can’t bring that food right away, nonetheless she is signaling something back
which gives him something to chew on while he is deciding whether to escalate up
to an even more annoyed response if she doesn’t move fast enough. This is a system
of steps. Even if he does escalate up to a real tantrum, he is not going from 0 to 60
in one second. More often than not, however, his mood and behavior will be regu-
lated as a part of this back-and-forth interaction. All the different feelings, from joy
and happiness to sadness to anger to assertiveness become a part of fine-tuned regu-
lated interactions where a subtle reciprocal pattern comes into play rather than an
all-or-nothing one.
Why would we have an all or nothing pattern if we didn’t have this chain of back-
and-forth interactions? What happens when a child doesn’t have access to sufficient
reciprocal interactions? Perhaps he can’t gesture or signal well or has an unrespon-
sive parent who is not signaling back or has a parent who is too intrusive and anxious
and can’t respond to his signals or is too self-absorbed or depressed to respond.
For any one of these reasons, we may see a compromise in this fine-tuned interactive

system involving the regulation of mood, feelings and behaviors and instead see
extreme or catastrophic type reactions.
When the child who is not learning how to engage in finely-tuned back-and-
forth interactions where there is lots of feedback has an emotion, he doesn’t impli-
citly expect it to lead to an immediate response from his environment. The
emotion consequently exists somewhat in isolation. Without the modulating influ-
ence of a reciprocal response, the feeling may simply get more intense. The child
is left to using more global responses of anger or rage, fear or avoidance, or with-
drawal or self-absorption. Very young infants are more prone to these extreme
reactions in the early months of life. When they cry, they cry very hard and loud
because they are very frustrated until we help settle them down. When they cry
with anger we can feel their rage in their vocal tone until we soothe them. These
extreme emotional reactions have certain similarities with what has been described
as flight/fight reactions, global reactions of the human nervous system in compar-
ison to finely-nuanced back-and-forth interactions with finely regulated emotions
and behaviors. Children are not limited to flight/fight reactions. They can have a
variety of global reactions including avoidance, withdrawal, self-absorption, and
fear, rage, or impulsivity.
The child graduates from global responses in the early months of life to nuanced
fine-tuned reciprocal interaction patterns (from about 3 months to 9 or 10 months).
By 9 or 10 months these finely-tuned reciprocal patterns become well established
and then become more complex, goal-directed, and problem-solving between 10
and 24 months. By the time the child is talking up a storm at age 2 and 21/2, he
already has the capacity to be involved in long chains of back and forth interactions
(back-and-forth reciprocal interactions) involving different emotions and behaviors.
If all has gone well, they are already part of a nuanced, finely-tuned system where
mother’s serious glance signals limits to the child; where her joyful expression sig-
nals permission for the child to be more expansive and try new things; where the
child’s annoyance signals the parent to have a quizzical look of “what can I do to
help you out?” and so forth. These series of actions and interactions involving emo-
tions, mood and behavior help the child to develop a finely-tuned, subtly-nuanced
regulatory capacity, rather than to retain and exaggerate early global catastrophic
What we have been describing is especially important when we see children who
are operating in a catastrophic or extreme manner, that is, extreme meltdowns or
tantrums, or getting carried away with excitement, anger, sadness, or even depres-
sion. Often, these extreme reactions will mean that not only are things happening at
the moment to cause anger or despair, but that the reactions are out of proportion to
the events of the moment, suggesting that some parts of the child’s feelings, mood
and behavior didn’t have a chance to become regulated through long chains of back-
and-forth reciprocal interactions.
For most families it’s easier to get involved in back-and-forth negotiations around
certain behaviors and feelings than others. Some children and caregivers develop
long chains of regulated reciprocal interactions around assertiveness and anger quite

well, but don’t do it as well around sadness or sense of loss. Other are just the reverse
and are better around sadness and loss and issues of intimacy and warmth, but are
very anxious and therefore don’t interact very much or don’t interact in a finely-
tuned nuanced way around assertiveness or anger. In this context, it is not surprising
that as children become older they have different modulation capacities for different
feelings. As adults many of us are better with one set of feelings than another set of
feelings in the way we regulate our moods, have a sense of being able to negotiate
feelings with others and understand them in ourselves.
As children learn to regulate their behavior and feelings as part of long chains
of back-and-forth interactions they can use this skill to get to the next level where
they can problem-solve with the feelings and try to change what’s happening in their
environment. If it is unpleasant, they can do things to change the situation and the
feeling. If it is a pleasant feeling, they can change their environment to bring on
more of those events. By 18 to 20 months, toddlers can already try to lessen those
conditions that make them feel sad or angry and increase those conditions that
make them feel happy. It could be the way they arrange the toys, the way they flirt
with and seek out a caregiver. Toddlers are already becoming active social problem-
solvers using finely regulated emotional interactions, now in a larger problem-
solving context.
Children progress further up to age 2 or 21/2, and, as indicated earlier, can form
images in their mind (symbols or ideas) and label the feelings that have come
under fine regulation. They understand these regulated feelings more fully than
unregulated ones. When a feeling is part of a regulated interaction with someone
else, one understands its context and gains perspective on it. We see this in the pre-
tend play of well-regulated children. They have many long chains of back-and-
forth interactions with many details in their dramas as they create scenes where
there are subtle interactions around anger, happiness or sadness. Children who are
more extreme in their reactions in contrast tend to evidence more global pretend
play patterns. The characters get excited and bang the floor, simply crash cars, or
just go to sleep. Their pretend play often reflects the degree to which they can regu-
late and modulate their feelings. For children capable of regulated reciprocal inter-
actions at the level of using ideas, they are able to label feelings and begin to
ponder them.
At the next level (building connections between ideas), as discussed earlier, chil-
dren can begin reasoning about their feelings, figuring out why they are happy or sad
or joyful. This occurs between ages 3 and 4. As they become older, they can further
reflect on these feelings and understand them in a larger context of their peer rela-
tionships. They can understand more subtlety in the gray area of feelings, in other
words, how angry, happy, or sad they feel. As a child becomes older this capacity for
reflective thinking with feelings becomes stronger and stronger.
These steps of labeling and reflecting on feelings depends on that important
earlier step of experiencing these feelings and behaviors as part of a back-and-forth
signaling system. Without that step, the feelings stay in a catastrophic global mode
(the all-or-nothing mode where I am extremely angry or excited). The continuation

of a global, unmodulated quality interferes with the progression to reflective thinking

because it’s hard to symbolize unmodulated extreme feelings. There is too much
pressure to put the feelings into action rather than into ideas and words. The feelings
are so strong in the unmodulated form that they create a reality about them, a sense
of urgency and immediacy that is stronger than the idea that could represent or sym-
bolize them. In other words, when the feelings are extreme, the tendency is to dis-
charge the feeling, put it into action or just shut down and avoid it entirely rather than
label and think about it. The child who is in the middle of a rage may be able to say,
“I’m angry,” but he is still biting, yelling, screaming and hitting. Even if he can label
a little bit, the labeling isn’t serving the purpose of using the idea or reflecting on it.
In contrast, the child who says, “I’m angry. I need my milk now, please” is using the
idea to convey the feeling. He is not being dominated by the feeling but is trying to
communicate something that will alter the feeling.
Perhaps most importantly, the capacity for a continuous flow of back-and-forth,
affective signaling is essential for establishing a sense of reality, reality testing (the
ability to distinguish fantasy from reality), the capacity for organized thought
(rather than fragmented thinking), and an organized sense of “self.” Children who
do not engage in a continuous flow of affective gesturing live in a world of islands
of back-and-forth episodic reality contacts with the world. Their sense of reality, as
defined by the continuous reciprocal exchange of information with another (to bal-
ance and define one’s inner life) is piecemeal. As a result, their thinking, sense of
reality, reality testing, and sense of self is fragmented. Not surprisingly, many chil-
dren with special needs who make outstanding language and cognitive gains have
difficulty with these higher-level capacities involving a stable sense of “self” and
“reality,” because their programs don’t sufficiently emphasize a continuous flow of
affective gesturing as a foundation for higher level capacities. In contrast, the con-
tinuous flow of affective gesturing, when coupled with emerging symbolic interac-
tion, creates the nexus for organized, reality-based thinking, a stable sense of self,
and the capacity to learn still higher level capacities for judgement, self reflection,
and insight.
It should also be pointed out that extreme anxiety tends to disrupt the child’s
capacity for long reciprocal chains of subtle and differentiated affect cueing.
Individuals, when they’re anxious, typically stiffen their facial muscles and become
more global (i.e., less differentiated) in their affect expressiveness, even though their
anxiety or fear may be quite observable in their facial features. The result of this anxi-
ety-based interruption in affect cueing is a tendency toward fragmented (rather than
integrated) thinking. Fragmented thinking, in turn, lends itself to distortions, exag-
gerations and polarizations, and further intensifies fear and anxiety. This pattern
leads to a cycle of increasing dysfunction.
Children with sensory processing difficulties, including visual-spatial processing,
auditory processing, or motor planning, or sensory modulation challenges are
especially prone to extreme anxiety because it’s more difficult for them to regulate,
comprehend, and operate on their social and physical worlds. It is, therefore, par-
ticularly easy for children with special needs to become involved in these cycles of

anxiety and fragmented thinking. As indicated, the fragmented thinking then leads
to more polarized perceptions and often catastrophic emotional states (which oper-
ate at an all-or-nothing level, rather than a finely regulated, modulated one) and the
intensification of the cycle of dysfunction.
Recognizing this pattern has important implications for intervention. The most
helpful intervention strategy is to calm the intense anxiety and fear through reestab-
lishing affective or emotional contact with the child in a highly soothing manner.
Then it is vital to begin the process of reciprocal, affective exchanges with an espe-
cially soothing manner and tone. One is simultaneously calming the child and bring-
ing the child back into reciprocal affective interchanges, which will enable him to
organize his thinking, become more reality based, and exercise more finely tuned
judgment (rather than operate under the agencies of catastrophic affects in highly
polarized, distorted thinking patterns). The long term goal is to facilitate the child’s
ability to better and better re-engage on his own so that he can become a better agent
at initiating long chains of soothing, affective interchanges.
A child who is operating under fine-tuned regulated affective interactions will be
in solid position to use logical reality-based and integrated thinking to gradually
increase his intellectual and social capacities. On the other hand, a child operating
under states of intense anxiety with polarized and catastrophic patterns will be com-
promised in his or her intellectual and social progress.
In addition to dealing with anxiety, the pattern of reciprocal affective interaction
can be a helpful therapeutic tool for working with problems with unstable or labile
mood, attention and activity problems, and impulsivity. The individual with labile
moods has difficulty sustaining a mutually soothing, calm, reciprocal affective pat-
tern. If the reciprocal partner (e.g., therapist, caregiver) can down-regulate (i.e., a
soothing tone of voice while maintaining a continuous flow of affective interchanges),
as the patient or child is revving up, over time a new capacity for more continuous,
soothing reciprocal interactions may be learned (often for the first time). As the
patient is helped to gradually experience emotions or thoughts associated with mood
shifts, these reciprocal patterns are especially helpful. Not infrequently, the original
parental response to what may have been a biologically based tendency was to recip-
rocate the child’s revving up with counter agitation or withdrawal, neither of which
would provide what the child needed.
For the inattentive, active child, the reciprocal pattern may need to emphasize
longer and longer chains of reciprocal interaction (to maintain attention), coupled
with modulation patterns where the therapist or caregiver changes the rhythm of
interaction (from fast to slow, slow to fast) and changes the intensity of response (such
as loud to soft sounds and vice versa). In a sense, one catches the other person into
longer and longer sequences with greater and greater modulation.
For the impulsive individual, the goal would be a combination of the interven-
tion described for the moody and inattentive, active individuals, with increased
soothing, longer sequences, and more modulation. This pattern would be implanted
as the individual was helped to gradually experience the emotion that precipitated
the impulsive behavior.

Affective Reciprocity, Evolution, and Intelligence

The capacity for a continuous flow of increasingly differentiated affective expres-

sions may be a universal process in the development of intelligent behavior in dif-
ferent members of the animal kingdom. This hypothesis would suggest that the
greater the capacity of the animal (or group of animals), including humans, to evi-
dence differentiated, affective expressions and use them in gestural interactions to
communicate and problem solve, the greater the animal’s intelligence. In addition,
the greater the capacity for a continuous flow of differentiated affective interaction,
the greater the animal’s ability to form complex social networks. Furthermore, the
greater the social networks, the greater the ability to deal with basic needs and, there-
fore, the greater the ability to employ mental resources for the development of higher
levels of intelligent behavior.
The capacity for affective reciprocity may, therefore, have been an important
step in evolution as well as in human development. In human development, it is a
critical step that enables exploration of the world and the regulation of behavior,
mood, and eventually, thought. In evolution, the capacity for affective reciprocal
interactions may have played an important role in the development of complex
social organizations and culture. Reciprocal affective interactions are necessary for
social groups (i.e., a method for quick and efficient communication of intents with-
out resorting to extreme behaviors). Aggression, dependency, sexuality, and the
like could be negotiated through affective cues. For example, an aggressive gesture
could lead the “other” to gesture submission without a violent confrontation, as is
often seen among mammals. Many such interchanges equal complex negotiations,
for example, the capacity to communicate and negotiate favors. Social organiza-
tions which, in turn, allow for more efficient self-and group protection, food, hous-
ing, and weaponry production. Freedom from meeting survival needs during all of
the waking time, in turn, would permit time for other endeavors, including more
complex social interactions, pattern recognition under non-catastrophic survival
emotions, and exploration of the next step beyond complex patterns, i.e., symbol

Affective Reciprocity and Symbol Formation: Perception

Without Action

This section will suggest a hypothesis about how symbol formation arose from
the capacity for affective reciprocity. The ability for affective reciprocity may be both
a critical step in the evolution and the human development of thought and sym-
bol formation. As with many human capacities, evolution may repeat itself in
the adaptive development of each infant and child. As indicated earlier, affective
reciprocity is associated with regulating emotions and behavior and, thereby, mas-
tering catastrophic behavioral discharge, feelings, and interactions. The regulation of
behavior through affective signaling enables the individual to have perceptions

without taking action. What is perception without action? Perception without action is
imagery. When the capacity for imagery is combined with the intentions (and their
related affects) that are involved in reciprocal affective interactions, symbolic mean-
ing develops.
Furthermore, while catastrophic survival emotions would keep one pretty well
tied to concrete behaviors, modulated reciprocal affective states, in contrast, would
promote higher-level mental (less concrete) capacities. It would promote associative
learning and allow time for associative mental processes because affects would be
motivating feedback-oriented interactions and learning from the environment and
others, rather than immediate survival-oriented action. Symbolic art characterizing
real events may have been an early form of symbolic expression. In other words, the
capacity for affective reciprocity would favor social organizations, pattern recogni-
tion, symbol formation, and other higher-level mental abilities—all of which, in turn,
would have evolutionary advantage.
In addition, the long interactive patterns that result from reciprocal interchanges
create the capacity for expectations (i.e., pattern recognition). It is only a small step,
then, to being able to abstract these “expected” patterns into a symbolic form. Both
the taming of the extreme catastrophic emotions and the formation of complex pat-
terns of expectations are important contributors to symbol formation.
Therefore, in both each child’s development and in the evolutionary patterns that
have preceded us, it is likely that the capacity for affective reciprocity enables the
regulation of behavior and mood (i.e., the mastery of catastrophic behavior discharge
emotions), and promotes social signaling, more efficient social organizations (which
frees individuals from survival necessities and promotes economic and cultural
growth), and the capacity for imagery and symbolic functioning. Much of this occurs
because affective reciprocity frees perception from action.

Variation in Affective Reciprocity and Symbol Formation

These high-level mental capacities are far from universal, even among human
beings. Affective reciprocity and the resultant capacity for symbolic functioning, as
well as derivative social, emotional, and intellectual capacities, vary in relationship to
experience. These capacities are not completely relative, however,(as one might
expect from the perspective of cultural relativism). In spite of variation, there are two
sources of similarity among peoples. One relates to the human perceptual apparatus.
There are certain commonalties among most people in terms of, for example, fre-
quencies of sound that will register and patterns of color or light that can be per-
ceived. Less obvious, there are also certain functional developmental capacities (in
terms of their basic structure) that are also shared by most people. These include, in
addition to the capacity for affective reciprocity discussed earlier, the capacities for
shared attention, relating to others, engaging in purposeful or willful interactions,
negotiating complex reciprocal affective problem-solving interactions, creating ideas,
building bridges between ideas, and progressing to some degree of higher-level

abstract and reflective thinking. It is interesting to look at how different cultures pro-
vide their own form of opportunities for babies to engage in experiences leading to
these basic functional developmental structures and to see if some cultures do not
provide such opportunities. In fact, the field of cultural anthropology may wish to
consider these structural capacities together with ones already being studied.
The common perceptual features (e.g., sound registration) are based largely on a
relatively similar biology. The common functional developmental capacities for relat-
ing, communicating, and thinking are based on common child-rearing and family
practices and include experiences such as nurturing human caregiving to promote
relating and sensitive affective signal reading and opportunities for interaction to pro-
mote purposeful, willful interactions and affective reciprocity. To the degree most
infants have access to nurturing interactive caregivers, these structural personality
characteristics will be at least somewhat similar among people.
At the same time, however, there will be variations on these capacities based on
both biology and experience Some babies, for example, have a narrower range of
functional sound recognition and are quite sensitive or reactive to high frequency
sounds. Many children will not have access to nurturing caregivers and, therefore,
will not learn to engage with others and participate in a continuous flow of affec-
tive interactions and, therefore, will not master the capacity for affective recipro-
city. Consequently, their capacities for symbolic functioning will also be quite
At present, however, only small numbers of children evidence these more
extreme biological or psychological patterns. What will happen if most children do
not have opportunities to learn how to engage with caregivers and master the capa-
city for affective reciprocity due to changing child rearing, such as poor quality day
care? What will happen if, due to toxic substances such as lead or dioxins, most chil-
dren evidence extreme difficulties in the way their central nervous systems func-
tion? In such circumstances, the common ground that supports shared human
capacities will decrease. This decrease will not lead simply to more human diver-
sity, however. It will erode a shared sense of what constitutes reality, rational
thought, or moral behavior. The glue that holds many societies together would be
in jeopardy.
In addition to these relatively stable features of human functioning (for the
moment at least), there are other features of human functioning that clearly vary
from person to person. Within each of the functional developmental capacities, such
as relating, affective reciprocity, and using symbols, each individual, based on their
unique experiences, will cast their own signature. One individual will interact more
around assertiveness, another around dependency, another around sexuality.
Unique cultural and family experiences will contribute to these personal signatures.
While sharing the capacity for thoughts and feelings, each individual’s specific
thoughts and feeling will, therefore, be uniquely personal and related to their per-
sonal history.
Importantly, however, the capacity to organize uniquely personal experience at a
gestural, affective reciprocal or symbolic level (i.e., form a symbolic sense of self)

depends on the basic functional developmental capacities described above. As indi-

cated earlier, perhaps the most critical of these functional developmental capacities for
human development and for evolution is the capacity for long chains of affective rec-
iprocity. It should not be surprising, therefore, that working with affective reciprocity
should be at the center of clinical work with children with autistic spectrum disorders.

From Affective Reciprocity to the Development of Ego

Signal Functions

Another important dimension of reciprocal, affective interactions is that they lead

to the capacity for an individual to use affect as a signal, which fosters anticipation
and consideration of alternatives rather than direct discharge, shut-down, or with-
drawal. The regulation made possible by back-and-forth affective exchanges, as indi-
cated, leads to symbolization and the symbolization of affect makes possible the use
of affect as a signal.
Consider the case of a latency-aged child who tended to be either agitated or
depressed. The goals of his psychotherapy were to help him progress to higher levels
in his functional emotional capacities (i.e., ego development). Specifically, one of
the goals was to enable him to engage in longer chains of regulated, reciprocal
affective exchanges. During longer and longer exchanges, the therapist would help
Andy better regulate his affective and behavioral expressions through critical pre-
verbal, as well as verbal responses. For example, when Andy would begin to evi-
dence more agitation in his voice and body movements, the therapist would
deliberately move towards a more soothing, comforting tone to attempt to down-
regulate the intensity of affect. When Andy would become more apathetic and self-
absorbed, the therapist would deliberately move towards a more energized rhythm
of preverbal and verbal exchange (e.g., more animated facial expressions and faster
tempo) to up-regulate. At the same time that the therapist was working at the pre-
verbal level, he would also, periodically, explore how Andy felt during these shifts
of affective rhythm and intensity. During these times, he was attempting to help
Andy symbolize and reflect on the subtle feeling states Andy was experiencing
when either agitated or apathetic.
Over a period of six months, Andy was able to make progress towards both of
these goals. He gradually responded to the therapist’s soothing, comforting tone of
voice and interactive rhythm by becoming more regulated (less agitated) when talk-
ing, for example, about his father being unfair or kids at school picking on him. He
was also able to begin verbalizing more abstracted feeling states, shifting from
somatic descriptions and descriptions of actions he was going to carry out to true
descriptions and reflections on inner feelings. For example, instead of talking about
his exploding insides or how he was going to punch so-and-so, he began describing
“Feeling like my insides were I was so mad.”
Interestingly, as Andy was able to symbolize affect, he began being able to use
affect as a signal (i.e., to both unconsciously and consciously anticipate next steps).

For example, he began to become aware of feeling angry and then could consider
alternate actions, when his brother came into his room uninvited. It’s important to
note that his capacity to use affect as a signal was based on his first learning to reg-
ulate reciprocal affective exchanges and then describe his affective states symboli-
cally. Both steps appear to be important. Without the regulation, affective states tend
to be intense and, therefore, are often experienced in an overwhelming or cata-
strophic manner, and there is a tendency toward discharge or somatization or inter-
personal withdrawal. The regulation of the affective interchanges enables shifts
towards symbolization (i.e., greater awareness and description of subtle affective
states) and, in turn, the symbolization enables the affects to also serve as intrapsy-
chic signals.

Implications for Attention and Processing Capacities

The importance of reciprocal emotional interactions also becomes central when

we are trying to help children who are having challenges, such as fears, anxieties, or
impulse control problems. These and other challenges often involve some degree of
unregulated feelings. Therefore, in addition to other therapeutic strategies deter-
mined by the child’s and family’s unique developmental profile, creating opportuni-
ties for long empathetic nurturing interactions around the child’s different feelings
can often go a long way to helping a child learn to regulate feelings and behavior
(e.g., parents get down on the floor, follow their child’s lead and tune into the child
at the level of gestures and, if the child is 2 or older, through pretend play as well).
Interestingly, back-and-forth, regulated, affective interactions also help the child
develop many of their most important information processing abilities, such as their
capacity for attention, planning and sequencing actions, understanding space (visual-
spatial thinking), and using words and language meaningfully.
Consider language. Words have to have meanings and meanings are conveyed
through the emotional context within which a word is used. A word that is used as
part of a back-and-forth pattern of negotiation will be a word with many emotional
subtleties to it. The child who learns the word “fair” or “unfair,” for example, can
learn this in an all-or-nothing way, this is simply “fair” or “unfair,” or can have subtle
gradations of what’s a little fair, very fair, extremely fair, or not fair, somewhat not
fair, etc. These subtle gradations will only be learned as part of interactive relation-
ships that involve a lot negotiation.
More illusive, though, is how these reciprocal interactions help a child learn to
pay attention and plan and sequence actions (e.g., executive functions). Many have
wondered why we are seeing more children with attention and motor planning prob-
lems. One contribution might be fewer opportunities for long chains of reciprocal
problem-solving interaction in child/caregiver relationships. Consider how this
process may work. In order for a child to carry out a two-step and then a three-step
and then a five-step plan, it is very helpful to have all the elements of that plan
invested with lots of emotion.

The ability to carry out a number of steps in a row, that is, to plan and sequence
actions, begins with simple reciprocal interactions. The emotional recognition that
one’s actions can have an impact on someone else is the foundation for sequencing
(i.e., the ability to carry out many steps in a row where each one is related to the prior
one). This is exactly what reciprocal interactions are. Smiles lead to smiles, frowns to
frowns, sounds to sounds, movement to movement. The harder it is for the child to
continue a long chain of interaction, the more important it is to practice long, regu-
lated sequences of interactions.
In order to practice affectively, however, one needs to put more and more emo-
tion into the interactions. The more vibrant and interesting and emotionally mean-
ingful the interaction is to the child, the longer the sequence the child will mobilize.
This doesn’t mean loud or extreme emotions, it means subtlety and expression (soft
or gentle, exciting or captivating, depending on the moment and the child). In
essence, some degree of emotion or intent is what carries the child’s ability to
sequence. Therefore, the emotional interest that the adult must bring to the situation
will have to be greater when the task is harder.
As a child graduates from simple reciprocal interactions to more problem-solv-
ing interactions, the ability to plan and sequence also improves because the child is
now sequencing longer chains of reciprocal interactions. At the next level where
ideas regulate behavior, a child can map out sequences with visual images and ver-
bal ideas. And as logical thinking begins to dominate, the child can plan strategies for
executing actions and consider a number of alternatives. The entire system of plan-
ning and executing problem-solving strategies, though, depends on the child being
able to link many pieces of behavior together into a long reciprocal pattern.
The degree to which these patterns can be elaborated in many different domains
of life increases the degree to which the child can plan and execute actions in these
different domains. For example, some children are very good at social negotiations,
but not very good at fixing a toy, figuring out how to find a hidden object, or self care
in the bathroom.
The domains that are more difficult for the child require more involvement with
long chains of reciprocal emotional interactions. In essence, one needs to build up
the basic patterns in areas that are hard for the child. For example, a child who is very
engaged, warm, and is very good at social sequencing, negotiation, and problem-
solving has a hard time with visual-spatial sequencing (he has never been able to
conceptualize how space is laid out). For this child, emotional interactions using
space, such as hide-and-go-seek games, treasure hunts, other search games and like,
would strengthen this basic capacity. This same principle applies to children who
have trouble understanding the spatial dimensions of their own bodies, integrating
their left and right sides, for example. Here, too, reciprocal interactions involving the
human body, games such as “touch my nose and I’ll touch your nose” would facili-
tate body-oriented visual-spatial problem-solving.
Many clinicians who work with exercises such as left hand to right shoulder or
different kinds of games involving different types of crawling and walking often
believe that all that is going on is exercising the body and processing capacities. The

gifted clinician working with the child, however, is often working in a highly recip-
rocal manner developing reciprocal sequences in the context of visual-spatial think-
ing, or other processing abilities. It is the chains of regulated reciprocal interaction in
that particular domain that creates the foundation for improved cognitive abilities in
that domain.
It would be useful to study caregiving patterns associated with different capacities
for attention and sequencing. The hypothesis would be that less reciprocal caregivers
are associated with greater attentional and sequencing problems in their children.
Furthermore, the optimal attention and sequencing capacity would be associated
with caregivers capable of long reciprocal chains in many functional areas (e.g.,
motor, language, visual-spatial, social) with a tendency to support the child’s gradu-
ally increasing assertiveness. The caregiver would begin with a continuous response
pattern, but gradually shift to a more intermittent pattern as the child became more
self sufficient. For example, the caregiver would reciprocate each gesture initially, but
as the child became able to sustain his or her initiative, reciprocate relatively inter-
mittently but sufficiently to help sustain the pattern. If the child were talking, one
would gradually shift to letting the child say more and more before commenting or
asking a question.

Affective Reciprocity, Motor Planning, Rhythmicity, and the

Development of Language

Long chains of reciprocal interaction support the toddler’s capacity for engag-
ing in planning and sequencing actions (i.e., long problem-solving sequences
involving social interaction and mastery of his physical environment). These
sequences of long problem-solving interactions, in turn, support many critical
aspects of language development. As discussed earlier, preverbal gestural (i.e.,
problem-solving actions), social, and problem-solving interactions create the con-
text for meaning of verbal symbols. Without this basic level of knowing through
doing, words, when they became possible, would have no experiences to draw on
to create meanings. In other words, the child knows what love is through his hugs,
cuddles, and flirtatious glances. The word summarizes the concept or meaning he
has been gradually forming. Similarly, words for “open,” “up,” or “door,” all have
their experiential contexts.
Reciprocal affective interactions also influence the basic grammar and semantic
aspects of language. We have found, for example, that children not capable of recip-
rocal affective interactions (e.g., children with autistic spectrum disorders), tend to
use words ungrammatically, repeating nouns or verbs perseveratively, for example.
Interestingly, if we try to simply correct their grammar, it doesn’t work very well.
They make progress, however, when we first help them engage in reciprocal affec-
tive gesturing and use their affect and gesturing purposefully (e.g., we get stuck
behind the door they are opening and closing and they eventually learn to push us

away). At that point, they begin to align their verbs and nouns in a grammatically
correct manner—“Daddy, go!” or “Leave me alone.” We observe the same patterns
in children from deprived backgrounds, such as orphanages. Whether the insults are
environmental or biological, it appears that a prerequisite for correct use of grammar
is the purposeful use of affects in interactive relationships. This fact may have been
missed by linguists who suggested grammar was largely innate and simply turned on
or off by global features of the environment because it’s easy to take reciprocal affect
cueing and other preverbal aspects of communication for granted. They occur so reg-
ularly. It’s only when we find circumstances where they don’t occur that we can see
their true impact.
Similar to grammar, the meaning of words, both the semantic and pragmatic
aspects are also imbedded in the earlier reality of gestural interactions, which are
used to explore and know the world. The literal meaning of a word or concept, for
example, the concept of a door or a table or a mommy or a daddy is first known
through gestural interactions with it. The capacity to form the word is then linked to
what is already partially known. The known entity takes on additional meaning
through context and further emotional experience with it. Therefore, both the literal
and the relative meaning of words and concepts emerge from reciprocal affective
interactions which provide the foundations and context for meanings.
The capacity for long chains of reciprocity and the basic capacity to plan and
sequence actions may also support the ability to sequence words or ideas and eventu-
ally concepts in a speech, essay, or debate, or simply a long conversation. Sequencing
ideas relates both to this basic ability to abstract meaning from earlier preverbal expe-
riences and then sequence them meaningfully.
Also, as discussed earlier, the very capacity for forming symbols derives from the
toddler’s ability to negotiate through reciprocal interactions involving long problem-
solving sequences. An image not tied to the need for direct discharge can become a
thought (i.e., a symbol). In addition, long chains of reciprocal communication
become pleasurable in their own right because they foster greater co-regulation of
relationships and provide safe and secure ways to express and negotiate feelings and
needs and therefore become an end in their own (rather than a means to some other
There are, however, other subtle contributions of affective reciprocity and
motor planning and sequencing to language development. Affective reciprocity and
planning and sequencing capacities occur in rhythmic patterns. Vocal and motor
interactions between caregiver and toddler exist in various rhythms, rapid
exchanges of vocalizations, slow exchanges, simple rhythms of monotones, novel
rhythms of animated dialogues, and so forth. We all get bored and tune out with a
slow, even, rhythmic monotone and perk up and listen better with a more rapid,
novel, changing rhythm, particularly when these are in tune with the meanings of
the words being expressed (e.g., speeding up or slowing down to emphasize an emo-
tional point). Some people, however, are better at listening, even if it’s a slow mono-
tone. Similarly, as infants and toddlers are identifying and using vocal patterns,
there is an intimate relationship between what can be perceived (i.e., processed

auditorially) and what can be articulated (a child hears sound patterns and tries to
imitate them). A toddler or child who has a hard time planning and sequencing
actions and therefore with reciprocal affect cueing, may have a relatively more dif-
ficult time engaging in and recognizing the rhythmic patterns underlying aspects of
auditory perception and expressive language.
In many respects, the capacity of rhythmic interactive patterns, which is a com-
ponent of motor planning and affective reciprocity, is a vital dimension of lan-
guage. There has been recent research on the temporal aspects of auditory
processing and language, including intervention strategies developed based on
temporal dimensions, for example, slowing down the presentation of sound
sequences (Tallal, et al., 1996). Another vital component of the temporal system,
however, is the rhythmicity of sound sequences. There is likely very different
thresholds of sound sequencing for perceiving patterns. Too slow a presentation,
for example, may make pattern recognition more difficult. Research in progress on
the Interactive Metronome (Shaffer, Jacokes, Cassily, & Greenspan, 2001) is reveal-
ing an optimal rhythmic range below which sound patterns are very difficult to per-
ceive. This research is also identifying differences in the degree to which
individuals with different processing challenges and diagnoses can perceive these
Reciprocal gesturing and motor planning and sequencing, particularly its rhyth-
mic dimensions, may also be especially important for oral-motor sequencing capa-
cities and overcoming oral-motor dysphoria, as well as stuttering and lack of
expressive intonation. Improving reciprocal affective gesturing and the related
capacities for motor planning and sequencing may, therefore, contribute in a num-
ber of ways to the processes that support language development. These may range
from the basics of auditory perception and imitative production of sounds and words
to symbol formation and meaning.
In summary, a baby uses simple exploration of relationships outside herself to
enter into reciprocal interactions. She then uses reciprocal interactions to explore her
world. Longer and longer and increasingly more complex chains of reciprocal inter-
actions, using the different senses and the motor system, then give birth to a variety
of the seemingly discrete cognitive, language, social, and visual-spatial and motor
planning skills. Gradually, the child progresses through the six stages described ear-
lier to higher levels of creative and reflective thinking.

Implications for Children with Special Needs

An important consequence of understanding how important reciprocal affective

interactions are for regulating emotions, mood, and behavior relates to therapeutic
and educational interventions for children with special needs.
Many children with developmental or emotional problems have difficulties at the
fundamental level of back-and-forth signaling globally or in a certain area, as indi-
cated above. Their difficulty may be an inability to engage in or sustain long chains

of affective reciprocal interactions or to mobilize them in specific processing or emo-

tional areas (e.g., around motor planning, visual-spatial thinking, or feelings of anger
or loss).
An appropriate therapeutic approach must work at the level of facilitating
reciprocal affective interactions generally and/or specific areas such as language,
motor planning, visual-spatial thinking, or expressing anger. The Developmental,
Individual-Difference, Relationship-Based model (DIR, i.e., floor time) conceptu-
alizes a comprehensive approach with reciprocal exchanges of affect cues at its
foundation (Greenspan, 1992; Greenspan & Wieder, 1998). Many therapies, how-
ever, try to help children without establishing reciprocal chains of affective interac-
tion generally or in the area where they are missing. For example, many very
structured approaches to therapy, such as behavioral approaches for children with
autistic spectrum disorders, fail to realize that one of the primary goals for many
of the children is to establish the ability for interacting in a continuous reciprocal
flow, that is, for entering into these long chains of reciprocal back and forth
communicating. In fact, the lack of reciprocal interaction is one of the key deficits
in autistic disorders. The goals of all interventions for autistic spectrum prob-
lems needs to involve reciprocal affective interaction, abstract and creative
thinking, true empathy, flexible peer relationships, and emotional flexibility and
Behavioral treatment tends to be a stop-start approach, rewarding the child for
matching a shape, for example, or for repeating a certain sound. Gifted therapists, in
spite of the curriculum sometimes enter into continuous flow with the child, which is
usually associated with better progress. Ivar Lovaas, at a meeting where he and I pre-
sented and compared the ABA Discrete Trial approach and the Floor Time approach
(and had a chance to discuss the relative strengths of both approaches), showed a
videotape of himself working with a child in his early years. In the tape, he was enter-
ing into a continuous flow of back-and-forth reciprocal communication, which is very
different from many of those who carry out his procedures. He was charming and
warm and exchanged lots of affect while purporting to offer rewards and punish-
ments. I gently pointed this out to him and he intuitively acknowledged the impor-
tance of getting into a relationship with a child where the child feels warm,
appreciated, and nurtured. This didn’t lead to a fundamental altering of his theoreti-
cal position about what he felt was important, however.
The stop-start approach advocated by formal behavioral theory and procedures
often interferes with caregivers or therapists establishing a continuous reciprocal flow
and affective rhythm. We’ve seen many children in consultation who have been
involved in intensive, 30 + hours/week behavioral (discrete trial) programs. We’ve
also observed children as part of a research program who were involved in intensive
30 + hours of behavioral-discrete trial programs. The children we observed in the
research program were described by their therapists and parents as having had very
good outcomes. The patterns we observed are informative and suggest some of
the strengths but also some of the weaknesses of behavioral approaches which did
not focus sufficiently on a continuous flow of spontaneous back and forth affective

gesturing and (and use of language). While there were exceptions the general trend
we observed are as follow: the children in these programs generally have learned to
use some words and even read and do math, but often cannot engage in a pattern of
continuous relating and back-and-forth gesturing; they have difficulty with sponta-
neous conversation; their behavior is often unregulated and unreciprocal in the sense
that they can only conform to specific rules under specific circumstances; new situa-
tions are difficult; creative and abstract thinking, including the capacity for making
inferences is severely restricted.
In fact, as indicated earlier, a deficit in reciprocal interactions and creative and
abstract thinking is a defining characteristic of autistic spectrum disorders
(Minshew & Goldstein, 1998; Dawson & Galpert, 1990; Tanguay, et al., 1998).
After years of exlusively very structured behavioral intervention (stop/start), rather
than a continuous flow of spontaneous reciprocal interactions, in our observation,
many children still present as self-absorbed, idiosyncratic and repetitive, even
though with the proper cues they can do very concrete and memory-based aca-
demic tasks.
When we then work with these children on their ability to interact with affective
gestures, starting with simple fun games, and work up to back and forth negotiations
(to open the door or to get the juice), we find that many children can move towards
a continuous flow of affective interaction. As we engage children in this way, their
repetitive, idiosyncratic and unreciprocal behavior begins to change. They begin
using their behavior and existing language and thinking skills in a more purposeful
creative and abstract manner. For children who begin this work at age 8 or 9, there
is lots of catch-up, however. It can take a number of years to help them develop the
basic skills for reciprocal affective gesturing that were skipped over. Many children
appear to develop these skills more quickly and fully when intervention is begun at
younger ages (Greenspan, 1992; Greenspan & Wieder, 1998). Many children bene-
fit from a balanced program where there is a focus on both spontaneous reciprocal
affective interchanges and semi structured problem solving activities. In our expe-
rience the key is that when working on specific goals in a semi structured way, the
semi structured activities be set up as a challenge which illicits enthusiastic affect and
a continuous flow of back and forth interaction while meeting the challenge. For
example, teaching a child to “open” in the context of his trying to open the door to
get his favorite toy which has been deliberately placed behind the door.

Implication for Child Care

The notion that emotional relationships are essential for regulating thoughts,
behavior, moods, and feelings is one that needs greater emphasis as we think about
child care settings and children’s activities. First, consider that the chains of long
back-and-forth affective interactions that are necessary for regulating our behavior,
moods and feelings can only occur with a loving caregiver who has lots of time to
devote to her child.

A busy day care provider with four babies or six or eight toddlers usually won’t
have the time for these long sequences of back-and-forth emotional interaction.
Similarly, a very depressed mom or dad or an overwhelmed caregiver with five chil-
dren may not have the needed time, energy, or know-how. Full-time working par-
ents, exhausted at the end of the day, may not be able to engage in these long
patterns of back-and-forth interaction. In addition, parents who are being told that
structured exercises with picture cards or computer games can improve their chil-
dren’s intelligence (rather than person-to-person interaction) may mistakenly take
time away from important spontaneous, reciprocal interactions in order to do “exer-
cises.” They often do not realize that what they are giving up is more important.

Implications for Children with Emotional and

Behavioral Problems

The importance of a continuous flow of affective communication is also espe-

cially relevant for many traditional types of psychotherapies that work with a range
of children, adolescents or adults with circumscribed challenges. The following com-
ments are, therefore, important for children or adults with special needs who have
progressed to being quite verbal and related but who have circumscribed emotional
challenges and for children and adults who do not have a history of special needs,
but evidence selective emotional difficulties.
Traditional psychotherapies tend to focus on listening, reflecting back and sum-
marizing what is being told to the therapist. For the child, this may involve reflecting
on patterns revealed in his play; for an adult, it may involve carefully listening and
summarizing back empathetically the feelings that are being shared with attempts at
insight and pattern recognition. Many therapies involve a fair amount of empathy,
warmth and support and sometimes selective guidance and advice.
What’s missing, however, as a point of emphasis in many child and adult insight-
oriented therapies is the importance of establishing a back-and-forth affective recip-
rocal flow. In other words, there will often be long times where the therapist is
relatively passive, taking in, perhaps empathizing, summarizing, and clarifying
periodically, but not working at establishing a dynamic back-and-forth affective inter-
action. Some child therapists do this as part of the give-and-take in play and are very
helpful to their children, but others do not. Some adult therapists who are generally
interactive people will do some of this intuitively, particularly for patients who are a
little more self-absorbed, withdrawn or depressed. By and large, however, reciprocal
affective interaction has not been an explicit focus of therapy, and there is a deliber-
ate stop-start quality to many therapeutic interactions (i.e., long passive pauses rather
than dynamic, gestural, back-and-forth, affective cueing and interactions).
Some patients come to treatment with well-developed reciprocal affective regu-
latory capacities and require help with circumscribed feelings, thoughts, or behavi-
ors. The traditional stop-start approach may not be a problem for such patients
and the patterns, clarifications, or insights discussed may prove helpful. For the

patients whose affects, behaviors or moods are not part of well-established regulatory
interactions, i.e., have not been part of this dynamic continuous flow of regulated
behavior since infancy and early childhood, however, the focus on reciprocity is
paramount for the success of the therapeutic endeavor. As a first step, simply enter-
ing into a reciprocal flow around particular emotional themes may be extraordinar-
ily helpful to the patient. One could postulate that the regulation and differentiation
occurs through reciprocal interactions around particular emotional areas, such as
assertiveness or aggression or sadness or loss or humiliation or shame, etc.
Establishing reciprocity in specific emotional areas is an early step in the therapeutic
hierarchy. Just as this step is necessary for normal development, it is necessary for
many patients who lack this ability in specific emotional areas or generally. The
unregulated patient is at the mercy of global or extreme feeling states. These must be
regulated before they can be fully understood symbolically.
Over time, regulated reciprocal interactions in emotionally and thematically
important areas lead to problem-solving around that particular emotional theme and
then to labeling with ideas and eventually to reasoning. This can lead to high levels
of reflection and insight. The process starts, however, with the ability to enter into
affective reciprocity around the emotional theme in question. The importance of
reciprocal affective interactions may explain why gifted intuitive therapists, even
without a lot of experience, may at times do better than experienced, but emotion-
ally less intuitive, therapists. We all know aides or helpers who don’t have an educa-
tional background who can often be quite therapeutic. Perhaps it is this ability to
establish affective communications in a particular thematic area and engage in a
reciprocal continuous flow in these areas that enables such people to be “therapeu-
tic.” Some individuals have a naturally bigger range of different emotional areas
They can empathetically respond in a continuous affective flow in the areas of
warmth, intimacy, aggression, assertiveness, loss, humiliation, shame, etc. Others are
more limited. The individual’s “range” may also be part of what makes some indi-
viduals gifted in terms of their therapeutic capacity regardless of their background
and training. Interestingly, the explicit training of the therapist may become more
important at the symbolic levels where we try to help the patient label the feelings,
engage in symbolic elaboration and pattern recognition, and develop reflective
capacities as well as insights. Many training programs help professionals develop
these traditional skills which are also important for successful therapy. These skills
can’t be successfully implemented, however, unless the patient comes in already
skillful at the earlier levels of affective reciprocity or is helped to work on these ear-
lier levels.
Some therapists who are less inclined to work on insight and reflection and more
on changing behavior will use cognitive behavioral approaches. Imagery is worked
on in specifically designed ways. Here, too, however, the ability to work successfully
may depend on the mastery of the earlier level of entering into long chains of affec-
tive reciprocity. With greater awareness of the importance of affective reciprocity,
perhaps it will be possible to train therapists more fully. What’s intuitive for some
perhaps can be learned by others.

The importance of affective reciprocity is relevant to other therapists as well, espe-

cially those who use structured techniques (e.g., speech pathologists, occupational
therapists, physical therapists, special educators, and early childhood educators).
Working in areas such as auditory processing, motor skills, or visual-spatial thinking
requires a highly interactive affective reciprocal manner. As one does this, that system
(the motor, visual-spatial processing or auditory processing) undergoes regulation and
differentiation. This enables the child to move up the ladder to symbolic elaboration
and reflection. Structured methods in these different therapies may be more useful
once the child has mastered a reasonable level of affective reciprocity.
Rather than think of this as two steps in a process, where the child goes from reci-
procity to symbol formation, the process can be thought of as operating at multiple
levels at the same time. The emphasis is on reciprocity when it is not present and
hasn’t been mastered. As one moves into the symbolic level, one continues working
on reciprocity but adds on symbolic elaboration and the capacity for reflection.
Structured techniques can also help further differentiate language, motor, sensory, or
visual-spatial thinking skills. They are most helpful, however, when operating in the
context of ongoing affective reciprocal interactions.


We have explored the role of affect in the core deficit in autism and in the devel-
opment of intelligence and social skills. In a sense, we have come full circle. We have
discussed how children with autistic spectrum disorders may uniquely, for biological
reasons, miss a critical developmental capacity, the ability to connect affect or intent
to motor planning and sequencing capacities and, therefore, have a difficult time
engaging in the long reciprocal chains of affective interaction so necessary for cre-
ative and abstract thinking and high-level social skills. We have also discussed how
these same affective interactions underlie intelligence and social development. To
improve assessments and interventions for children with a variety of challenges
including autistic spectrum disorders, it is imperative to appreciate the role of affec-
tive interchanges in disordered and healthy development. The Developmental,
Individual-Difference, Relationship-Based model (DIR) enables us to appreciate the
role of affect in development by systematizing its dimensions (i.e., its functional
developmental level, individual difference, and interactive relationship patterns)
(Greenspan, 1992, 1997b; Greenspan & Wieder, 1997, 1998, 1999).


Bauman, M. (2000). Genetic and neuroanatomical and biochemical pathways. Presented at the
Interdisciplinary Council on Developmental and Learning Disorders’ Fourth
International Conference, Autism: What Are the Causes; Why the Increase in Cases; Redefining
the Standards of Care. November 10–12.

Baranek, G. T. (1999). Autism during infancy: a retrospective video analysis of sensory-motor

and social behaviors at 9–12 months of age. Journal of Autism and Developmental Disorders,
29(3), 213–224.
Baron-Cohen, S. (1994). Mindblindness: An essay on autism and theories of mind. Cambridge,
MA: MIT Press.
Bell, S. (1970). The development of the concept of object as related to infant-mother attach-
ment. Child Development, 41, 219–311.
Benson, F. & Zaidel, E. (1985). The dual brain. New York: Guilford Press.
Bowlby, J. (1952). Maternal health and mental health. WHO Monograph Series No. 2. Geneva:
World Health Organization.
Brazelton, T. B., & Cramer, B. (1990). The earliest relationship: Parents, infants, and the drama of
early attachment. Reading, MA: Addison Wesley.
Chomsky, N. (1966) Cartesian linguistics. New York: Harper & Row.
Chomsky, N. (1980) Rules and Representations. New York: Columbia University Press
Condon, W. S., & Sander, L. (1974). Synchrony demonstrated between movements of the
neonate and adult speech. Child Development 45, 456–462.
Condon, W. S. (1975). Multiple response to sound in dysfunctional children. Journal of Autism
and Childhood Schizophrenia 5, no. 1, 43ff.
Courchesne, E., Akshoomoff, N., Egaas, B., Lincoln, A.J., Saitoh, O., Schreibman, L.,
Townsend, J., & Yeung-Courchesne, R. (1994). Role of cerebellar and parietal dysfunc-
tion in the social and cognitive deficits in patients with infantile autism. Paper presented
at the Autism Society of America Annual Conference. Las Vegas, NV.
Damasio, A. (1994). Descartes’ error: Emotion, reason, and the human brain. New York: Putnam.
Davis, J. O., Phelps, J. A., & Bracha, S. (1995). Prenatal development of monozygotic twins
and concordance for schizophrenia. Schizophrenia Bulletin, 21, 357–366.
Dawson, G., & Galpert, I. (1990). Mother’s use of imitative play for facilitating social responsive-
ness and toy play in young autistic children. Developmental and Psychopathology, 2, 151–162.
Dawson, G., Warrenburg, S., & Fuller, P. (1982). “Cerebral lateralization in individuals diag-
nosed as autistic in early childhood.” Brain and Language. 15:353–368.
Emde, R. N., Biringen, Z., Clyman, R.B., & Oppenheim, D. (1991). The moral self of infancy:
Affective core and procedural knowledge. Developmental Review 11:251–270.
Escalona, S. (1968). The roots of individuality. Chicago: Aldine.
Gardner, H. (1983). Frames of mind: The theory of multiple intelligences. New York: Basic Books.
Goleman, D. (1995). Emotional Intelligence. New York: Bantam.
Greenough, W.T. & Black, J.E. (1992). Induction of Brain Structure by Experience: Substrates
for Cognitive Development,” Developmental Behavioral Neuroscience. 24:155–299.
Greenspan, S.I. (1979). Intelligence and adaptation: An integration of psychoanalytic and
Piagetian developmental psychology. Psychological Issues, Monograph 47/68. New York:
International Universities Press.
Greenspan, S.I. & Greenspan, N.T. (1985). First feelings: Milestones in the emotional development of
your infant and child from birth to age 4. New York: Viking Press.
Greenspan, S.I. (1989). The development of the ego: Implications for personality theory, psychopathol-
ogy, and the psychotherapeutic process. Madison, CT: International Universities Press.
Greenspan, S.I. (1992). Infancy and early childhood: The practice of clinical assessment and intervention
with emotional and developmental challenges. Madison, CT: International Universities Press.
Greenspan, S.I. (1997a). Developmentally based psychotherapy. New York: International
Universities Press.
Greenspan, S.I. (1997b). The growth of the mind and the endangered origins of intelligence. Reading,
MA: Addison Wesley Longman.
Greenspan, S.I. & Lourie, R.S. (1981). Developmental structuralist approach to the classifica-
tion of adaptive and pathologic personality organizations: Application to infancy and
early childhood. American Journal of Psychiatry, 138(6), 725–735.

Greenspan, S.I. & Wieder, S. (1997). Developmental patterns and outcomes in infants and chil-
dren with disorders in relating and communicating: A chart review of 200 cases of chil-
dren with autistic spectrum diagnoses. The Journal of Developmental and Learning Disorders.
Greenspan, S.I. & Wieder, S. (1998). The child with special needs: Intellectual and emotional growth.
Reading, MA: Addison Wesley Longman.
Greenspan, S. I. & Wieder, S. (1999). A functional developmental approach to autism spectrum
disorders. Journal of the Association for Persons with Severe Handicaps (JASH), 24 (3); 147–161.
Hartmann, H. (1939). Ego psychology and the problem of adaptation. New York: International
Universities Press.
Hein A. & Diamond, R. M. (1983) Contribution of eye movement to the representation of space.
In A. Hein and M. Jeannerod (eds) Spatially Oriented Behavior. 119–34. New York: Springer.
Hofer, M.A. (1988). On the nature and function of prenatal behavior. In W. Smotherman &
S. Robinson (Eds.), Behavior of the Fetus. Caldwell, NJ: Telford.
Hofer, M.A. (1995). Hidden regulators: Implications for a new understanding of attachment,
separation, and loss. In S. Goldberg, R. Muir, & J. Kerr (Eds.) Attachment Theory: Social,
Developmental, and Clinical Perspectives. Pp. 203–230. Hillsdale, NJ: Analytic Press.
Holloway, R. L. (1966). Dendritic branching: Some preliminary results of training and com-
plexity in rat visual cortex. Brain Research 2: 393–96.
Kagan, J. (1981). The second year: The emergence of self-awareness. Cambridge, MA: Harvard
University Press.
Kanner, L (1943). Autistic disturbances of affective contact. Nervous Child. 2:217–250.
Lewy, A. L., & Dawson, G. (1992). Social stimulation and joint attention in young autistic chil-
dren. Journal of Abnormal Child Psychology, 20 (6), 555–566.
Mahler, M.S., Pine, F., & Bergman, A. (1975). The psychological birth of the human infant: Symbiosis
and individuation. New York: Basic Books.
Minshew N. & Goldstein, G. (1998). Autism as a disorder of complex information processing.
Mental Retardation and Developmental Disabilities, 4:129–136.
Mundy, P., Sigman, M., & Kasari, C. (1990). A longitudinal study of joint attention and lan-
guage development in autistic children. Journal of Autism and Developmental Disorders,
20 (1), 115–128.
Murphy, L. (1974). The individual child. Department of Health, Education, and welfare, publi-
cation no. OCD 74–1032. Washington, DC: U.A. Government Printing Office.
NICHD Early Child Care Research Network. (1998). Early child care and self-control, com-
pliance and problem behavior at twenty-four and thirty-six months. Child Development,
69 (3), 1145–1170.
NICHD Early Child Care Research Network. (1999, April). Effect sizes from the NICHD Study
of Early Child Care. Paper presented at the Biennial Meeting of the Society for Research
in Child Development, Albuquerque, NM.
NICHD Early Child Care Research Network. (2000). The relation of child care to cognitive
and language development. Child Development, 71 (4), 960–980.
Osterling, J., & Dawson, G. (1994). Early recognition of children with autism: a study of first
birthday home videotapes. Journal of Autism and Developmental Disorders, 24(3), 247–257.
Peisner-Feinberg, E. S., Burchinal, M. R., Clifford, R. M., Culkin, M. L., Howes, C., Kagan, S.
L., Yazejian, N., Byler, P., Rustici, J., and Zelazo, J. 1999. The Children of the Cost, Quality,
and Outcomes Study Go to School: Technical Report.
Piaget, J. (1962). The stages of intellectual development of the child. In S. Harrison & J.
McDermott (Eds). Childhood Psychopathology. New York: International Universities Press.
Piaget, J. (1981/1954). Intelligence and affectivity: Their relationship during child development. Palo
Alto, CA, Annual Reviews.
Pinker, S. (1994). The language instinct. New York: William Morrow.

Rakic, P., Bourgeois, J., & Goldman-Rakic, P. (1994). Synaptic development of the cerebral
cortex: Implications for learning, memory, and mental illness. In J. van Pelt, M. A.
Corner, H.B. M. Uylings, & F. H. Lopes da Silver (Eds) The self-organizing brain: From
growth cones to functional networks. Pp. 227–243. New York: Elsevier Science.
Rappaport, D. (1960). The structure of psychoanalytic theory: A systematizing attempt. New York:
International University Press.
Sameroff, A., Seifer, R., Barocas, R., Zax, M., Greenspan, S. I. (1986). IQ scores of 4-year-old
children: Social-environmental risk factors. Pediatrics, 29, 343–350.
Sameroff, A., Seifer, A., Baldwin, A., & Baldwin, C. (1993). Stability of intelligence from pre-
school to adolescence: The influence of social and family risk factors. Child Development
64, 80–97.
Sameroff, A., Seifer, R., Barocas, R., Zax, M., Greenspan, S. I. (1986). IQ scores of 4-year-old
children: Social-environmental risk factors. Pediatrics, 29, 343–350.
Schanberg, S. M. & Field, T. M. (1987). Sensory deprivation stress and supplemental stimula-
tion in the rat pup and preterm human neonate. Child Development 58: 1431–47.
Shaffer, R. J., Jacokes, L. E., Cassily, J. F., & Greenspan, S. I. (2001). Effect of Interactive
Metronome(r) IM Training on Children with ADHD. American Journal of Occupational
Therapy, March.
Shonkoff, J. P. & Phillips, D. A. (eds.) (2000). From Neurons to Neighborhoods: The Science of Early
Childhood Development. National Research Council, Committee on Integrating the
Science of Early Childhood Development.
Singer, W. (1986). Neuronal activity as a shaping factor in postnatal development of visual cor-
tex. In W. T. Greenough and J. M. Juraska (eds) Developmental Neuropsychobiology,
271–293. Orlando: Academic Press.
Spelke, E. (1976). Infants’ Intermodal Perception of Events. Cognitive Psychology, 8, 553–60.
Sperry, R.W. (1985). Consciousness, personal identity, and the divided brain. In Frank Benson,
M.D. & Eric Zaidel, Ph.D. (Eds.) The Dual Brain. 11–27. New York: Guilford Press.
Spitz, R. (1945). Hospitalism. Psychoanalytic study of the child, 1, 53.
Tallal, P., Miller, S. L., Bedi, G., Byma, G., Wang, X., Nagarajan, S.S., Schreiner, C., Jenkins,
W. M., Merzenich, M.M. (1996). Language comprehension in language-learning
impaired children improved with acoustically modified speech. Science, 271(5245):81–4.
Tanguay, P. E. (1999). The diagnostic assessment of autism using social communication
domains. Paper presented at the Interdisciplinary Council on Developmental and
Learning Disorders’ Third Annual International Conference, Autism and Disorders of
Relating and Communicating. November 12–14, McLean, VA.
Tanguay, P. E., Robertson, J., & Derrick, A. (1998). A dimensional classification of autism spec-
trum disorder by social communication domains. Journal of the American Academy of Child
and Adolescent Psychiatry, 37(3):271.
Thinus-Blanc, C. (1981). Volume discrimination learning in golden hamsters: Effects of the
structure of complex rearing cages. Developmental Psychobiology 14: 397–403.
Turner, A. M. & Greenough, W. T. (1983). Synapses per neuron and synaptic dimensions in
occipital cortex of rats reared in complex, social, or isolation housing. Acta Stereologica 2,
suppl. 1: 239–244.
Turner, A. M. & Greenough, W. T. (1985). Differential rearing effects on rat visual cortex
synapses. I: Synaptic and neuronal density and synapses per neuron. Brain Research 329:
Vandell, D. L., & Wolfe, B. (2000). Child Care Quality: Does It Matter and Does It Need to be
Improved? SR 78. Institute for Research on Poverty, University of Wisconsin-Madison.
Weiler, I.J., Hawrylak, N. & Greenough, W.T. (1995). Morphogenesis in Memory Formation:
Synaptic and Cellular Mechanisms,” Behavioural Brain Research. 66:1–6.
Wiesel, T. N. & Hubel, D. H. (1963). Single-cell responses in striate cortex of kittens deprived
of vision in one eye. Journal of Neurophysiology 26: 1003–17.

Werner, H., & Kaplan, B. (1963). Symbol formation. New York: Wiley.
Wetherby, A. M., & Prizant, B. M. (1993). Profiling communication and symbolic abilities in
young children. Journal of Childhood Communication Disorders, 15, 23–32.
Winnicott, D.W. (1931). Clinical notes on disorders of childhood. London: Heineman.
Young, T. T. (1943). Emotions in man and animal. New York: Wiley.
Zimmerman, A., & Gordon, B (2000). Neuromechanisms in autism. In The Interdisciplinary
Council on Developmental and Learning Disorders (ICDL) Clinical Practice Guidelines:
Redefining the Standards of Care for Infants, Children, and Families with Special Needs.
Bethesda, MD: ICDL Press.

Mailing Address:
Stanley I. Greenspan, M.D.
7901 Glenbrook Road
Bethesda, Maryland 20814-4292
Are Intuitive Physics and Intuitive
Psychology Independent?

Simon Baron-Cohen, Ph.D., Sally Wheelwright, M.A.,

Amanda Spong, M.A., Victoria Scahill, M.A., and John
Lawson, M.A.

Abstract. According to the framework of evolutionary psychology, the human mind

should be considered in terms of its evolved adaptedness to the environment (Karmiloff-
Smith, Grant, Bellugi & Baron-Cohen, 1995). Two postulated neurocognitive adaptations
are intuitive (or folk) psychology, for inferring social causality; and intuitive (or folk)
physics, for inferring physical causality. In this paper we test these two aspects of our causal
cognition in children with Asperger Syndrome (AS). To do this, we employ new tests of intu-
itive physics and intuitive psychology. Results show that children with AS are impaired in
folk psychology whilst being superior in folk physics. Future work needs to test if intuitive
psychology and physics are truly independent of one another (implying separate underlying
mechanisms) or are inversely related to one another (implying a single underlying
mechanism for both).

The Evolutionary Framework

The model guiding this study holds that there are specialized neurocognitive
mechanisms that have evolved to enable rapid discrimination of two classes of
entity: agents vs. non-agents (Baron-Cohen, 1994; Leslie, 1995; Premack, 1990).
This follows from the classical view that in this universe there are only two

Acknowledgments: SBC and SW were supported by the MRC and the McDonnell-Pew Foundation
during the period of this work. VS was supported by the Anglia and Oxford NHS. JL was supported
by the Isaac Newton Trust. Steve Cooter, Rachel Treadaway, Becky Hannon, and Nic Alexander car-
ried out invaluable pilot studies for this study, as part of their final year undergraduate project, for
which we are grateful. We are also indebted to Southlands School, Hampshire, and Wolverhampton
Grammar School for their co-operation. Special thanks to Maggi Rigg, Ann Wakeling, and Tony
Bennett for opening their schools to us.


kinds of entity: those that have intentionality, and those that do not (Brentano,
1874/1970)1. The animate-inanimate distinction doesn’t quite cover the inten-
tional/non-intentional distinction in that plants are of course alive, so the distinc-
tion is better covered by the concept of agency (Premack, 1990). Agents have
intentionality, whereas non-agents do not. This also means that when agents and
non-agents move, their motion has different causes (Csibra, Gergely, Biro, Koos &
Brockbanck, 1999; Gelman & Hirschfield, 1994). Agents can move by self-propul-
sion, which we naturally interpret as driven by their goals and desires, whilst non-
agents can reliably be expected not to move unless acted upon by another object
(e.g., following a collision).
We assume that the task for hominids as information-processors, over millions
of years, has been to compute the causes of these two distinct classes of motion
rapidly, since to fail to do so would be self-evidently maladaptive. Dennett’s claim
is that humans from infancy onwards use folk (or intuitive) psychology to deduce
the cause of an agent’s actions, and use folk (or intuitive) physics 2 to deduce the
cause of a non-agent’s movement (Dennett, 1987). Thus, if we see a rock rolling
down the hill, and an agent is present then the event could be interpreted as hav-
ing been caused by an intention (e.g. to throw it, roll it, kick it, etc.,). If no agent
is present, the event could be interpreted in terms of a physical causal force (e.g. it
was hit by another object, gravity, etc.,). Sperber, et. al., suggest that humans alone
have the reflective capacity to be concerned about causality, and that “causal cog-
nition” broadly falls into at least these two types (Sperber, Premack & Premack,
It may be that aptitudes in folk psychology and folk physics are independent of
one another, or are inversely related to one another. The experiments reported
below are relevant to these possibilities, since they investigate if some individuals are
impaired in their folk psychology but not in their folk physics, and if others show the
opposite pattern.
Folk psychology and folk physics may also be under some degree of genetic con-
trol. That is, they may comprise modules in a minimally innate sense (Baron-Cohen,
1999). One way one can test if such mechanisms are under some degree of genetic
control is by testing for dissociations in individuals who are known to have a genetic
disability. In the studies to be reported, we take this approach by investigating folk
psychology and folk physics in children with Asperger Syndrome (AS). Before turn-
ing to AS, it is necessary to define folk psychology and folk physics, and to consider
the evidence from normal development which points to these being “core domains
of cognition” (Wellman & Inagaki, 1997).

Intentionality is defined as the capacity of something to refer or point to things other than itself. A
rock cannot point to anything. It just is. In contrast, a mouse can “look” at a piece of cheese, and can
“want” the piece of cheese.
We use the terms “folk psychology” “intuitive psychology”, and “theory of mind” interchangeably.
We also intend the terms “folk physics” and “intuitive physics” to be interchangeable.

Defining Folk Psychology and Folk Physics

We define folk psychology as comprising both low-level social perception, and

higher-level social intelligence. Low-level here broadly refers to skills present in
human infancy ( Johnson, 2000). These include being able to judge (a) if something
is an agent or not (Premack, 1990); (b) if another agent is looking at you or not
(Baron-Cohen, 1994); (c) if an agent is expressing a basic emotion (Ekman, 1992),
and if so, what type. It also includes (d) engaging in shared attention, for example
by following gaze or pointing gestures (Mundy & Crowson, 1997; Scaife & Bruner,
1975; Tomasello, 1988); (e) showing concern or basic empathy at another’s distress,
or responding appropriately to another’s basic emotional state (Yirmiya, Sigman,
Kasari & Mundy, 1992); (f) being able to judge an agent’s goal or basic intention
(Premack, 1990). Higher-level here refers to skills present from early childhood and
which continue to develop throughout the lifespan. These include the following: (i)
Attribution of the range of mental states to oneself and others, including pretence,
deception, belief (Leslie, 1987). (ii) Being able to recognize and respond appropri-
ately to complex emotions, not just basic ones (Harris, Johnson, Hutton, Andrews &
Cooke, 1989). (iii) Being able to link such mind-reading to action, including lan-
guage, and therefore to understand and produce pragmatically appropriate lan-
guage (Tager-Flusberg, 1993). (iv) Using mind-reading not only to make sense of
others’ behaviour, but also to predict it, and even manipulate it (Whiten, 1991). (v)
Our sense of what is appropriate in different social contexts, based on what others
will think of our own behaviour. (v) Empathic understanding of another mind. In
short, it includes the skills that are involved in normal reciprocal social relationships
(including intimate ones) and in communication. We recognise that we have defined
folk psychology broadly, such that it is unlikely to hinge on a single cognitive
process. However, we argue that the domain is quite focused and narrowly defined,
namely, understanding social causality.
We define folk physics as comprising both low-level perception of physical
causality, and higher-level understanding of physical-causality. Low-level here
refers broadly to skills present in human infancy, such as the perception of physi-
cal causality (Leslie & Keeble, 1987) and expectations concerning the motion and
properties of physical objects. Higher-level here refers to skills present from early
childhood and which continue to develop throughout the lifespan. These include
concepts relating to mechanics (Karmiloff-Smith, 1992). Like folk psychology, folk
physics is unlikely to hinge on a single cognitive process. However, like folk psy-
chology, we argue that the domain is quite focused and narrowly defined, namely,
understanding how things work.
Both folk physics and folk psychology have been proposed as “core domains of
human cognition” because they share seven features (Carey, 1985; Gelman &
Hirschfield, 1994; Sperber et al., 1995; Wellman & Inagaki, 1997). Both domains
(1) are aspects of our causal cognition, (2) demonstrate precocity in human infancy,
(3) are acquired or develop universally, (4) show little if any cultural variability,
(5) have a specific but universal ontogenesis, (6) are adaptive, and (7) may be open to

neurological dissociation. The first of these features (causal cognition) is definitional:

folk psychology involves searching for the mental or intentional causes behind agen-
tive events, whilst folk physics involves searching for the physical causes of non-agen-
tive event. Evidence for features 2–5 applying to folk physics and folk psychology
comes from studies in developmental psychology, reviewed in the next section.
The sixth feature (adaptiveness) may not be directly testable but has inherent
plausibility and we imagine is non-contentious. The final feature (neurological disso-
ciability) is tested with respect to these two domains in the experiment reported later
in this paper.

Developmental Evidence

Folk psychology appears to be present from at least 12 months of age (Baron-

Cohen, 1994; Premack, 1990). Thus, infants show dishabituation to actions of
“agents” who appear to violate goal-directedness (Gergely, Nadasdy, Gergely &
Biro, 1995; Rochat, Morgan & Carpenter, 1997). They also expect agents to “emote”
(express emotion), and expect this to be consistent across modalities (between face
and voice) (Walker, 1982). They are also highly sensitive to where another person is
looking, and by 14 months will strive to establish joint attention (Butterworth, 1991;
Hood, Willen & Driver, 1997; Scaife & Bruner, 1975). By 14 months they also start
to produce and understand pretence (Bates, Benigni, Bretherton, Camaioni &
Volterra, 1979; Leslie, 1987). By 18 months they begin to show concern at the dis-
tress of others (Yirmiya et al., 1992). By 2 years old they begin to use mental state
words in their speech (Wellman & Bartsch, 1988). By 3 years old they can under-
stand relationships between mental states such as seeing leads to knowing (Pratt &
Bryant, 1990). By 4 years old they can understand that people can hold false beliefs
(Wimmer & Perner, 1983). By 5–6 years old they can understand that people can
hold beliefs about beliefs (Perner & Wimmer, 1985). By 7 years old they begin to
understand what not to say in order to avoid offending others (Baron-Cohen,
O’Riordan, Stone, Jones & Plaisted, 1999a). With age, mental state attribution
becomes increasingly more complex (Baron-Cohen, Jolliffe, Mortimore &
Robertson, 1997a; Happe, 1993).
Folk physics is also present very early in human ontogeny as manifested in the
infant’s sensitivity to apparent violations of the laws of physics. Thus, infants show
dishabituation to the unexpected events of larger objects going into smaller ones,
objects being unsupported, two objects occupying the same space, one object pass-
ing through another, or one inanimate object moving without being touched by
another (Baillargeon, Kotovsky & Needham, 1995; Leslie & Keeble, 1987; Spelke,
Phillips & Woodward, 1995). That is, even in infancy, humans appear to be sensitive
to physical causality. With age, children’s understanding of mechanics grows
(Karmiloff-Smith, 1992), but again the precocity of folk physics argues strongly for its
status as a core domain. The little cross-cultural evidence that exists suggests a simi-
lar picture in very different cultures (Avis & Harris, 1991).

These data have been interpreted in terms of two innate, independent modules
being part of the infant cognitive architecture: a theory of mind mechanism (ToMM)
and a theory of bodies mechanism (ToBy) (Leslie, 1995). A theory of mind takes sev-
eral years to develop, but a more restricted Intentionality Detector (or ID) (Baron-
Cohen, 1994; Premack, 1990) may be part of the starting state of our causal cognition
in infancy. The studies reported in this paper are not concerned with whether folk
psychology and folk physics are modular systems; rather, they are concerned with
these two core domains of cognition in children with Asperger Syndrome. Asperger
Syndrome is conceptualised as a variant on the autistic spectrum. This group of chil-
dren are chosen for two reasons: previous work suggests that a dissociation between
these core domains might characterise them; and that this might occur for genetic
and neurodevelopmental reasons. We elaborate on these points next.

Asperger Syndrome

Asperger Syndrome (AS) was first described by Asperger (Asperger, 1944), and the
descriptions of the children he documented overlapped considerably with the accounts
of childhood autism (Kanner, 1943). Little was published on AS in English until rela-
tively recently (Frith, 1991; Wing, 1981). Current diagnostic practice recognises AS as
meeting the same criteria for autism but with no history of language or communication
delay, and with no cognitive delay (APA, 1994; ICD-10, 1994). Although some studies
have claimed a distinction between AS and high-functioning autism (HFA) (Klin,
Volkmar, Sparrow, Cicchetti & Rourke, 1995), the majority of studies have not demon-
strated any significant differences between these. For this reason we use the term AS
(for present purposes) as overlapping with HFA. In the experiment reported below,
we test folk psychology and folk physics in these children. But first, why should one
suspect a dissociation between these two domains will be found in such individuals?
Since the first test of folk psychology in children with autism (Baron-Cohen, Leslie
& Frith, 1985), there have been more than 30 experimental tests, the vast majority
revealing profound impairments in the development of their folk psychological under-
standing. These are reviewed elsewhere (Baron-Cohen, 1995; Baron-Cohen, Tager-
Flusberg & Cohen, 1993) but include deficits in: joint attention (Baron-Cohen, 1989d;
Sigman, Mundy, Ungerer & Sherman, 1986); use of mental state terms in language
(Tager-Flusberg, 1993); production and comprehension of pretence (Baron-Cohen,
1987; Wing & Gould, 1979); understanding that “seeing-leads-to-knowing” (Baron-
Cohen & Goodhart, 1994; Leslie & Frith, 1988); distinguishing mental from physical
entities (Baron-Cohen, 1989a; Ozonoff, Pennington & Rogers, 1990); making the
appearance-reality distinction (Baron-Cohen, 1989a); understanding false belief
(Baron-Cohen et al., 1985); understanding beliefs about beliefs (Baron-Cohen,
1989b); and understanding complex emotions (Baron-Cohen, 1991). Some adults with
AS only show their deficits on age-appropriate adult tests of folk psychology (Baron-
Cohen et al., 1997a; Baron-Cohen, Wheelwright & Jolliffe, 1997b). This deficit in their
folk psychology is thought to underlie the difficulties such children have in social and

communicative development (Baron-Cohen, 1988; Tager-Flusberg, 1993), and the

development of imagination (Baron-Cohen, 1987; Leslie, 1987).
The above evidence points to an impairment in folk psychology, but it says noth-
ing about how circumscribed this is. Does it leave their folk physics intact? Or might
their folk physics even be super-developed? We predicted the latter, for reasons
explained next.

Autism and Folk Physics

If children with autism had an impairment in their folk physics, this might sug-
gest that the cause of their problems in the intentional domain was a problem in “the-
ory-building” per se (Carey, 1985). However, there are reasons to suspect that not
only is their folk physics intact but may even be superior, relative to normally devel-
oping children.
First, there is no shortage of clinical descriptions of children with autism being
fascinated by machines (the paragon of non-intentional systems). One of the earliest
clinical accounts was by Bettelheim (Bettelheim, 1968) who describes the case of
“Joey, the mechanical boy”. This child with autism was obsessed with drawing pic-
tures of machines (both real and fictitious), and with explaining his own behaviour
and that of others in purely mechanical terms. On the face of it, this would suggest
he had a well-developed folk-physics. The clinical literature reveals hundreds of
cases of children obsessed by machines. Parents’ accounts (Hart, 1989; Lovell, 1978;
Park, 1967) are a rich source of such descriptions. Indeed, it is hard to find a clinical
account of autism that does not involve the child being obsessed by some machine or
another. Typical examples include extreme fascinations with electricity pylons, bur-
glar alarms, vacuum cleaners, washing machines, video players, trains, planes, and
clocks. Sometimes the machine that is the object of the child’s obsession is quite sim-
ple (e.g., the workings of drainpipes, or the design of windows, etc.,).
Of course, a fascination with machines need not necessarily imply that the child
understands the machine, but in fact most of these anecdotes also reveal that children
with autism have a precocious understanding too. The child (with enough language,
such as is seen in children with AS) may be described as holding forth, like a “little
professor”, on their favourite subject or area of expertise, often failing to detect that
their listener may have long since become bored of hearing more on the subject. The
apparently precocious mechanical understanding, whilst being relatively oblivious to
their listener’s level of interest, suggests that their folk physics might be outstripping
their folk psychology in development. The anecdotal evidence includes not just an
obsession with machines, but with other kinds of physical systems. Examples include
obsessions with the weather (meteorology), the formation of mountains (geography),
motion of the planets (astronomy), and the classification of lizards (taxonomy).
Clinical/anecdotal evidence must however be left to one side, as this may not
prove anything. More convincing is that experimental studies converge on the same
conclusion: children with autism not only have an intact folk physics, they have

accelerated or superior development in this domain (relative to their folk psychology

and relative to their mental age, both verbal and nonverbal). First, using a picture
sequencing paradigm, children with autism performed significantly better than men-
tal-age matched controls in sequencing physical-causal stories (Baron-Cohen, Leslie
& Frith, 1986). The children with autism also produced more physical-causal justifi-
cations in their verbal accounts of the picture sequences they made, compared to
intentional accounts3.
Second, two studies found children with autism showed good understanding of a
camera (Leekam & Perner, 1991; Leslie & Thaiss, 1992). In these studies, the child is
shown a scene where an object is located in one position (A). The child is encour-
aged to take a photo of this scene, using a Polaroid camera. Whilst the experimenter
and the child are waiting for the photo to develop, the scene is changed: the object
is now moved to a new position (B). The experimenter then turns to the child and
asks where in the photo the object will be. These studies found that children with
autism could accurately infer what would be depicted in a photograph, even though
the photograph was at odds with the current visual scene. Again, this contrasted with
their poor performance on False Belief tests.
These “false photo” tasks (Zaitchik, 1990) closely parallel the structure of the false
belief task. The key difference is that in the (folk psychological) false belief test, a
person sees the scene, and then the object is moved from A to B whilst that person is
absent. Hence the person holds a belief that is at odds with the current visual scene.
In the false photo task a camera records the scene, and then the object is moved from
A to B whilst the camera is not in use. Hence the camera contains a picture that is at
odds with the current visual scene. The pattern of results by the children with autism
on these two tests was interpreted as showing that whilst their understanding of men-
tal representations was impaired, their understanding of physical representations was
not. This pattern has been found in other domains (Charman & Baron-Cohen, 1992;
Charman & Baron-Cohen, 1995). But the False Photo Test is also evidence of their
folk physics outstripping their folk psychology and being superior to mental age
(MA) matched controls.
Family studies add to this picture. Parents of children with Asperger Syndrome
(AS) also show mild but significant deficits on an adult folk psychology task (the adult
version of the “Reading the Mind in the Eyes” task). This mirrors the deficit in folk
psychology seen in patients with autism or AS (Baron-Cohen & Hammer, 1997). This
familial resemblance at the cognitive level is assumed to reflect genetic factors, since
autism and AS appear to have a strong heritable component (Bailey et al., 1995;
Bolton et al., 1994; Folstein & Rutter, 1977; Le Couteur et al., 1996). One should
also expect that parents of children with autism or AS to be over-represented in
occupations in which possession of superior folk physics is an advantage, whilst a
deficit in folk psychology would not necessarily be a disadvantage. The paradigm
occupation for such a cognitive profile is engineering.

This study however did not involve a chronological age (CA) matched control group, so the appar-
ent superiority in folk physics in autism may simply have reflected their higher CA.

A recent study of 1000 families found that fathers and grandfathers (patri- and
matrilineal) of children with autism or AS were more than twice as likely to work in
the field of engineering, compared to fathers and grandfathers of children with other
disabilities (Baron-Cohen, Wheelwright, Stott, Bolton & Goodyer, 1997c). Indeed,
28.4% of children with autism or AS had at least one relative (father and/or grandfa-
ther) who was an engineer. Related evidence comes from a survey of students at
Cambridge University, studying either sciences (physics, engineering, or maths) or
humanities (English or French literature). When asked about family history of a range
of psychiatric conditions (schizophrenia, anorexia, autism, Down’s Syndrome, or
manic depression), the students in the science group showed a six-fold increase in the
rate of autism in their families, and this was specific to autism (Baron-Cohen et al.,
This raises the possibility that the cognitive phenotype of autism spectrum
conditions may involve superiority in folk physics alongside a relative deficit in folk
psychology, relative of course to MA. In this paper we report an experimental test of
the prediction that children with AS will have superior folk physics in the presence
of impaired folk psychology.

The Experiment


We tested 2 groups of subjects.

Group 1 comprised 15 children (all male) with a clear diagnosis of Asperger
Syndrome (AS), defined according to internationally recognised, established criteria
(APA, 1994; ICD-10, 1994). They were all attending a special school for Asperger
Syndrome (the only one in the UK). This is a residential provision, reflecting the sever-
ity of their symptoms and the disruptive impact these had had on their previous school-
ing and on their family life. They were aged between 8 and 14 years of age, and were
all of at least average intelligence. They had all received an IQ test within the previous
2 years, with a standard instrument (WISC-R). Their ages and IQ data are shown in
Table 1. They had a range of obsessional interests, consistent with their diagnosis, and
these included military tanks, explosives, the periodic table, historical dates, football,
electricity, relativity, vehicles, and machines. They had all received a special educational
needs statement specifying that they needed special provision as a result of their AS.
Group 2 were pupils attending state primary and secondary schools in
Cambridge and Wolverhampton, selected at random. They comprised n⫽63 male
and n⫽40 female pupils, age range 12–13 (see Table 1), and n⫽53 children, age
6–10 yrs (see Table 2). The children in the 12–13 yr age group (in Table 1) were
given the Folk Physics Test, whilst the children in the 6-10 yr age group (in Table 2)
were given the Folk Psychology Test (see below). None had been statemented for
having special educational needs. They were not given any IQ test because of time

Subjects Participating in the Folk Physics Test.



GROUP x (sd) x (sd) x (sd) x (sd)

AS (n⫽15 males) 13.35 (1.18) 102.4 (10.1) 93.6 (8.8) 96.9 (9.9)
Males (n⫽63) 12.4 (0.3) –
Females (n⫽40) 12.8 (0.3) –
VIQ⫽Verbal IQ
PIQ⫽Performance IQ

Subjects Participating in the Folk Psychology Test.



GROUP x (sd) x (sd) x (sd) x (sd)

AS (n⫽15 males) 13.35 (1.18) 102.4 (10.1) 93.6 (8.8) 96.9 (9.9)
Normals (n⫽53)
6-8 yr. olds
Males (n⫽9) 7.3 (0.7)
Females (n⫽11) 6.8 (0.6)
8-10 yr. olds
Males (n⫽8) 8.9 (0.6)
Females (n⫽6) 9.0 (0.3)
10-12 yr. olds
Males (n⫽9) 11.0 (0.5)
Females (n⫽10) 10.7 (0.6)

constraints but were assumed to have normal intelligence by virtue of their educa-
tional placement.


The children with AS were all given two tests:

1. The Folk Physics Test (see full test in Appendix A): This comprises 20 ques-
tions drawn from a variety of sources, with multiple choice format. This was piloted
with a range of age groups of normal subjects, revealing meaningful results only above
age 10 years. We consider it as a test of folk physics for two reasons. (a) All the prob-
lems could be solved from everyday real world experience of the physical-causal
world. (b) The teachers of physics in the schools where our subjects were studying
confirmed that these problems had not been taught as part of any school curriculum.

Since the folk physics task was visual, as a control test for perceptual processing the
children with AS were also give the Raven’s Coloured Matrices (Raven, 1956).
2. The Folk Psychology Test: this comprised the children’s version of the Reading
the Mind in the Eyes Test, adapted from the adult version (Baron-Cohen et al., 1997a,
2001). Examples are shown in Appendix B. This comprises 28 photographs of the
eye region of the face. The subject is asked to pick which of 4 words best describes
what the person in the photo is thinking or feeling. The test is the result of piloting
with normal children. 3 of the 4 words are foil mental state terms, and the other word
is deemed “correct”. (See below for how “correct” was established). Position of the 4
words are randomised for each item. This is not simply a complex test of emotion
recognition (although it is this in part) because the mental state words included both
affective and non-affective (cognitive) mental state terms.
The Eyes task included a control for non-mentalistic social intelligence: the children
were asked to judge the person in the photo’s gender, from their eyes alone. Mental state
words were not displayed on this control task, and instead the words “male” and
“female” appeared as a forced choice. This latter control test was given in full to the chil-
dren with AS, but because of limited testing time with the normal children, just a ran-
dom selection of 8 items were given to the normal subjects, all of whom performed at
ceiling on this control test. While we do not claim that the Gender Recognition task is
matched for complexity with the mentalizing condition, it nevertheless involves a (non-
mentalistic) social judgement from the eyes, and attention to the relevant stimuli.
Regarding children in Group 2 (normal controls), only children over 12 were
given the folk physics test, as piloting showed that prior to this age normal perform-
ance is poor. Thus, only children in the age groups 6-10 were given the Eyes test, in
order to obtain normative developmental data on this test from this age range.


Folk Physics Test

A one-way ANOVA was used to compare the AS group with the control males
and females. There was a significant difference between the groups, F(2, 117)⫽30.4,
p⬍0.0001. Post hoc Student Newman-Keuls tests, with significance set at p⬍0.05,
indicated that the AS group performed significantly better than both the control
males and the control females, who did not differ from each other. See Table 3 for

Results of the Folk Physics Test: Means
and Standard Deviations

GROUP x (sd)

AS (n⫽15 males) 16.3 (3.1)

Males (n⫽63) 10.6 (2.8)
Females (n⫽40) 9.9 (2.8)

Folk Physics Test:
Item Analysis (Normal group only)


1 60.4
2 76.4
3 73.6
4 16.0
5 73.6
6 89.6
7 46.2
8 18.9
9 69.8
10 50.0
11 83.0
12 67.9
13 25.5
14 61.3
15 13.2
16 44.3
17 42.5
18 19.8
19 39.6
20 65.1

results. An item analysis was carried out. This did not reveal any specific item was
failed significantly more often than any other item, in the AS group. In the normal
group, there was more variability. (Table 4 shows the item analysis for the normal
group.) On the Raven’s Coloured Matrices, the AS group performed in line with (but
not significantly above) their mental age (scores, x⫽30, sd⫽4).
Folk Psychology Test
The test was first developed to parallel the (revised) Adult Version4 with 36 items,
but using child-level vocabulary. It was piloted on a small group of normal children
(n⫽6) age 8-12, to identify candidate target and foil words. In the larger sample tested
here (n⫽53), all 36 items were screened in two ways. (a) By checking that in all cases
the majority of normal subjects (more than 50%) in the 10-12 yr. age group (n⫽19)
identified the target word as correct (more than 10 children out of 19 identifying this
word). (b) By checking that the second word most commonly identified was chosen
by no more than a third of normal subjects in this age group (i.e. no more than 5 chil-
dren). This was true for all items analysed below (i.e., for 28 items). The remaining 8

The published version of this test (Baron-Cohen et al., 1997a) has 25 items. The revised version of
this test improves on this by having 36 items. “(Baron-Cohen et al, 2001).”

items failed to meet these criteria and were therefore dropped from any further
analysis. Table 5 shows the item analysis for these remaining 28 items.
In a test of 28 items, with 4 response options, scoring 9 or more out of 28 is above
chance (Binomial Test, p⬍0.05). 4 normal children in the 6-8 yr. age group (all girls)
failed to score above chance. In the AS group, 2/15 boys failed to score above
chance. A one-way ANOVA comparing performance of the AS group and the 3
groups of normal children on the Eyes Test was significant, F(3, 64)⫽16.0, p⬍0.0001.
Post hoc Student Newman-Keuls tests, with significance set at p⬍0.05, indicated that
the two oldest groups of normal children scored significantly higher than both the AS
group and the youngest group of normal children. This was the only significant dif-
ference found in the post hoc analysis. The AS group scored close to ceiling on the
Gender Judgement Control Test (x⫽25.3, sd⫽0.3).

Item Analysis from the Normal Group on the Folk Psychology Test
(Data from 10–12 Year Olds Only).
Group 3 Correct Answers Shown in Bold


1 26.3 0.0 57.9 15.8

2 0.0 5.3 5.3 89.5
3 63.2 5.3 0.0 31.6
4 0.0 100.0 0.0 0.0
5 36.8 63.2 0.0 0.0
6 5.3 0.0 63.2 31.6
7 10.5 15.8 68.4 5.3
8 78.9 0.0 0.0 21.1
9 5.3 0.0 0.0 94.7
10 21.1 21.1 57.9 0.0
11 0.0 84.2 0.0 15.8
12 10.5 0.0 0.0 89.5
13 84.2 15.8 0.0 0.0
14 5.3 94.7 0.0 0.0
15 52.6 5.3 21.1 21.1
16 57.9 0.0 15.8 26.3
17 5.3 0.0 5.3 89.5
18 89.5 5.3 5.3 0.0
19 5.3 31.6 5.3 57.9
20 5.3 0.0 94.7 0.0
21 52.6 0.0 26.3 21.1
22 0.0 0.0 5.3 94.7
23 21.1 68.4 0.0 10.5
24 68.4 15.8 15.8 0.0
25 15.8 15.8 10.5 57.9
26 10.5 5.3 63.2 21.1
27 31.6 0.0 52.6 15.8
28 5.3 0.0 73.7 21.1

Results of Folk Psychology Test: means and standard deviations.

GROUP x (sd) x (sd)

AS (n⫽15 males) 12.6 (3.3) 26.5 (2.3)

Normals (n⫽53)
6–8 yr. olds
Males (n⫽9) 14.6 (5.1)
Females (n⫽11) 12.5 (5.6)
8–10 yr. olds
Males (n⫽8) 18.1 (4.7)
Females (n⫽6) 17.7 (3.5)
10–12 yr. olds
Males (n⫽9) 20.2 (2.4)
Females (n⫽10) 21.0 (2.4)

Results are shown in Table 6. An Age by Gender ANOVA in the normal group
alone, on the Eyes Scores, found a significant main effect of Age (F(2,52)⫽13.9,
p⬍0.0001), with the youngest age group scoring significantly lower on the Eyes Test
compared to the older two age groups, who did not differ from each other (Student
Newman-Keuls Test, p⬍0.05). There was no effect of Gender (F(1, 52)⫽0.3, p⫽0.62),
and no Age by Gender interaction (F(2,52)⫽0.6, p⫽0.60).

Correlation Between Folk Physics and Folk Psychology in AS

A test of correlation between folk psychology and folk physics in the normal group
was not possible since different groups of children were given each of these tests. In the
group with AS, the two tasks were strongly inversely correlated (r⫽-0.63, p⫽0.001).


The experiments in this paper derive from the model that the human brain has
evolved at least two independent modes of causal cognition: folk psychology and folk
physics. In the extreme case, severe autism may be characterised by almost no folk
psychology (and thus “mindblindness”). Autism spectrum conditions come by
degrees, so different points on the autistic spectrum may involve degrees of deficit in
folk psychology (Baron-Cohen, 1995). We predicted that in those individuals who
have no accompanying mental handicap (i.e., whose intelligence is in the normal
range), the child’s folk physics would develop not only normally, but even at a superi-
or level. This was tested in a group of children with Asperger Syndrome (AS). This
prediction was confirmed: children with AS were functioning significantly above
their mental age (MA) in terms of folk physics, but significantly below their MA in
terms of folk psychology.

The control tasks in this experiment enable us to conclude that children with AS
are not superior in all visuo-spatial tasks (since they were normal but not above aver-
age on the Raven’s Matrices). On the Eyes Task they were clearly attending to the
eyes well enough to judge gender. This pattern of results suggests their understand-
ing in folk physics may represent an islet of ability; and that their difficulties on the
Eyes Task may be specifically linked to mind-reading.
We take seriously the notion that this profile in AS (impaired folk psychology,
together with superior folk physics) might be partly the result of a genetic liability.
This is because AS appears to be heritable (Gillberg, 1991), and because there is every
reason to expect that individuals with such a cognitive profile could have been
selected for in hominid evolution. Good folk physics would have conferred important
advantages to an individual’s inclusive fitness (e.g., tool use, hunting skills, construc-
tion skills, etc.,), even if that individual’s folk psychology skills were less proficient.
Note that a genetic factor could operate in at least two different ways: (a) An indi-
vidual might have a genetically-based impairment in folk psychology; or (b) a genet-
ically-based talent for folk physics. This second alternative derives from the idea of
an independent module for folk physics (Leslie, 1995). It is possible that in autism
spectrum conditions we see the twin genetic anomalies of impaired folk psychology
co-occurring with superior folk physics. But whether the present results reflect (a) or
(b) above, or both, such genotypes would lead the individual to spend less time inter-
acting with the social environment, and more time interacting with the physical enviro-
nment, since he or she would understand the latter better. A gene-environment
interaction could then explain why such a brain, developing along an abnormally
one-sided trajectory, would lead to a superiority in folk physics.
What is the extra explanatory scope of documenting superior folk physics in
autism spectrum conditions, over and above the (now standard) demonstration of a
theory of mind deficit in autism? The theory of mind account has been virtually silent
on why such children should show “repetitive behaviour”, a strong desire for rou-
tines, and a “need for sameness”. To date, the only cognitive account to attempt to
explain this aspect of the syndrome is the executive dysfunction theory (Ozonoff,
Rogers, Farnham & Pennington, 1994; Pennington et al., 1997; Russell, 1997). This
paints an essentially negative view of this behaviour, assuming that it is a form of
“frontal lobe” perseveration or inability to shift attention.
Whilst some forms of repetitive behaviour in autism, such as “stereotypies” (e.g.,
twiddling the fingers rapidly in peripheral vision) may be due to executive deficits,
the executive account has traditionally ignored the content of “repetitive behaviour”.
The current account draws attention to the fact that much repetitive behaviour
involves the child’s “obsessional”5 or strong interests with mechanical systems (such
as light switches or water faucets) or other systems that can be understood in physi-

Elsewhere (Baron-Cohen, 1989c) we review the argument for why the term “obsession” can only be
used in the context of autism with some qualifications. This centers on the traditional definition of an
obsession being “egodystonic” (or unwanted). In autism, there is no evidence that the child’s strong inter-
ests are unwanted. Rather, those individuals with autism or AS who can report on why they engage in
these activities report that they often derive some pleasure from them. They are therefore probably

cal-causal terms. Rather than these behaviours being a sign of executive dysfunction,
these may reflect the child’s intact or even superior development of their folk physics.
The child’s obsession with machines and systems, and what is often described as their
“need for sameness” in attempting to hold the environment constant, might be signs
of the child as a superior folk-physicist: conducting mini-experiments in his or her
surroundings, in an attempt to identify physical-causal principles underlying events.
Certainly, our recent study of obsessions suggests that these are not random with
respect to content (which would be predicted by the content-free executive dysfunc-
tion theory), but that these test to cluster in the domain of folk physics (Baron-Cohen
& Wheelwright, 1999).
In this paper we have not presented a task analysis of folk physics, and it could
be argued that the good folk physics skills seen here are simply an expression of an
anomaly previously documented, namely “weak” central coherence (Frith, 1989;
Happe, 1996). Weak central coherence refers to the individual’s preference for local
detail over global processing. This has been demonstrated in terms of an autistic
superiority on the Embedded Figures Task (EFT) and the Block Design Subtest
( Jolliffe & Baron-Cohen, 1997; Shah & Frith, 1983; Shah & Frith, 1993). Both of these
are interpreted as evidence of good segmentation skills, and superior attention to
detail. The latter has also been demonstrated on visual search tasks (Plaisted,
O’Riordan & Baron-Cohen, 1998a; Plaisted, O’Riordan & Baron-Cohen, 1998b).
The question is whether superior folk physics, like weak central coherence, might
simply reflect this superior attention to detail. This is a strong possibility, and merits
direct testing in the future. If confirmed, this would not invalidate the usefulness of
studying folk physics in autism spectrum conditions. Rather, it may show strong folk
physics as an upstream benefit of weak central coherence.

Developing a New Model

If folk psychology and folk physics are independent dimensions it is possible to

plot on orthogonal axes possible scores from possible tests assessing these two abili-
ties. Figure 1 provides a visual representation of this model of the relationship
between folk psychology and folk physics. It suggests appropriate labels for different
possible patterns of scores. The axes show number of standard deviations from the
mean. The scale of the diagram is less important than the principle underlying it.
We have used the terms Balanced Brain, Social Brain, Technical Brain, Extreme Social
Brain and Extreme Technical Brain as short-hand to describe these different possible
patterns of scores. The terms describe the discrepancy between the folk psychology
score and the folk physics score. In the Balanced Brain, there is no difference
between scores. In the Social Brain, folk psychology is one or two standard devia-
tions higher than folk physics. In the Extreme Social Brain, this discrepancy is greater
than two standard deviations. The same pattern is used for the Technical Brain and
the Extreme Technical Brain. In the Technical Brain, folk physics is one or two stan-
dard deviations higher than folk psychology whilst for the Extreme Technical Brain,
this discrepancy is greater than two standard deviations.

F IGURE 1. Folk psychology—folk physics dimensions.

It is worth underlining the fact that the key point is the discrepancy between the
scores rather than the absolute scores themselves. For example, someone could score
two standard deviations above the mean on folk psychology (a very high score) but
if they scored three standard deviations above the mean on folk physics, they would
be described as having the Technical Brain. Thus, the key issue is possible asymmet-
ries of ability.

Our ongoing work is testing this model. If folk physics and folk psychology are
truly independent, then it should be possible to find individuals in every square
within the diagram (no correlation). If there is some trade-off between abilities in folk
physics and folk psychology, then the majority of individuals should have results
which fit into the top left quadrant or bottom right quadrant of the diagram (negative
correlation). If folk physics and folk psychology abilities are subserved by the same
underlying system, then the majority of individuals should have results which fit into
the top right quadrant or bottom left quadrant (positive correlation). It is important
to clarify that we conceptualise folk physics and folk psychology will both vary with
mental age (MA). Therefore, standardised norms will need to be obtained for each
level of MA.
Evidence from sex difference research (Kimura, 1992) suggests that the Technical
Brain type is more commonly found in males whilst the Social Brain type is more fre-
quent in females. For this reason we can also use the terminology Female Brain and
Male Brain types as synonyms for the Social and Technical Brains, respectively. This
claim is also being tested as part of ongoing work, using a wider variety of tests and
Autism has been described as the extreme form of the male brain (Baron-Cohen
& Hammer, 1997). Figure 2 illustrates where we predict the vast majority of people
with autism will be located in this (MA-matched) framework. Although this area
overlaps with the Extreme Technical brain they are not exactly the same. This is
because we predict that people with autism will always score more than one standard
deviation below the mean on folk psychology and also that they will always score
more than one standard deviation above the mean folk physics. We tested this pre-
diction using the AS group data and control data from above. The scores for the AS
group on the Eyes test were transformed to z-scores using the mean and standard
deviation from the eldest group of control children (so the children were as closely
matched on MA as possible). All the scores from the control children who did the
folk physics test were used to produce the mean and standard deviation to standard-
ise the AS group on the folk physics test. Figure 3 shows the relationship between the
standardised folk physics and folks psychology scores in the AS group. Note that all
but one of the children appear in the bottom right quadrant. Obviously, the number
of subjects in this test is limited and in future studies, a battery of tests is likely to be
used to make up the folk physics and folk psychology scores. However, this first test
of the model is encouraging.
It is important to stress that this approach is in no way diagnostic. We do not
intend to imply that someone who scores in the Extreme Technical Brain area should
be diagnosed with autism. Rather, we are simply predicting that people with autism
are more likely to score in the Extreme Technical Brain area than in any other area.
Figure 2 also illustrates where the contrast case to autism is located. This area is
the exact opposite of the predicted autism area, overlapping with the Extreme Social
Brain, but not matching it exactly. Some people have speculated as to whether peo-
ple with Williams syndrome might have the Extreme Social Brain, (Karmiloff-Smith
et al., 1995), though this is debated (Tager-Flusberg, Boshart & Baron-Cohen, 1998).

F IGURE 2. The predicted location for autism spectrum conditions and the contrast

Note that in the same way that autism can be considered either from the perspective
of difficulties (folk psychology ) or strengths (folk physics), so can the contrast case.
In the latter case, the difficulties are predicted to be in folk physics (we could think
of this as technical-blindness) whilst the strengths are remarkable empathy. Such a
case is predicted by this model but has not yet been documented. Our ongoing work
will go out to test for such cases.

F IGURE 3. Distribution of standardised scores on the physics and eyes tests in the AS group.

Note that the two dimensions of folk psychology and folk physics are conceptu-
alised as independent of IQ. We certainly know of cases in individuals with Asperger
syndrome whose IQ is in the superior range but who perform in the low range on
empathy (folk psychology) tests (Baron-Cohen et al., 1999c). We speculate that whilst
IQ may influence folk physics ability to a greater extent, there may be individuals
whose folk physics is out of keeping with their IQ. The children with Asperger
syndrome in the above study are an example of this, since they were normal on
the Raven’s matrices.
This article has focused on folk physics and folk psychology, because they are
two forms of causal cognition. As has been discussed by others (Hatano & Inagaki,
1994; Sperber et al., 1995; Wellman, 1990), other universal cognitive domains may
also exist. The principal other candidates are folk mathematics (counting) and folk
biology (classification of the animate world into species, predators, prey, etc.). We
remain to be persuaded that these are independent domains, since it is plausible that
folk mathematics is simply part of folk physics, for example. However, in the same
way that a deficit in folk psychology should leave folk physics either unaffected or
superior in autism, the same arguments should lead to unaffected or superior develop-
ment of folk mathematics and folk biology in such individuals.
This model of the independence of folk physics and folk psychology (or social
and non-social intelligence) also predicts the existence of very high functioning indi-
viduals with AS, who may be extreme high achievers in domains such as mathemat-
ics and physics—equivalent to Nobel Prize winners even - but who have deficits in
folk psychology. Our recent case studies are beginning to identify such very high-
functioning individuals (Baron-Cohen et al., 1999c).
In conclusion, the present data suggest folk psychology is impaired in individuals
with AS, whilst their folk physics is superior. This is consistent with recent neuro-
logical reports of the effects of specific lesions to the amygdala causing specific
impairments in social perception (Damasio, Tranel & Damasio, 1990). It is of some
interest that using functional magnetic resonance imaging (fMRI) the normal brain
shows activation of the amygdala when performing the Eyes Task, whilst individuals
with AS show significantly reduced amygdala activity (Baron-Cohen et al., 1999b).
Whilst the brain basis of folk psychology is gradually being unravelled, the brain
basis of folk physics is as yet unknown.

Appendix A: The Folk Physics Test


Appendix B: Folk Physics Test Answers

1. c
2. c
3. b
4. b
5. c
6. a
7. a
8. d
9. b
10. a
11. b
12. a
13. a
14. a
15. c
16. a
17. c
18. a
19. d
20. c

Appendix C: Examples from the Children’s

Version of the Reading the Mind in the Eyes
(Folk Psychology) Test (Revised).

1. Female

Correct answer⫽sure about something

surprised sure about something

joking happy

Appendix C: Examples from the Children’s

Version of the Reading the Mind in the Eyes
(Folk Psychology) Test (Revised).

2. Male

Correct answer⫽Friendly

friendly sad

surprised worried


APA. (1994). DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th Edition.
Washington DC: American Psychiatric Association.
Asperger, H. (1944). Die “Autistischen Psychopathen” im Kindesalter. Archiv fur Psychiatrie und
Nervenkrankheiten, 117, 76–136.
Avis, J., & Harris, P. (1991). Belief-desire reasoning among Baka children: evidence for a uni-
versal conception of mind. Child Development, 62, 460–467.
Bailey, T., Le Couteur, A., Gottesman, I., Bolton, P., Simonoff, E., Yuzda, E., & Rutter, M.
(1995). Autism as a strongly genetic disorder: evidence from a British twin study.
Psychological Medicine, 25, 63–77.
Baillargeon, R., Kotovsky, L., & Needham, A. (1995). The acquisition of physical knowl-
edge in infancy. In D. Sperber, D. Premack, & A. Premack (Eds.), Causal Cognition:
A Multidisciplinary Debate,. Oxford: Oxford University Press.
Baron-Cohen, S. (1987). Autism and symbolic play. British Journal of Developmental Psychology, 5,
Baron-Cohen, S. (1988). Social and pragmatic deficits in autism: cognitive or affective? Journal
of Autism and Developmental Disorders, 18, 379–402.
Baron-Cohen, S. (1989a). Are autistic children behaviourists? An examination of their mental-
physical and appearance-reality distinctions. Journal of Autism and Developmental
Disorders, 19, 579–600.
Baron-Cohen, S. (1989b). The autistic child’s theory of mind: a case of specific developmen-
tal delay. Journal of Child Psychology and Psychiatry, 30, 285–298.
Baron-Cohen, S. (1989c). Do autistic children have obsessions and compulsions? British Journal
of Clinical Psychology, 28, 193–200.
Baron-Cohen, S. (1989d). Perceptual role-taking and protodeclarative pointing in autism.
British Journal of Developmental Psychology, 7, 113–127.
Baron-Cohen, S. (1991). Do people with autism understand what causes emotion? Child
Development, 62, 385–395.
Baron-Cohen, S. (1994). How to build a baby that can read minds: Cognitive mechanisms in
mindreading. Cahiers de Psychologie Cognitive/ Current Psychology of Cognition, 13, 513–552.
Baron-Cohen, S. (1995). Mindblindness: an essay on autism and theory of mind. Boston: MIT
Press/Bradford Books.
Baron-Cohen, S. (1999). Does the study of autism justify minimalist innate modularity?
Learning and Individual Differences, 10, 179–191.
Baron-Cohen, S., Bolton, P., Wheelwright, S., Short, L., Mead, G., Smith, A., & Scahill, V.
(1998). Autism occurs more often in families of physicists, engineers, and mathemati-
cians. Autism, 2, 296–301.
Baron-Cohen, S., & Goodhart, F. (1994). The “seeing leads to knowing” deficit in autism: the
Pratt and Bryant probe. British Journal of Developmental Psychology, 12, 397–402.
Baron-Cohen, S., & Hammer, J. (1997). Parents of children with Asperger Syndrome: what is
the cognitive phenotype? Journal of Cognitive Neuroscience, 9, 548–554.
Baron-Cohen, S., & Hammer, J. (1997). Is autism an extreme form of the male brain? Advances
in Infancy Research, 11, 193–217.
Baron-Cohen, S., Jolliffe, T., Mortimore, C., & Robertson, M. (1997a). Another advanced test
of theory of mind: evidence from very high functioning adults with autism or Asperger
Syndrome. Journal of Child Psychology and Psychiatry, 38, 813–822.
Baron-Cohen, S., Leslie, A. M., & Frith, U. (1985). Does the autistic child have a “theory of
mind”? Cognition, 21, 37–46.
Baron-Cohen, S., Leslie, A. M., & Frith, U. (1986). Mechanical, behavioural and Intentional
understanding of picture stories in autistic children. British Journal of Developmental
Psychology, 4, 113–125.

Baron-Cohen, S., O’Riordan, M., Stone, V., Jones, R., & Plaisted, K. (1999a). Recognition of
faux pas by normally developing children and children with Asperger Syndrome or
high-functioning autism. Journal of Autism and Developmental Disorders, 29, 407–418.
Baron-Cohen, S., Ring, H., Wheelwright, S., Bullmore, E., Brammer, M., Simmons, A., &
Williams, S. (1999b). Social intelligence in the normal and autistic brain: an fMRI study.
European Journal of Neuroscience, 11, 1891–1898.
Baron-Cohen, S., Tager-Flusberg, H., & Cohen, D. (Eds.). (1993). Understanding other minds:
perspectives from autism: Oxford University Press.
Baron-Cohen, S., & Wheelwright, S. (1999). Obsessions in children with autism or Asperger
Syndrome: a content analysis in terms of core domains of cognition. British Journal of
Psychiatry, 175, 484–490.
Baron-Cohen, S., Wheelwright, S., & Hill, J. (2001). The ?Reading the Mind in the Eyes? Test
Revised Version: A study with normal adults, and adults with Asperger syndrome or
High-Functioning Autism, Journal of Child Psychiatry and Psychiatry, 42, 241–252.
Baron-Cohen, S., Wheelwright, S., & Jolliffe, T. (1997b). Is there a “language of the eyes”?
Evidence from normal adults and adults with autism or Asperger syndrome. Visual
Cognition, 4, 311–331.
Baron-Cohen, S., Wheelwright, S., Stone, V., & Rutherford, M. (1999c). A mathematician,
a physicist, and a computer scientist with Asperger Syndrome: performance on folk
psychology and folk physics test. Neurocase, 5, 475–483.
Baron-Cohen, S., Wheelwright, S., Stott, C., Bolton, P., & Goodyer, I. (1997c). Is there a
link between engineering and autism? Autism: An International Journal of Research and
Practice, 1, 153–163.
Bates, E., Benigni, L., Bretherton, I., Camaioni, L., & Volterra, V. (1979). Cognition and com-
munication from 9–13 months: correlational findings. In E. Bates (Ed.), The emergence of
symbols: cognition and communication in infancy,. New York: Academic Press.
Bettelheim, B. (1968). The empty fortress. Chicago: The Free Press.
Bolton, P., MacDonald, H., Pickles, A., Rios, P., Goode, S., Crowson, M., Bailey, A., & Rutter,
M. (1994). A Case-Control Family History Study of Autism. Journal of Child Psychology
and Psychiatry, 35, 877–900.
Butterworth, G. (1991). The ontogeny and phylogeny of joint visual attention. In W. A (Ed.),
Natural theories of mind,. Oxford.
Carey, S. (1985). Conceptual change in childhood. Cambridge, Mass: Bradford Books MIT Press.
Charman, T., & Baron-Cohen, S. (1992). Understanding beliefs and drawings: a further test of
the metarepresentation theory of autism. Journal of Child Psychology and Psychiatry, 33,
Charman, T., & Baron-Cohen, S. (1995). Understanding models, photos, and beliefs: a test of
the modularity thesis of metarepresentation. Cognitive Development, 10, 287–298.
Csibra, G., Gergely, G., Biro, S., Koos, O., & Brockbanck, M. (1999). Goal attribution without
agency cues: the perception of “pure reason” in infancy. Cognition, 72, 253–284.
Damasio, A., Tranel, D., & Damasio, H. (1990). Individuals with sociopathic behaviour caused
by frontal lobe damage fail to respond autonomically to socially charged stimuli.
Behavioural Brain Research, 14, 81–94.
Dennett, D. (1987). The Intentional Stance. Cambridge, Mass: MIT Press/Bradford Books.
Ekman, P. (1992). Facial expression of emotion: an old controversy and new findings. In
V. Bruce (Ed.), The Face,: Royal Society.
Folstein, S., & Rutter, M. (1977). Infantile Autism: A Genetic Study of 21 Twin Pairs. Journal of
Child Psycholology and Psychiatry, 18, 297–321.
Frith, U. (1989). Autism: explaining the enigma. Oxford: Basil Blackwell.
Frith, U. (1991). Autism and Asperger’s Syndrome. Cambridge: Cambridge University Press.
Gelman, S., & Hirschfield, L. (1994). Mapping the Mind. Cambridge: Press Syndicate, University
of Cambridge.

Gergely, G., Nadasdy, Z., Gergely, C., & Biro, S. (1995). Taking the intentional stance at
12 months of age. Cognition, 56, 165–193.
Gillberg, C. (1991). Clinical and neurobiological aspects of Asperger syndrome in six family
studies. In U. Frith (Ed.), Autism and Asperger Syndrome,. Cambridge: Cambridge
University Press.
Happe, F. (1993). Communicative competence and theory of mind in autism: A test of
Relevance Theory. Cognition, 48, 101–119.
Happe, F. (1996). Studying weak central coherence at low levels: children with autism do not
succumb to visual illusions. A research note. Journal of Child Psychology and Psychiatry, 37,
Harris, P., Johnson, C. N., Hutton, D., Andrews, G., & Cooke, T. (1989). Young children’s the-
ory of mind and emotion. Cognition and Emotion, 3, 379–400.
Hart, C. (1989). Without reason. New York: Harper & Row, Inc.
Hatano, G., & Inagaki, K. (1994). Young children’s naive theory of biology. Cognition, 50, 171–188.
Hood, B., Willen, J., & Driver, J. (1997). An eye-direction detector triggers shifts of visual
attention in human infants. Unpublished ms., Harvard University.
ICD-10. (1994). International classification of diseases. (10th ed.). Geneva, Switzerland: World
Health Organisation.
Johnson, S. (2000). The recognition of mentalistic agents in infancy. Trends in Cognitive Sciences,
4, 22–28.
Jolliffe, T., & Baron-Cohen, S. (1997). Are people with autism or Asperger’s Syndrome faster
than normal on the Embedded Figures Task? Journal of Child Psychology and Psychiatry, 38,
Kanner, L. (1943). Autistic disturbance of affective contact. Nervous Child, 2, 217–250.
Karmiloff-Smith, A. (1992). Beyond modularity: MIT Press/Bradford Books.
Karmiloff-Smith, A., Grant, J., Bellugi, U., & Baron-Cohen, S. (1995). Is there a social module?
Language, face-processing and theory of mind in William’s Syndromeand autism, in
press. Journal of Cognitive Neuroscience, 7, 196–208.
Klin, A., Volkmar, F., Sparrow, S., Cicchetti, D., & Rourke, B. (1995). Validity and neuropsy-
chological characterization of Asperger Syndrome: convergence with nonverbal learn-
ing disabilities syndrome. Journal of Child Psychology and Psychiatry, 36, 1127–1140.
Kimura, D. (1992). Sex differences in the brain. Scientific American, September, 119–125.
Le Couteur, A., Bailey, A., Goode, S., Pickles, A., Robertson, S., Gottesman, I., & Rutter, M.
(1996). A broader phenotype of autism: the clinical spectrum in twins. Journal of Child
Psychology and Psychiatry, 37, 785–801.
Leekam, S., & Perner, J. (1991). Does the autistic child have a metarepresentational deficit?
Cognition, 40, 203–218.
Leslie, A. (1995). ToMM, ToBy, and Agency: core architecture and domain specificity. In L.
Hirschfeld & S. Gelman (Eds.), Domain specificity in cognition and culture,. New York:
Cambridge University Press.
Leslie, A., & Keeble, S. (1987). Do six-month old infants perceive causality? Cognition, 25,
Leslie, A. M. (1987). Pretence and representation: the origins of “theory of mind”. Psychological
Review, 94, 412–426.
Leslie, A. M., & Frith, U. (1988). Autistic children’s understanding of seeing, knowing, and
believing. British Journal of Developmental Psychology, 6, 315–324.
Leslie, A. M., & Thaiss, L. (1992). Domain specificity in conceptual development: evidence
from autism. Cognition, 43, 225–251.
Lovell, A. (1978). In a summer garment. London: Secker & Warburg.
Mundy, P., & Crowson, M. (1997). Joint attention and early social communication. Journal of
Autism and Developmental Disorders, 27, 653–676.
Ozonoff, S., Pennington, B., & Rogers. (1990). Are there emotion perception deficits in young
autistic children? Journal of Child Psychology and Psychiatry, 31, 343–363.

Ozonoff, S., Rogers, S., Farnham, J., & Pennington, B. (1994). Can standard measures identify
subclinical markers of autism? Journal of Autism and Developmental Disorders.
Park, C. (1967). The Siege. London: Hutchinson.
Pennington, B., Rogers, S., Bennetto, L., Griffith, E., Reed, D., & Shyu, V. (1997). Validity Test
of the Executive Dysfunction Hypothesis of Autism. In J. Russell (Ed.), Executive
Functioning in Autism,. Oxford: Oxford University Press.
Perner, J., & Wimmer, H. (1985). “John thinks that Mary thinks that...” Attribution of second-
order beliefs by 5–10 year old children. Journal of Experimental Child Psychology, 39,
Plaisted, K., O’Riordan, M., & Baron-Cohen, S. (1998a). Enhanced discrimination of novel,
highly similar stimuli by adults with autism during a perceptual learning task. Journal of
Child Psychology and Psychiatry, 39, 765–775.
Plaisted, K., O’Riordan, M., & Baron-Cohen, S. (1998b). Enhanced visual search for a con-
junctive target in autism: A research note. Journal of Child Psychology and Psychiatry,
39 (777–783).
Pratt, C., & Bryant, P. (1990). Young children understand that looking leads to knowing (so
long as they are looking into a single barrel). Child Development, 61, 973–983.
Premack, D. (1990). The infant’s theory of self-propelled objects. Cognition, 36, 1–16.
Raven, J. C. (1956). Coloured Progressive Matrices. London: H.K. Lewis and Co.
Rochat, P., Morgan, R., & Carpenter, M. (1997). Young infants’ sensitivity to movement infor-
mation specifying social causality. Cognitive Development, 12, 537–561.
Russell, J. (1997). How executive disorders can bring about an inadequate theory of mind. In
J. Russell (Ed.), Autism as an executive disorder, . Oxford: Oxford University Press.
Scaife, M., & Bruner, J. (1975). The capacity for joint visual attention in the infant. Nature, 253,
Shah, A., & Frith, U. (1983). An islet of ability in autism: a research note. Journal of Child
Psychology and Psychiatry, 24, 613–620.
Shah, A., & Frith, U. (1993). Why do autistic individuals show superior performance on the
block design test? Journal of Child Psychology and Psychiatry, 34, 1351–1364.
Sigman, M., Mundy, P., Ungerer, J., & Sherman, T. (1986). Social interactions of autistic,
mentally retarded, and normal children and their caregivers. Journal of Child Psychology
and Psychiatry, 27, 647–656.
Spelke, E., Phillips, A., & Woodward, A. (1995). Infants’ knowledge of object motion and
human action. In D. Sperber, D. Premack, & A. Premack (Eds.), Causal Cogntion: A
Multidisciplinary Debate,. Oxford: Oxford University Press.
Sperber, D., Premack, D., & Premack, A. (Eds.). (1995). Causal Cogntion: A Multidisciplinary
Debate. Oxford: Oxford University Press.
Tager-Flusberg, H. (1993). What language reveals about the understanding of minds in chil-
dren with autism. In S. Baron-Cohen, H. Tager-Flusberg, & D. J. Cohen (Eds.),
Understanding other minds: perspectives from autism,: Oxford University Press.
Tager-Flusberg, H., Boshart, J., & Baron-Cohen, S. (1998). Reading the windows of the soul:
evidence of domain specificity sparing in Williams syndrome. Journal of Cognitive
Neuroscience, 10, 631–639.
Tomasello, M. (1988). The role of joint-attentional processes in early language acquisition.
Language Sciences, 10, 69–88.
Walker, A. S. (1982). Intermodal perception of exptessive behaviours by human infants.
Journal of Experimental Child Psychology, 33, 514–535.
Wellman, H. (1990). Children’s theories of mind. Bradford: MIT Press.
Wellman, H., & Bartsch, K. (1988). Young children’s reasoning about beliefs. Cognition, 30,
Wellman, H., & Inagaki, K. (1997). The emergence of core domains of thought: Children’s reasoning
about physical, psychological, and biological phenomena. (Spring ed.). (Vol. 75): Jossey-Bass Inc
Whiten, A. (1991). Natural theories of mind. Oxford: Basil Blackwell.

Wimmer, H., & Perner, J. (1983). Beliefs about beliefs: Representation and constraining func-
tion of wrong beliefs in young children’s understanding of deception. Cognition, 13,
Wing, L. (1981). Asperger Syndrome: a clinical account. Psychological Medicine, 11, 115–130.
Wing, L., & Gould, J. (1979). Severe impairments of social interaction and associated abnor-
malities in children: epidemiology and classification. Journal of Autism and Developmental
Disorders, 9, 11–29.
Yirmiya, N., Sigman, M., Kasari, C., & Mundy, P. (1992). Empathy and cognition in high func-
tioning children with autism. Child Development, 63, 150–160.
Zaitchik, D. (1990). When representations conflict with reality: the preschooler’s problem with
false beliefs and “false” photographs. Cognition, 35, 41–68.

Mailing Address:
Simon Baron-Cohen, Ph.D.
Autism Research Centre
Departments of Experimental Psychology and Psychiatry,
University of Cambridge, Downing St,
Cambridge, CB2 3EB, UK
Shared Deficits or Different Disorders?

Christopher Gillberg, M.D., Ph.D.

Abstract. This paper provides a selective review and personal point of view about the rela-
tionship between Asperger syndrome and high-functioning autism. The terms are more likely syn-
onyms than labels for different disorders. Most cases of Asperger syndrome and high-functioning
autism are unlikely to be caused by one set of identical factors. We may be looking for genetic fac-
tors or brain damage factors, or other factors, not yet studied, but it is unlikely, particularly with
the range of comorbidity demonstrated that we are dealing with one set of identical factors.
Overlapping and shared brain circuitries are likely to be involved in autism and some of its over-
lapping and comorbid conditions. For instance, the link with ADHD, DCD and Tourette syn-
drome suggests a common dopamine abnormality, and the link with affective disorders and the
obsessive compulsive symptoms suggests serotonin dysfunction. Specific disorders, such as tuberous
sclerosis and certain chromosomal disorders may provide very important links when it comes to
the location of the brain lesion and also to the fine mapping of genetic factors.


Hans Asperger’s original paper from 1944 (Asperger, 1944) is interesting in that, long
before anybody thought of including very high-functioning individuals with autistic
features in the broader group with “autism” he described the characteristics of “high-
functioning autism”. True, he referred to “autistic psychopathy”, and most of us work-
ing in the field now use Lorna Wing’s term “Asperger’s syndrome”, but it has
becoming increasingly clear that the constellation of symptoms he described reflects
“autism” in individuals who are not moderately or severely mentally retarded
(Gillberg & Ehlers, 1998).
Having said this, I should perhaps add that it was not Asperger who first
described the syndrome that now bears his name. A Russian neurologist, Eva
Ssucharewa, published a paper in the mid 1920s in which she described “schizoid
personality disorder” in children. Reading Sula Wolff’s translation of that paper it
becomes clear that Ssucharewa described the core deficits and major hallmarks of
autism long before Asperger or Kanner (Ssucharewa 1926, Wolff 1995).


The Diagnosis of Asperger Syndrome and

High-Functioning Autism

Asperger Syndrome

Asperger never outlined diagnostic criteria for “autistic psychopathy”. The first
operationalized criteria for diagnosing Asperger syndrome were outlined by our
group at the first international conference on Asperger’s syndrome which was held
in London in 1988. These criteria, which are based specifically on Asperger’s own
case reports, appeared in print in 1989 and were elaborated in more detail in 1991
(Gillberg & Gillberg, 1989, Gillberg, 1991). The Gillberg criteria for Asperger syn-
drome have been systematically tested against other criteria for the disorder and
against criteria for autism (typical or atypical). They delineate a group of highly
verbal, ritualistic, socially severely malfunctioning individuals with special narrow
interests and odd ways of communicating with other people (Table 1). About half
of all high-functioning individuals in the autism spectrum meet these criteria
(Leekam, et al., 2000). Many of these also meet criteria for autistic disorder or atyp-
ical autism (roughly equivalent to pervasive developmental disorder not otherwise
specified). The Gillberg criteria are problematic in that they require a very severe
and very specific type of symptomatology and that they do not rule out the diag-
nosis of autistic disorder. We have recently thought of changing the algorithm so
that meeting criteria for five out of the six (rather than all six) symptom areas would
suffice for caseness if the social criterion is met and reserve the diagnosis for indi-
viduals whose symptomatology does not better fit the criteria/clinical picture of
autistic disorder.
Peter Szatmari published his criteria for Asperger’s syndrome in 1989
(Szatmari, et al., 1989). They are similar to those of our own group but can be used
only for individuals who do not meet criteria for autistic disorder (Table 2). The
ICD-10 (WHO 1992) and DSM-IV (APA 1994) criteria for the syndrome are vir-
tually identical to each other (Table 3). They are problematic in that they specifi-
cally exclude cases with signs of early language, developmental or social delays.
Virtually nobody with an autism spectrum disorder fits these criteria (Leekam,
et al., 2000). Asperger’s own cases do not meet criteria for DSM-IV Asperger’s dis-
order (Ozonoff and Miller, 1997).
In clinical practice, the DSM-IV criteria for Asperger’s disorder are not helpful.
There is a problem stemming from the fact that clinicians are trying to squeeze
patients into the DSM-IV/ICD-10 framework, which really requires that individuals
with the syndrome should have been normal up until age three years. I have yet to
meet a patient with the clinical presentation that Asperger described who was com-
pletely normal in his development early on. Another problem is that the actual
symptom threshold for qualifying for a diagnosis is very low; only two social and one
behavioural symptoms are required to reach diagnostic status. The ICD-10 has the
additional problem that there is no specification that symptoms have to be handi-
capping in daily life.

Diagnostic criteria for Asperger syndrome (Gillberg & Gillberg)

1. Social impairment (extreme egocentricity)

(at least two of the following):
(a) difficulties interacting with peers
(b) indifference to peer contacts
(c) difficulties interpreting social cues
(d) socially and emotionally inappropriate behaviour
2. Narrow interest
(at least one of the following):
(a) exclusion of other activities
(b) repetitive adherence
(c) more rote than meaning
3. Compulsive need for introducing routines and interests
(at least one of the following):
(a) which affect the individual’s every aspect of every-day life
(b) which affect others
4. Speech and language peculiarities
(at least three of the following):
(a) delayed speech development
(b) superficially perfect expressive language
(c) formal pedantic language
(d) odd prosody, peculiar voice characteristics
(e) impairment of comprehension including misinterpretations of literal/implied meanings
5. Non-verbal communication problems
(at least one of the following):
(a) limited use of gestures
(b) clumsy/gauche body language
(c) limited facial expression
(d) inappropriate facial expression
(e) peculiar, stiff gaze
6. Motor clumsiness
poor performance in neuro-developmental test

There are no universally agreed criteria for high-functioning autism, but it

would seem reasonable to reserve this label for those individuals who meet full cri-
teria for autistic disorder, and who, in addition, have full-scale IQs in the non-
retarded range.
It would seem obvious from this overview of diagnostic criteria that (a) there is
no clear consensus as to the diagnostic criteria that are most appropriate for Asperger
syndrome, (b) there are major problems fitting DSM-IV/ICD-10 criteria for
Asperger syndrome to real live patients/individuals, and (c) there is no clear border-
line between Asperger syndrome and high-functioning autism (in fact, in many
instances, the two terms are used interchangeably for the same kind of behavioural

Diagnostic criteria for Asperger syndrome according to Szatmari et al. (1989)

1. Social isolation
(at least two of the following):
(a) no close friends
(b) avoids others
(c) no interest in making friends
(d) a loner
2. Impaired social interaction
(at least one of the following):
(a) approaches others only to have own needs met
(b) clumsy social approach
(c) one-sided responses to peers
(d) difficulty sensing feelings of others
(e) indifference to the feelings of others
3. Impaired non-verbal communication
(at least one of the following):
(a) limited facial expression
(b) impossible to read emotions through facial expression of the child
(c) inability to convey message with eyes
(d) avoids looking at others
(e) does not use hands to aid expression
(f) large and clumsy gestures
(g) infringes on others people’s physical space
4. Speech and language peculiarities
(at least two of the following):
(a) abnormalities of inflection
(b) over-talkative
(c) non-communicative
(d) lack of cohesion to conversation
(e) idiosyncratic use of words
(f) repetitive patterns of speech

The Clinical Presentation

Asperger Syndrome

There is always marked social impairment in Asperger syndrome, usually show-

ing as extreme egocentricity. There is mostly a much decreased ability to interact
with peers, often coupled with a lack of desire to interact with peers, a lack of appre-
ciation of social cues, and socially and emotionally inappropriate behaviors. The
social impairment may actually be just as marked in Asperger syndrome as in clas-
sic autism, but the overall better functioning and the higher level of skills in the for-
mer syndrome may go along way in obfuscating the major problems. However,

ICD-10, WHO, 1993 Diagnostic criteria for Asperger syndrome

1. There is no clinically significant general delay in spoken or receptive language

or cognitive development. Diagnosis requires that single word should have developed
by 2 years of age or earlier and that communicative phrases be used by 3 years of age or
earlier. Self-help skills, adaptive behaviour, and curiosity about the environment during
the first three years should be at a level consistent with normal intellectual development.
However, motor milestones may be somewhat delayed and motor clumsiness is usual
(although not a necessary diagnostic feature). Isolated special skills, often related to
abnormal preoccupations, are common, but are required for diagnosis.
2. There are qualitative abnormalities in reciprocal social interaction in at least two
of the following areas:
(a) failure adequately to use eye-to-eye gaze, facial expression, body posture, and gesture
to regulate social interaction
(b) failure to develop (in a manner appropriate to mental age, and despite ample
opportunities) peer relationships that involve a mutual sharing of interests,
activities, and emotions
(c) lack of social-emotional reciprocity as shown by an impaired or deviant response
to other people’s emotions, or lack of modulation of behaviour according to social
context; or a weak integration of social; emotional, and communicative behaviours
(d) lack of spontaneous seeking to share enjoyment, interests, or achievements with
other people (e.g. lack of showing, bringing, or pointing out to other people
objects of interest to the individual)
3. The individual exhibits an unusually intense, circumscribed interest or restricted,
repetitive, and stereotyped patterns of behaviour, interests, and activities in at
least on two of the following:
(a) an encompassing preoccupation with one or more stereotyped and restricted patterns
of interest that are abnormal in their intensity and circumscribed nature though not in
their content or focus
(b) apparent compulsive adherence to specific, non-functional routines or rituals
(c) stereotyped and repetitive motor mannerisms that involve either hand- or finger-
flapping or -twisting, or complex whole-body movements
(d) preoccupations with part-objects or non-functional elements of play materials (such as
their odour, the feel of their surface, or the noise or vibration that they generate

several years may pass before family, peers, relatives and teachers understand that
something is seriously amiss, and it may only be with hindsight that they realize that
there was never a period of normal development.
The narrow interest pattern, was something that Asperger himself put a lot of
emphasis on. He felt that this interest pattern should lead to the exclusion of other
activities, or be very repetitive, or be more relying on memory than underlying
meaning. Even though the narrow interest pattern is highly characteristic of the
most typical cases of males with Asperger syndrome, there are those, and particu-
larly those females, who otherwise fit the criteria for the disorder, who do not
demonstrate this feature. Some girls (and a very few boys) with the other core fea-

tures of the condition have a strong tendency to avoid demands and to always say
“no”. It is as though their main interest in life is to say no, negative things, and to
go around “being bored”. They themselves cannot seem to find anything to interest
them. Just occasionally, other people may find something to attract their interest,
and they can then become totally absorbed for a while, whereafter the “boring state”
In spite of the fact that “his” children, as he sometimes called them, were usually
good with formal language, Asperger noted that they had language and speech pecu-
liarities. He drew attention to the possibility that, even though they had this formal
language level, which was usually very good, they were very often slightly delayed,
particularly as compared to their brothers and sisters. It is common for a child with
Asperger syndrome to have “delayed” expressive language development; they do
not speak at an early age, even though you had the feeling that they would have been
able to if they wanted to/felt the need to. I certainly see a number of kids who say
nothing for two-three years and then suddenly start speaking because they “have
something to say”. Some of them actually say: “Why should I speak before I have
something important to say?”. There are also children who, are able to read before
they start speaking. They have superficially expressive perfect language and they are
very often formal and pedantic in their style and they have, most of them, this very
odd prosody. More often than not there is impairment of comprehension, also, in the
face of—sometimes extremely—good expressive skills. They sometimes have very
mild, but often very marked comprehension problems.
There are also non-verbal communication problems in this group of individuals,
and Asperger put a lot of emphasis on the odd gaze behaviors, in particular the star-
ing gaze and the fixed gaze of the young boys that he saw. Incidentally, it was only
towards the end of his life that he believed that the syndrome could occur in any
form in females. He thought it was an exclusively male disorder. Odd and awkward
gait, strange posturing (sometimes approaching catatonia) and a mask-like facial
expression are all common features.
Asperger also noted the motor clumsiness. This may be most marked in social
situations and be reflected in clumsy feeding behaviors, spilling, stumbling, tripping,
or in inability to swim, ride a bike etc. Our own group always felt that this was an
important part of the syndrome.

High-Functioning Autism

The description of the typical presentation of Asperger syndrome is also one that would fit
high-functioning autism in most domains. If there are differentitating features and the two con-
ditions are separate from each other (at least in some cases), language competence and full scale
IQ would probably be the most likely differentiators, both being at a lower level of functioning
in high-functioning autism than in Asperger syndrome.


Asperger Syndrome

The prevalence of this disorder, as diagnosed on the Gillberg criteria, has been
in the range of 0.26 through 0.71 per cent (Gillberg & Gillberg, 1989, Ehlers et al.,
1993, Magnusson & Gudjonsdottir, 1994, Landgren, et al., 1996, Kadesjö & Gillberg,
1999). The five prevalence studies—which have all come from the Nordic countries—
are in relatively good agreement, and it would seem reasonable to conclude that the
rate of the disorder is about half a per cent of the general population of school age
children. Nevertheless, further prevalence studies from other geographical regions
are clearly needed before firm conclusions can be drawn in this respect.

High-Functioning Autism

The most recent studies of autistic disorder have quoted prevalence figures of about 0.1 per
cent. About 30% of cases in those studies have had IQs over 70. These would constitue “high-
functioning” cases, meaning that the prevalence of high-functioning autism (not necessarily
meeting criteria for Asperger syndrome) would be 0.03 per cent.

Mental Retardation

Asperger Syndrome

The vast majority of individuals who receive a diagnosis of Asperger syndrome have IQs in
the normal or near normal range (or are of superior intelligence). Only a few per cent of all
published cases have associated mental retardation.

High-Functioning Autism

Almost by definition, cases with high-functioning autism would not have mental retarda-
tion. However, in clinical practice, it is not uncommon for individuals with this diagnostic label
to have normal performance IQ, but verbal IQ may be in the retarded range, so that full scale
IQ is just under 70.

The Rate of Mental Retardation in Autism

If, as most authors seem to agree, Asperger syndrome is part of the autism spectrum, then
we need to change our concepts as to how often mental retardation is associated with autism.

Counting autistic disorder (low- and high-functioning cases) and Asperger syndrome, the rate
of associated mental retardation in the autism spectrum would be about 15%, far different from
the much quoted figures of 70–80% mental retardation in autism. Now, 15% is well above the
prevalence of mental retardation in the general population, but certainly nothing near the level
that we have been used to thinking about when discussing autism. Thus, the majority of indi-
viduals in the autism spectrum, are NOT mentally retarded.

Male:Female Ratios

The boy:girl ratio in most studies of Asperger syndrome have been reported to
be 5-10:1 (even up to 47:1). However, in the Nordic population studies, there are
usually about 3–4 boys for every girl. If, instead, we go to our clinic, where all of
the children are very severely impaired, the ratio of boys to girls goes up drastically.
It is likely that girls/women with the core features of Asperger syndrome/high-func-
tioning autism may go unrecognised to a much larger extent than boys. This, in turn,
may be due to the fact that both Asperger and Kanner described mostly prototypi-
cal boys, and that, therefore, our way of conceptualising the syndromes is based on
the phenotypical expression in males, not in females. Interestingly, we have found a
high rate of Asperger syndrome/high-functioning autism/PDD NOS in female
teenagers/adults with anorexia nervosa (Wentz, et al., 2000) and selective mutism
(Kopp & Gillberg, 1997). They had had core autistic features all their lives, but it was
only when they were meticulously worked-up for the other condition that their
autism spectrum disorder was really disclosed.


Asperger syndrome is very often comorbid with at least one other condition
(Figure 1). The findings that apply to Asperger syndrome are probably translatable
to high-functioning autism also.
According to the population-study by our own group, tics (including full-
blown Tourette syndrome) and ADHD are the most common comorbidities
(Ehlers et al., 1993), each occurring in more than half of all clear-cut or suspected
cases. Developmental coordination disorder (DCD) is almost universal, but then,
at least according to some of the diagnostic algorithms, clumsiness is part and par-
cel of Asperger syndrome, and it might therefore seem redundant to list DCD as
a separate disorder. In the Nordic countries ADHD with DCD is usually regis-
tered as “DAMP” (“Deficits in Attention, Motor control and Perception”). Several
studies have shown DAMP to be strongly associated with autism spectrum prob-
lems, and the rate of Asperger syndrome in DAMP is increased about thirtyfold
compared with the general population (Gillberg & Gillberg, 1989, Kadesjö &
Gillberg, 2000).

FIGURE 1. Overlap of mild to moderate, and severe) DAMP (ADHD+DCD), autistic

features, and Asperger syndrome (data from Gillberg C., 1983 and Gillberg I.C. and
Gillberg C., 1989) and possible overlap with ICD-10 hperkinetic syndrome and DCD.
(Adapted from Clinical Child Neuropsychiatry, Gillberg C., 1995, Cambridge:
Cambridge University Press, by permission.)

However, there are many other conditions that sometimes can occur in Asperger
syndrome. Some of these may be chance coincidences, but in many cases there
appears to be a more meaningful link between the autism spectrum disorder and the
other condition (Gillberg & Billstedt, 2000). This may be particularly true for obses-
sive-compulsive disorder (OCD) and obsessive-compulsive personality disorder
(OCPD), for anorexia nervosa and selective mutism (Gillberg, 2000).
Depression quite often develops in the pre-adolescent or adolescent period in
children who have high-functioning autism or Asperger syndrome. This may either
be a reflection of comorbidity with manic-depressive illness (deLong, et al., 1986) or
be seen as a reactive condition following in the footsteps of feeling socially awkward
and of being an “outsider”.
Many adults with Asperger syndrome or high-functioning autism apply for
psychiatric help but are only occasionally correctly diagnosed as having an autism

spectrum disorder. We have seen our previous patients diagnosed with borderline
personality disorder, antisocial personality disorder, paranoid disorder, psychosis,
and schizophrenia. Only rarely would these diagnostic labels seem appropriate.
There is, of course, the occasional patient who, having had an autism spectrum dis-
order from the very first years of life, does develop schizophrenic symptoms in late
adolescence, in which case both the autism spectrum disorder and schizophrenia
should be diagnosed. However, much more often, the schizophrenia label is not
really appropriate. It may be a matter of the adult psychiatrist not being familiar with
the history and symptoms of an individual with Asperger syndrome/high-function-
ing autism which makes him/her liable to make a diagnosis of a condition with some
overlapping symptomatology for which there is a well-known framework.
Alcoholism appears to be much overrepresented in Asperger syndrome as com-
pared with the general population (Wing, 1981, Hellgren, et al., 1994), but the stud-
ies published to date have either been on potentially biased groups or included very
small samples making it impossible to draw any generalized conclusions at this stage.
Learning disorders (including problems with reading, writing, mathematics, phys-
ical education) are probably overrepresented, particularly in high-functioning autism,
but the systematic evidence in the field is limited. Even in Asperger syndrome (where
IQ is usually normal or even very high, and hyperlexia is very common) there may
be severe academic difficulties. However, these seem usually to stem more from the
social and behavioural problems rather than from any basic cognitive deficit (physi-
cal education may be an exception though, and the problems encountered in this field
might well be a consequence of underlying neurological dysfunction).

Genetic Factors

There is good evidence that genetic factors account for a majority of all cases with
autistic disorder, including those that are designated as high-functioning (Gillberg &
Coleman, 2000). The evidence comes from population studies of twins, family stud-
ies, studies of sib-pairs and epidemiological studies of autism in the general popula-
tion. The search for susceptibility genes is on, and it appears that certain
chromosomal regions may be much more interesting than others.
There is widespread agreement that Asperger syndrome is probably a genetic
disorder/variant in a majority of all cases. However, there is nothing near the
amount of evidence that exists for classic autistic disorder. Autistic disorder and
Asperger syndrome quite often appear in one and the same extended family. There
is, quite possibly, a genetic link with manic-depressive illness in Asperger syndrome
(deLong, et al., 1986), at least in a subgroup. Nevertheless, studies reporting on
series of clinical cases have also found a high index of perinatal distress (Wing 1981,
Gillberg 1989, Rickarby, et al., 1991), and one of these (Rickarby, et al., 1991) found
no indication of familial loading. Perinatal stress may sometimes be a reflection of a
genetic defect. However, in other cases, it may contribute to brain damage in itself.
Thus, any generalized conclusion in respect of the genetics of Asperger syndrome
would seem premature.

Neurology and Neuropsychology

In classic autism, there is quite often an associated medical condition such as

tuberous sclerosis or a chromosomal disorder. Population studies have indicated that
such an association applies in about one in four of all cases, whereas in clinical stud-
ies the rate drops to about one in eight. (This may be because clinicians who work
up and follow a child with tuberous sclerosis may not refer him/her for a separate
diagnosis of autism since they feel that there is already an “established diagnosis”—
tuberous sclerosis—that accounts for the child’s problems.)
The rate of associated medical conditions in the high-functioning cases—including
Asperger syndrome—is lower (probably at or under one in eight cases) (Gillberg, et al.,
1987). However, occasionally, you do come across Asperger syndrome phenotypic
presentations with underlying tuberous sclerosis (Gillberg, I.C., et al., 1994), fetal alco-
hol syndrome (Aronson, et al., 1997), CATCH-22-syndrome (Niklasson, et al., 2000) or
other chromosomal disorders (Annerén et al., 1998). In some instances, these underly-
ing disorders provide clues as to the basic etiology of the autism spectrum disorder.
Epilepsy is common in autism spectrum disorders. Several studies indicate that
about one in four to one in three of all individuals with autistic disorder has epilepsy,
which can have its onset in the first two years or (usually much) later (Gillberg, 1991).
The rate is much lower in those with higher IQ: one study reported ten per cent of
those with high-functioning autism and nobody with Asperger syndrome had
epilepsy (Ehlers, et al., 1997).
Many studies have found indices of anatomical brain abnormality in Asperger
syndrome. However, all of these have been on potentially selected clinical samples and
it has not yet been possible to tease out whether referral factors/comorbid conditions
rather than Asperger syndrome per se account for the findings. Our own group com-
pared brain CT scans of 18 Asperger syndrome and 22 high-functioning children with
autistic disorder and found some cerebral atrophy in 17% of the former and 22% of the
latter group (Gillberg, 1989). Another group found MRI abnormalities (mostly right-
sided cortical) in the brains of five out of seven individuals with Asperger syndrome and
comorbid Tourette syndrome. Other MRI and PET-scan studies have yielded variable
results (Piven, et al., 1992, Herold, et al., 1988, Horwitz, et al., 1988). Our own group
studied adults with Asperger syndrome and compared them with normal adults on PET-
scan measures in connection with theory of mind tasks (Happé, et al., 1996). The
Asperger group showed lack of activation of a relatively distinct area in the left frontal
lobe, an area which was markedly activated in the normal group. In a SPECT-study of
high- and low-functioning cases with autism (with and without epilepsy), all individuals
showed temporal lobe hypoperfusion and frontal hypoperfusion in one in three cases
regardless of IQ-level and presence of epilepsy (Gillberg I.C., et al., 1993).
Here are just some of the things that have been associated in a few studies with
Asperger’s Syndrome. The study we just completed on Megalencephalus, that is large
head size in Asperger’s and Autism suggest that it’s only in the very high function-
ing group that you’ll find a large excess of individuals with big heads. In fact 24% of
classic Asperger’s cases we have consistently macrocephalus. They have it both at
birth and later so we did not replicate the findings by a few other groups that we mega-

lencephalus comes at a later age. It was there from the very beginning and it was there
when we followed them up much later. No consistent findings have yet emerged in
respect of other neurophysiological or neurochemical studies in the field of high-func-
tioning autism or Asperger syndrome.
There is now an enormous wealth of data relating to the neuropsychology of
autism (Happé, 2000). Conversely, there is a very limited neuropsychological litera-
ture that has attempted to differentiate high-functioning autism from Asperger syn-
drome (Nydén, 2000). In one study, children with high-functioning autism showed
the typical autism profile on the WISC with a peak on Block Design and troughs on
Comprehension and Picture Arrangement. Individuals with Asperger syndrome
showed much better performance on Comprehension but relatively poor on Object
Assembly, Digit Span, and Digit Symbol. These latter results were similar to those of
an ADHD (attention-deficit/hyperactivity disorder group), placing Asperger syn-
drome in the middle of a spectrum with autistic disorder at one end and ADHD at
the other (Ehlers, et al., 1997). According to another study, school age children with
high-functioning autism and Asperger syndrome could not be differentiated from
those with ADHD on most measures of executive functioning, even though they dif-
fered markedly from a normal comparison sample (Nydén et al., 2000).

Outcome Studies

The results of the general population follow-up studies of individuals in the autism
spectrum that we are now doing, and which are not yet completed, suggest that out-
come is restricted, not only in low-functioning cases of autism but also in those that are
most high-functioning. This may hold particularly true for those, who, at one or other
time during development, applied for help. Now, that should not be taken to mean that
if you apply for help you have a poorer outcome because the people who help you turn
you into some kind of monster. But, it does imply that if you come to a hospital setting,
for instance, or a clinic, you are more likely to have more severe problems. Some data
(e.g. Ehlers & Gillberg, 1993) indicate that only about half of all individuals who meet
full criteria for Asperger syndrome (and who are, in a sense, “clinically” impaired),
actually apply for professional help in childhood or adolescence.
The follow-up studies reporting on outcome in clinical series of patients with high-
functioning autism also suggest that prognosis is not very good, even though about one
in four may function relatively well in adult age (Rumsey, et al., 1985, Szatmari, 1989).
I have met many individuals with Asperger syndrome who have superficially
excellent outcomes, who have had a university degree, an excellent job, and a fam-
ily. Nevertheless, the spouses of the young Asperger syndrome men that we followed
up still feel that their husbands have some terrible problems. The men themselves
feel that they are fine and they’re doing well and there is not a major problem. Some
of them have children. Unfortunately, at least in one case, that child has turned out
to have severe autism.

There is obviously a range of outcomes in Asperger syndrome and you will see
that the variability is much broader than in so-called classical autism, where the
range, although wide, is much more narrow.
The results from our ongoing follow-up study suggest that 50% do relatively well
in adult age. There seems to be a relatively stable prognosis in that, if you had
Asperger’s or high-functioning autism as a child, you are very likely to have those
same problems when you’re 20 years older, even though you may get a very differ-
ent diagnosis once you go to see adult psychiatrist.

High-Functioning Autism and Asperger Syndrome:

Where are the Boundaries?

Lorna Wing, Judy Gould and I have recently done a study with Sue Leekam and
others in London (Leekam, et al., 2000). We looked at 200 cases that attended the
Center for Social and Communication Disorders in London over a period of years.
They were all referred with a suspicion of an autism spectrum disorder and were
worked up using the DISCO, the Diagnosis of Social and Communication Disorders
(Wing, et al., 2000). This instrument allows you to make an algorithm diagnosis by
computer: you get a printout of exactly who meets what criteria. So we used all the
possible criteria that were in the literature, ICD 10-Childhood Autism, ICD 10-
Asperger’s Syndrome, ICD 10-Atypical Autism, Gillberg-Asperger syndrome etc.
Out of those 200 cases, 174 met full Childhood Autism criteria by ICD-10 criteria.
(Parenthetically, I might add that we are currently doing a study in Gothenburg
where we are comparing the results of the DISCO to those of the ADI-R and you
get very similar results although the ADI-R tends to give a diagnosis of Childhood
Autism slightly more often than the DISCO. Now, one thing that most people have
not observed is that diagnostic algorithms for atypical autism and PDD-NOS do not
rule out a diagnosis of Asperger’s, meaning that atypical autism and Asperger’s can
be diagnosed in one and the same individual.) Only three individuals out of the 200
met full criteria for Asperger’s according to the ICD 10. That is only because of the
criterion of normal development in the first three years of life. However, there were
91 that met the Asperger criteria according to our own group. We looked more care-
fully at this group, and found that they were more intelligent than the remaining 117
(two thirds of the Asperger’s group had normal IQ, whereas only 50% in the group
of cases meeting Childhood Autism/Atypical autism criteria, but not our criteria for
Asperger’s, had normal IQ). There was significant difference also with regard to lan-
guage estimate (the Gillberg Asperger group had much better language develop-
ment). These were really the only major findings that differentiated the Gillberg
Asperger syndrome cases from the other cases of Autism and Atypical Autism in the
whole study. Interestingly, the majority of those with Gillberg’s Asperger syndrome
also met full criteria for childhood autism.

So, in conclusion, there is no good evidence, I think, that so-called high func-
tioning autism and Asperger’s are different disorders. When we use these concepts,
we are probably referring to the same group of individuals, and depending on where
we work, and what history we have, we tend to use one diagnosis more than the
other. The only possible difference that we could tease out from this rather large scale
study was in regard to IQ or verbal IQ, and even that could only distinguish between
groups, not between individuals.


Most cases of Asperger syndrome and high-functioning autism are unlikely to be

caused by one set of identical factors. The terms are more likely synonyms than
labels for different disorders. We may be looking for genetic factors or brain damage
factors, or other factors, not yet studied, but it is unlikely, particularly with the range
of comorbidity demonstrated that we are dealing with one set of identical factors.
Overlapping and shared brain circuitries are likely to be involved in autism and
some of its overlapping and comorbid conditions. For instance, the link with ADHD,
DCD and Tourette syndrome suggests a common dopamine abnormality, and the
link with affective disorders and the obsessive compulsive symptoms suggests sero-
tonin dysfunction. Specific disorders, such as tuberous sclerosis and certain chromo-
somal disorders may provide very important links when it comes to the location of
the brain lesion and also to the fine mapping of genetic factors.
The label Asperger syndrome may, to some purists, present a major obstacle:
why use it if it is only another word for high-functioning autism? Asperger syndrome
was the term introduced by Lorna Wing at a time when very few people realised that
autism stretches far beyond the boundaries of the condition outlined by Kanner. It
has been very useful in that it has served as an eye-opener. In the old days, mere
mention of the word autism produced images of extreme impairment, non-speaking,
completely withdrawn individuals. You could not tell parents or teachers that this
high-functioning child had that. Asperger syndrome was more “neutral”. By using the
term Asperger syndrome for two decades now, we have moved forward in the field
of autism. It is now seen as a spectrum of conditions in which there are severe and
milder variants, and a wide variety of clinical presentations. I doubt that this leap for-
ward would have been accomplished so quickly if the term had not been there to
help us disregard the long shadow of the past that was cast by those who insisted on
the old myths about “classic autism”.


American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders. Fourth
Edition. Washington, D.C.: American Psychiatric Association, 1994.
Annerén G, Dahl N, Uddenfelt U (1995) Letter to the Editor: Asperger syndrome in a boy
with a balanced de novo translocation: t(17;19)(p13.3; p11). American Journal of Medical
Genetics 56: 1–8.

Aronson M, Hagberg B, Gillberg C. Attention deficits and autistic spectrum problems in chil-
dren exposed to alcohol during gestation: a follow-up study. Developmental Medicine and
Child Neurology 1997; 39: 583–587.
Asperger H. Die autistischen Psychopathen im Kindesalter. Archiv für Psychiatrie und
Nervenkrankheiten 1944; 117: 76–136.
DeLong GR, Dwyer JT. Correlation of family history with specific autistic subgroups:
Asperger’s syndrome and bipolar affective disease. Journal of Autism and Developmental
Disorders 1988; 18: 593–600.
Ehlers S, Gillberg C. The epidemiology of Asperger syndrome. A total population study.
Journal of Child Psychology and Psychiatry 1993; 34: 1327–1350.
Ehlers S, Nydén A, Gillberg C, Dahlgren-Sandberg A, Dahlgren S-O, Hjelmquist E, et al.
Asperger syndrome, autism and attention disorders: A comparative study of the cognitive pro-
file of 120 children. Journal of Child Psychology and Psychiatry 1997; 38: 207–217.
Gillberg C. Asperger syndrome in 23 Swedish children. Developmental Medicine and Child
Neurology 1989; 31: 520–531.
Gillberg C. Clinical and neurobiological aspects of Asperger syndrome in six family studies.
In: Gillberg C (ed(s)). Autism and Asperger Syndrome. Cambridge: Cambridge University Press.
1991. 122–146.
Gillberg C. The treatment of epilepsy in autism. Journal of Autism and Developmental Disorders
1991; 21: 61–77.
Gillberg C. The Emanuel Miller Memorial Lecture 1991: Autism and autistic-like conditions:
subclasses among disorders of empathy. Journal of Child Psychology and Psychiatry 1992;
33: 813–842
Gillberg C, Billstedt E. Autism and Asperger syndrome: the coexistence with other clinical dis-
orders. Acta Psychiatrica Scandinavica 2000; 102: 1–10.
Gillberg C, Coleman M. The Biology of the Autistic Syndromes. Third Edition. London:
Cambridge University Press, 2000.
Gillberg C, Ehlers S. High-functioning people with autism and Asperger syndrome. A litera-
ture review. In: Gillberg C and Ehlers S (ed(s)). Asperger Syndrome or High
Functioning Autism? Special Issue of Journal of Autism and Developmental Disorders. New
York: Plenum Press. 1998. 79–106.
Gillberg C, Steffenburg S, Jakobsson G. Neurobiological findings in 20 relatively gifted chil-
dren with Kanner-type autism or Asperger syndrome. Developmental Medicine and Child
Neurology 1987; 29: 641–649.
Gillberg IC, Bjure J, Uvebrant P, Gillberg C. SPECT (Single Photon Emission Computed
Tomography) in 31 children and adolescents with autism and autistic-like conditions.
European Child & Adolescent Psychiatry 1993; 2: 50–59.
Gillberg IC, Gillberg C. Asperger syndrome—some epidemiological considerations: a research
note. Journal of Child Psychology and Psychiatry 1989; 30: 631–638.
Gillberg IC, Gillberg C, Ahlsén G. Autistic behaviour and attention deficits in tuberous sclero-
sis. A population-based study. Developmental Medicine and Child Neurology 1994; 36: 50–56.
Happé F. An overview of the psychology of autism. Papers read at the 6th International Congress
Autism-Europe. Glasgow, May 2000.
Happé F, Ehlers S, Fletcher P, Frith U, Johansson M, Gillberg C, et al. “Theory of mind” in the brain.
Evidence from a PET scan study of Asperger syndrome. NeuroReport 1996; 8: 197–201.
Hellgren L. Psychiatric Disorders in Adolescence. Longitudinal Follow-up Studies of Adolescent
onset Psychoses and Childhood onset Deficits in Attention Motor Control and Perception. M.D.
Thesis. Göteborg University. 2000.
Herold S, Frackowiak RSJ, Le Couteur A, Rutter M, Howlin P. Cerebral blood flow and
metabolism of oxygen and glucose in young autistic adults. Psychological Medicine 1988;
18: 823–831.
Horwitz B, Rumsey JM, Grady CL, Rapoport SI. The cerebral metabolic landscape in autism.
Intercorrelations of regional glucoseutilization. Archives of Neurology 1988; 45: 749–755.

Kadesjö B, Gillberg C. The comorbidity of ADHD in the general population of swedish school
age children. Journal of Child Psychology and Psychiatry 2000; In press.
Kadesjö B, Gillberg C, Hagberg B. Brief report: Autism and Asperger syndrome in Seven-
Year-Old Children. A Total Population Study. Journal of Autism and Developmental
Disorders 1999; 29: 327–331.
Kopp S, Gillberg C. Selective mutism: A population-based study: Research note. Journal of
Child Psychology and Psychiatry 1997; 38: 257–262.
Landgren M, Pettersson R, Kjellman B, Gillberg C. ADHD, DAMP and other neurodevelop-
mental/neuropsychiatric disorders in six-year-old children. Epidemiology and comor-
bidity. Developmental Medicine and Child Neurology 1996; 38: 891–906.
Leekam S, Libby S, Wing L, Gould J, Gillberg C. Comparison of ICD-10 and Gillberg’s cri-
teria for Asperger syndrome. Autism 2000; 4: 11–28.
Magnusson P, Gunnarsdóttir K. The epidemiology of autism and Asperger syndrome in Iceland:
A summary of a pilot study. Arctic Medical Research 1994; 53, Suppl. 2: 464–466.
Miller JN, Ozonoff S. Did Asperger’s cases have Asperger Disorders? A research note. Journal
of Child Psychology and Psychiatry 1997; 38: 247–251.
Niklasson L, Rasmussen P, Óskarsdóttir S, Gillberg C. The chromosome 22q11 deletion
syndrome (CATCH 22)—Some neuropsychiatric and neuropsychologic aspects.
Developmental Medicine and Child Neurology 2000; Submitted.
Nydén A. Autism spectrum disorders. Developmental, cognitive, and neuropsychological
aspects. M.D. Thesis. Göteborg. Department of Child and Adolescent Psychiatry. Göteborg
University. 2000. pp 63.
Nydén A, Gillberg C, Hjelmquist E, Billstedt E. Neurocognitive abilities in boys with Asperger
syndrome, attention disorder, and reading and writing disorder: Stability over time.
Developmental Medicine and Child Neurology 2000; Accepted.
Piven J, Nehme E, Simon J, Barta P, Pearlson G, Folstein SE. Magnetic resonance imaging in
autism: measurement of the cerebellum. Biological Psychiatry 1992; 31: 491–504.
Rickarby G, Carruthers A, Mitchell M. Brief report: biological factors associated with
Asperger syndrome. Journal of Autism and Developmental Disorders 1991; 21: 341–348.
Rumsey JM, Rapoport JL, Sceery WR. Autistic children as adults: psychiatric and behavioral
outcomes. Journal of The American Academy of Child Psychiatry 1985; 24: 465–473.
Ssucharewa GE. Die schizoiden Psychopathien im Kindesalter. Monatsschrift für Psychiatrie and
Neurologie 1926; 60: 235–261.
Szatmari P, Brenner R, Nagy J. Asperger’s syndrome: a review of clinical features. Canadian
Journal of Psychiatry 1989; 34: 554–560.
Wentz Nilsson E, Gillberg C, Gillberg IC, Råstam M. Ten Year Follow-up of Adolescent-Onset
Anorexia Nervosa: Personality Disorders. Journal of the American Academy of Child and
Adolescent Psychiatry 1999;38: 1389–95
Wing L. Asperger’s syndrome: a clinical account. Psychological Medicine 1981; 11: 115–129.
Wolff S. Loners. The Life Path of Unusual Children. London: Routledge, 1995.

Mailing Address
Christopher Gillberg, M.D., Ph. D.
Professor of Child and Adolescent Psychiatry
Department of Child and Adolescent Psychiatry
Göteborg University
Kungsgatan 12
411 19 Göteborg
A System of Classification for Autism
Spectrum Disorder

Peter E. Tanguay, M.D., Julia Robertson, M.D., Ann

Derrick, A.R.N.P., M.S., C.S.

Abstract. Although there is no conclusive evidence that autism is a spectrum disorder, the
concept can be clinically useful in classifying persons who show a broad range of social com-
munication deficits. We have used the Autism Diagnostic Interview—Revised and the Autism
Diagnostic Observation Schedule to assess affective reciprocity, joint attention and theory of
mind in persons diagnosed as having autism, Asperger’s Disorder and Pervasive Developmental
Disorder—Not Otherwise Specified. Persons with more severe forms of autism show marked
deficits in all three domains, while milder forms of autism may primarily show handicaps in
social cognition and joint attention. We present a case study of an autistic child who had some
affective reciprocity, but poor joint attention and imaginative play. Based on his skills and
handicaps, we suggest what specific interventions would be indicated. A social communication
model may help to quantify core social communication deficits in the autism spectrum.

Although case descriptions of children whom we recognize as possibly autistic have

appeared from time-to-time in the early psychiatric literature of the last hundred
years (Barr, 1898), it was only in 1943 that a description of what would come to be
known as autism was independently delineated by Kanner (1943) and by Asperger
(1991). Descriptions of “psychotic” children were presented in the 1950’s and 1960’s
under the terms autism, childhood schizophrenia, atypical psychosis and various
other labels. The first official classification of autism did not appear in DSM systems
until into the 1970’s (American Psychiatric Association, 1980). This classification
was based upon a small number of well-executed epidemiological studies of chil-
dren with early onset psychoses. Although the clinical description of autism has
been refined through further epidemiological work into the DSM-IV classification
(American Psychiatric Association, 1994), its current official definition is similar to
that reported in DSM-III. Like other conditions in DSM-III and DSM-IV, the diag-
nosis is a categorical one, based upon a person’s meeting a set of diagnostic criteria
whose number and degree of severity are specified. The category of autism has been
found to be one of the most robust in the DSM system in terms of its diagnostic
specificity and sensitivity (Volkmar et al., 1994), and it continues to serve us well for
research purposes.


Over the past twenty years there has been a growing recognition among clini-
cians that some persons might have symptoms and handicaps similar in nature to
those found in autism, but in a way that does not meet the full DSM criteria for the
diagnosis. Other categories, both official and unofficial, have been suggested in the
DSM-IV and ICD-10. These include Pervasive Developmental Disorder—Not
Otherwise Specified (PDD-NOS) and Asperger’s Disorder, while “high-functioning
autism” (Eisenmajer et al., 1996), Non-Verbal Learning Disorder (Klin, Volkmar,
Sparrow, Cicchetti, & Rourke, 1995), Pragmatic-Semantic Disorder (Mayes,
Volkmar, Hooks, & Cicchetti, 1993), Multiple Complex Developmental Disorder
(Van-der et al., 1995) and “atypical autism” are used unofficially among clinicians
and scientists. The persons so designated may not be fully autistic, but they are usu-
ally so handicapped that they often cannot function in society and require a lifetime
of special services. Recently, population-based studies have suggested that the
“broader” classification of autistic-like disorders may have a prevalence rate of
1/1000 or even 1/500, as compared to a rate of 1/2000 for “classical” autism
(Fombonne, Du, Cans, & Grandjean, 1997; Honda, Shimizu, Misumi, Niimi, &
Ohashi, 1996).
If we are to move ahead in understanding these early-onset developmental dis-
orders, we may need to re-think the categorical approach of the DSM-IV and the
ICD-10 in favor of an approach which would allow classification of autism in terms
of a spectrum of severity. Although some family studies of persons with autism
have suggested that there might be a range of social oddities or increased cognitive
and language problems in the relatives of persons with autism (Bailey et al., 1995;
Boutin et al., 1997; Piven et al., 1994; Szatmari et al., 1996), there are no data to
prove that autism is, in fact, a spectrum disorder. It might be that what we see as a
“spectrum” is a number of distinct conditions whose apparent overlap is clinically
artificial. Only when we have a much better understanding of the underlying
genetic and environmental factors which influence the clinical manifestations of
autism will it be possible to know whether at least some forms of Pervasive
Developmental Disorder (PDD) may be described along a spectrum of severity, or
whether most forms of PDD will be found to represent specific conditions such as
Fragile-X Syndrome, Rett’s Disorder or other genetically-influenced entities
wherein a small set of genetic abnormalities can lead to a variety of broad clinical
The results of recent multi-center family pedigree studies have been reported to
be compatible with a model in which genetic defects in autism occur at a large num-
ber of loci, perhaps 15 or more (Risch, Spiker, Lotspeich, Nouri, & Hinds, 1999). If
this is true, then it may be unlikely that we will be able to closely match phenotypic
expressions of autism with specific gene abnormalities. The outcome of this research
may be a system in which a specific gene abnormality will be understood as confer-
ring a degree of susceptibility to a specific deficit or symptom in autism.
Currently, our best practical approach to characterizing autism may be as a spec-
trum disorder. Aside from its immediate clinical usefulness, study of the “spectrum”
model might lay the groundwork for defining those phenotypic characteristics of the

population which might eventually be linked to those genes and gene products which
do play a role in autism.
But a spectrum of what? One suggestion derives from studies of social com-
munication in childhood which have been carried out in the past several decades
(for a review see Sigman & Capps, 1997; Baron-Cohen S, Tager-Flusberg, &
Cohen, 1993). This literature is already well-developed, and while there are many
arguments as to the nature of the phenomena in question, certain observations are
agreed upon. Children are born with what appear to be innate behavioral propen-
sities to interact with others, using facial expression, gesture, prosody and “body
language”. Within the first year of life, and before the development of speech or
of propositional thought, children learn a rich repertory of spontaneous social
interactions with their caregivers. Children with severe autism, even if they have
apparent normal intellectual capacity, fail to show these behaviors (Charman,
1997; Mundy, 1995; Lewy & Dawson, 1992). Possibly they come into the world
without the necessary behavioral propensity, or perhaps they have the propensity
but their threshold for perceiving facial, gestural and verbal cues is so high that
the latter are ineffective.
For the past several years, we have been attempting to develop a method to
clinically conceptualize and measure social communication in children who have
autism and related forms of PDD. Our measurement instruments have been the
Autism Diagnostic Interview—Revised (ADI-R) (Lord, Rutter, & Le, 1994) and
versions of the Autism Diagnostic Observations Schedule (ADOS) (Lord et al.,
1989; DiLavore, Lord, & Rutter, 1995). We chose these instruments because both
contain many questions about or observations of the elements of social communi-
cation in children. We found (Tanguay, Robertson, & Derrick, 1998; Robertson,
Tanguay, L’Ecuyer, Sims, & Waltrip, 1999) that a factor analysis of a subset of the
social-communication items from each of the instruments yielded three domains.
The factors mirrored phenomena described in the social communication litera-
ture: affective reciprocity, joint attention and theory of mind. We believe that the three
domains are not specific stages in development, but merely a way which the field
has chosen to characterize the salient characteristics seen in what is otherwise a
seamless developmental trajectory.

Affective Reciprocity

This encompasses the extent to which the person sends various social signals
to others, through facial expression, tone of voice and social and emotional ges-
tures. It could be seen as a type of instinctual drive whose function is to cause a
child to send social signals to others and to look for social signals. The more the
child is driven to interact with others, the more he or she can learn the meaning
of such signals. Affective reciprocity is also shown by affectionate and empathic
behaviors, greeting others with pleasure and spontaneously offering to share toys
or food with others.

Joint Attention

By 15 months of age, children are eager to share their interests with others. They
show things to others, they try to share their enjoyment, they babble reciprocally
and they direct others’ attention to objects which interest them. Later, when they
have language, they ask about others’ interests and ideas, and they share others’

Theory of Mind

Pragmatic skill represents knowledge about the “rules of social communication”.

In its simplest form, this includes knowing how to begin a conversation, how to con-
tinue a conversation, how to end a conversation and how to choose topics of conver-
sation which are appropriate to the situation at hand. Theory of mind implies the
ability to correctly infer what another person’s intentions may be. It allows one to auto-
matically and successfully respond to others while furthering, through the interaction,
one’s own needs and goals. One must have an adequate theory of mind in order to use
“good common sense”. Theory of mind is learned through years of interacting with
others. Those who lack pragmatic skills and an adequate theory of mind are com-
pletely at sea when it comes to conversing, chatting and negotiating with others.
The nature of “theory of mind” has been hotly debated (Baron-Cohen S et al.,
1993). Is it the ability to empathically put oneself into another person’s frame of mind?
Or is it a product of logical cognitive operations operating on known observations?
There is evidence that the phenomena is first seen early in life (Sorce, Emde, Campos,
& Klinnert, 1985). By age three it has been shown that children have a strong propen-
sity to classify the world of people and the world of objects in very different ways (see
Baron-Cohen’s article elsewhere in this issue). By age four children believe that peo-
ple act as a result of “motives,” while they understand that objects do not have a mind
of their own and can only respond to mechanical actions. The former phenomena has
been called “folk psychology,” while the latter has been referred to as “folk physics.”
High-functioning autistic persons appear to have a serious lack of understanding of
“theory of mind” or of a propensity to think in terms of “folk psychology” even if their
understanding of “folk physics” may appear advanced or even hypertrophic.
We have observed that children with poor affective reciprocity do not develop
good joint attention or good imagination/play; but persons with reasonable affective
reciprocity may develop variable degrees of joint attention and imagination/play.
Persons with the least degree of social communication handicap, who might be classi-
fied as Asperger’s Disorder or PDD-NOS in DSM-IV, have greater deficits in imagi-
nation and play than in joint attention. It should be noted that non-autistic
mentally retarded children, as well as children with deafness, have been shown by
previous investigators to have poor “theory of mind” and poor imagination and play
skills (Yirmiya, Solomonica-Levi, Shulman, & Pilowsky, 1996; Russell et al., 1998). It
has also been our clinical impression that while social communication can be a useful

dimension for classifying persons with autism and Asperger’s Disorder, it is inadequate
unless it is supplemented by specific information about intellectual level and verbal
(semantic-syntactic) language skills. It is also crucial that the presence of other devel-
opmental phenomena, e.g. hypersensitivity to sensory input, poor gross and fine motor
skills and unusual motor stereotypies be included in individual case descriptions.

Case Presentation

Developmental and Social History from the Autism Diagnostic

Interview—Revised (ADI-R)

William Jones is a 45-month-old child who lives with his mother and grand-
mother. His father and mother are divorced. He attends a day care program. The
program is not staffed to care for children with special needs. He runs away from the
class, ignores the other children and “refuses” to play games. Based on what she read
on the Internet, Willie’s mother suspected that he might be autistic.
When Willie was six months of age his mother recalls that he seemed responsive:
he played baby games with her and looked her in the eye. But by twelve months of
age he had stopped interacting with others. He no longer paid any attention to his
mother, was not interested in sharing objects or events with her and did not develop
social language or even social babble. By three years of age he would have periods
when he would be excited by some repetitive action he saw on TV, at which time he
would flap his hands while staring at the TV out of the corner of one eye. Currently,
he has no interest in other children and avoids their social overtures. When frus-
trated, he engages in tantrums in which he bangs his head against objects. He does
not talk spontaneously and is echolalic. He is fascinated with the Seven Dwarfs which
he watches on videos over and over again. He tries to order all his toys or objects
which he finds about the house in sevens, and he repeats fragments of what he heard
on the videotapes. These vocalizations have very limited social use.
In her replies to questions on the ADI-R, William’s mother indicated that he has lit-
tle evidence of affective reciprocity, i.e., drive or skills that would indicate that he reaches
out to others using eye contact, facial expression or gesture. What speech he has is
echolalic, with sing-song prosody. He does not nod his head yes or shake his head no.
He does not reach up to others when he wishes to be picked up, though he did some-
times echo his mother’s gesture of reaching out to him. He ignores others in distress and
fails to respond to excitement in other children, even at his own third birthday party.
His joint attention is similarly deficient. Even children who do not have language
can show extensive social interaction with their caretakers, but Willie does not. He
does not show things that interest him to others, nor does he point to others to show
them things which interest him. His lack of joint attention was said to be profound.
He appears to have little social imagination or theory of mind, inasmuch as he
engages in no imaginative play. He is described as lining up objects and being very
upset if others disturbed his lines. He is also upset if others do not put his toys away

in a particular order. He has been upset by changes in routine. A teacher who is

absent from school can send him into a state of acute anger and despondency.

Family History

A maternal uncle is quite certainly autistic by Ms. Jones’ description. His par-
ents have refused to have him evaluated, however; and he lives at home as an
adult. He is not employed. He has good language and is a walking encyclopedia
of baseball statistics and information about recorded music. He does not use this
knowledge to interact normally with others. There is also an 18-year-old paternal
cousin who may have autistic traits. He also has low intelligence and has dropped
out of school.

ADI-R Algorithm Scores

The data gathered on the ADI-R were subjected to analysis using algorithms
which come with the ADI-R for the diagnosis of autism. William’s scores were within
the range for autism in all three DSM-IV domains: impairment in reciprocal social
interaction, impairment in verbal and non-verbal communication and presence of a
markedly restricted repertoire of activities and interests.

Observations Based on the Autism Diagnostic Observation

Schedule (ADOS)

William was an attractive, well-groomed boy, looking a bit older than he is.
When the examiner greeted him and introduced herself in the lobby, he looked at
her and said “Hi.” He was playing with a Tarzan cause-and-effect toy, making a mon-
key go up and down the tree; he brought a slide over for the monkey to slide on. He
ran down the hall once then reluctantly held the examiner’s hand and his mother’s
on the way to the playroom. In the playroom he ran right to the toys to close and
open the Poppin’ Pals.
From the outset it was clear that Willie was quite echolalic, using various phrases
and words in a rote way. Despite his lack of conversational ability, however, he was
capable of using rote phrases and greetings in a way which made him initially appear
more socially in touch than he was. His echolalic phrases included “want to play with
the blocks?” and “all the little pieces.” Spontaneous phrases included: “Let’s play the
bunny,” “Here, Mom,” “I’ve got a three,” “Hey, where the bubbles go?” and “Let’s
put it on the plate.” Willie also repeated a number of apparently spontaneous
phrases, e.g., during the birthday party, he said, about five times: “Let’s play with the
cake.” His prosody was quite good (the tone, rhythm, volume, rate of his speech). His
articulation varied from crystal clear to mumbling

Affective Reciprocity

Willie demonstrated affective reciprocity but not consistently. He frequently emit-

ted normal facial expressions of affect but failed to direct them to anyone. He had a
beautiful smile and very expressive face. When the examiner inflated a balloon, he
imitated the manner in which she opened her mouth, and when she released the bal-
loon, he ran enthusiastically to pick it up and hand it back to her to blow up again.
Willie enjoyed social routines like Ring-around-the-Roses and anticipated the “all
fall down” right on cue. He enjoyed the make-believe birthday party and knew to
place the candles on the cake, sing the song and blow out the candles. He showed
eager anticipation for the routines. He was not the least bit interested in feeding or
playing with the doll, however. With much encouragement he started to give the doll
a drink but then pretended to drink himself. He did offer his mother a drink, and he
initiated a play interaction with her, saying, “Here, Mom,” and handed her a stack-
ing cup, meshing gesture and vocalization, although he did not look at her.
Although Willie clearly enjoyed some of the activities, the examiner was unable
to elicit a social smile from him. He did not point, nor did he use any other gestures.
He did not wave goodbye even with many prompts. When the examiner said “the
doll is sleepy. What’s she going to do now?”, he looked at the doll, waved his hand
around and said, “Hey, baby, wake up now.”

Joint Attention

Willie had major deficits in joint attention. Occasionally he made eye contact but
only in a fleeting and unsustained way. Even during peak activities of enjoyment,
such as blowing bubbles, he never looked to his mother to share the event with her,
nor did he make eye contact with her. He made no attempts to direct the attention
of the examiner or of his mother to what he was doing or what caught his interest. If
he needed something, he indicated it mechanically, e.g., when he wished the exam-
iner to blow up the balloon, he handed it to her without sending or looking for any
facial signals. When he wished to leave the room, he hung onto the door but did not
otherwise attempt to communicate his wish.

Theory of Mind

William readily played with the cause-and-effect toys: the Poppin’ Pals, remote
control bunny and the toy phone. He rang the bell on the phone but did not try to
use it symbolically, even when he was urged to do so with examples and gestures.
He lined up the stacking toys mechanically and recited the number printed on the
bottom of each cup. He ignored the doll, truck and cars.
As Willie and his mother left the ADOS session, he raced down the hall towards
the escalator. His mother caught him and held his hand. The examiner shouted,

“Bye, Willie.” He turned and looked at her and shouted back, “Bye, Ann,” in a burst
of reciprocal social interaction.
ADOS scores: Willie’s scores on the autism algorithm of the ADOS were better
than they had been for the ADI-R but were still within the range required for the
diagnosis of autism.

Observations from the ADOS

Despite his having fairly good affective reciprocity on the ADOS, Willie is
severely impaired in reciprocal social interaction and in theory of mind. It is not that
he failed entirely to engage with others, but he did so obliquely, using echolalic
phrases, rote routines and simple requests for help without accompanying eye con-
tact, body language or gesture. He did not seek to read meaning in the facial expres-
sion or gesture of others. His play behaviors indicated that he had some familiarity
with how objects work and an interest in manipulating them, but little propensity to
use objects in a social way. At his age children might be expected to run the cars
about on imaginary roads or feed the baby doll with the bottle. They show objects to
others and expect a response; they act as if they know that others have minds which
differ from their own and that they can be interested, amused or otherwise reactive
in response to social overtures.
Willie’s initial greeting to the examiner, as well as his farewell response (“Bye,
Ann”) were somewhat out-of-keeping with his lack of joint attention during the ses-
sion. Clearly, he could show joint attention under certain circumstances. We specu-
late that his realization that the session was over may have led to the sudden release
of his best interactional skills as he left the scene. We have noted many times that
anxiety can be a potent inhibitor of affective reciprocity and joint attention. The fact
that under certain circumstances Willie could engage in such behaviors may have
salutary implications in terms of his being able to benefit from treatment. It doesn’t
mean that he is not autistic but that his degree of autism is not profound.
There are important differences in Willie’s mother’s perceptions of him as
recorded on the ADI-R and our observations on the ADOS. Ms. Jones described
Willie as seriously impaired in all three domains of social communication on the
ADI-R. In contrast, we found him to have good affective reciprocity, as reflected in
his having normal facial expression (though he failed to direct the expression at oth-
ers); normal prosody; good use of instrumental gestures (handing the balloon to be
blown up again) and good greeting behaviors, both to the examiner and to his
mother. In our experience this difference between the ADI-R and ADOS observa-
tions is unusual, in that the differences are usually in the opposite direction: the par-
ents describe the child as more functional than we observe on the ADOS. This latter
could be a result of the child’s failing to perform on the ADOS because of anxiety or
shyness, or they might be due to the parent’s having an overly optimistic view of the
child’s abilities and behaviors. In the present instance, Willie’s mother seems to have
been overly pessimistic in her appreciation of his skills.

Other Skill Domains

Willie’s language skills (vocabulary and grammar) are delayed. He is echolalic

when it would be expected that he use socially meaningful phrases to interact with
Non-verbal IQ testing indicated that Willie was functioning in the low normal
range. This is a positive prognostic sign in terms of his future schooling, especially if
his language skills improve.
Temperament: Willie is easy-going. He is not especially negative and intense, which
will make it easier for him to fit into a classroom. He likes routines, which we attrib-
ute to his lack of social knowledge. Routines allow a person with autism to anticipate
what may be expected in a social situation, and they provide a framework around
which to build a response.
Willie is not hyper-responsive to auditory, visual or tactile sensations. This is
good, as such difficulties can add to whatever other impairments a person may have.


Willie meets the diagnostic criteria for autism, but where does he fit on the autism
spectrum? This is difficult to answer, given that our nosology does not currently
address this issue. We suggest that a four-year-old with good affective reciprocity and
beginning joint attention behaviors may be said to have a moderate degree of social
communication disorder. If Willie had had good affective reciprocity and good joint
attention, plus some imaginative play, then we might have considered him as being
mildly impaired. We are far from having a classification of social communication that
would resemble that for intellectual impairment with its mild, moderate, severe and
profound degrees.
The positive family history of autism in a maternal uncle, and perhaps in a
cousin, would be consistent with Willie’s diagnosis. His history of apparent normal
social development prior to six months can be noted but not explained.


Willie should receive speech therapy aimed at improving his language skills. He
should leave his private preschool where there are no special education teachers and
enroll in a therapeutic preschool. In his home town the school district operates a pre-
school programs for children with developmental disabilities. It is important that
Willie’s mother participate in the program since she seems to have a somewhat
overly pessimistic view of the degree of his handicaps. Parents usually find it helpful
to observe others with their child. The teacher may have play materials and tech-
niques that Ms. Jones could use at home.

If Willie were profoundly autistic and had no affective reciprocity skills, it

would be necessary for him to be in a program that focused mainly on catching his
attention and engaging him at a sensory-motor level, much as mothers do with their
normal three- to six-month-old infants in what has been called “motherese.” Willie
does not need this in that he is already responsive to others, though he has no the-
ory of mind and little knowledge about the rules of social interactions or interest in
imaginative play. His teachers, and his family, will need to focus on joint interac-
tions which capitalize on his interest in toys such as the toy telephone, the trucks and
the cars. They will need to interact with him using games involving songs, gestures,
dance and gross motor activities. He is ready for what has been called intense social-
pragmatic teaching, as opposed to massed discrete-trial learning (see Prizant & Wetherby,
1998, for a discussion of these contrasting approaches). The latter emphasize learn-
ing such things as words, facts and specific skills. Later, when Willie has better lan-
guage and joint attention skills, he will be ready for more didactic and traditional
learning approaches.


Though we hope that Willie’s affective reciprocity skills augur well in terms of his
learning language, joint attention and social knowledge, there are few studies which
have followed children’s progress in terms of their earlier level of affective reciproc-
ity and joint attention skills. Does a four-year-old with good affective reciprocity have
a better prognosis than one who has poor affective reciprocity? We don’t know,
although common sense would suggest that this might be so.
The system of classification outlined above is of interest to us because it fits with
our notions of inborn behavioral propensities and because it is oriented to what
seems to be useful clinical observations. It does not appear to stray too far from what
we conceive to be the core problems of PDD and autism. We believe the concept
deserves debate and further study.


American Psychiatric Association (1980). Diagnostic and Statistical Manual of Mental Disorders
(3rd Edition). Washington, D.C.: American Psychiatric Association.
American Psychiatric Association (1994). Diagnostic and Statistical Manual of Mental Disorders
(4th Edition). Washington, D.C.: American Psychiatric Association.
Asperger, H. (1991). ‘Autistic psychopathology’ in childhood. In Frith Uta (Ed.), Autism an
Asperger Syndrome (pp. 37–92). Canbridge: Cambridge Univerity Press.
Bailey, A., Le, C. A., Gottesman, I., Bolton, P., Simonoff, E., Yuzda, E., & Rutter, M. (1995).
Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med,
25, 63–77.
Baron-Cohen S, Tager-Flusberg, H., & Cohen, D. (1993). Understanding Other Minds. New York:
Oxford University Press.

Barr, M. W. (1898). Some notes on echolalia, with the report of an extraordinary case. Journal
of Nervous and Mental Diseases, 25, 21–30.
Boutin, P., Maziade, M., Merette, C., Mondor, M., Bedard, C., & Thivierge, J. (1997). Family
history of cognitive disabilities in first-degree relatives of autistic and mentally retarded
children. Journal of Autism & Developmental Disorders, 27, 165–176.
Charman, T. (1997). The relationship between joint attention and pretend play in autism.
[Review] [110 refs]. Development and Psychopathology, 9, 1–16.
DiLavore, P. C., Lord, C., & Rutter, M. (1995). The pre-linguistic autism diagnostic observa-
tion schedule. Journal of Autism & Developmental Disorders, 25, 355–379.
Eisenmajer, R., Prior, M., Leekam, S., Wing, L., Gould, J., Welham, M., & Ong, B. (1996).
Comparison of clinical symptoms in autism and Asperger’s disorder. Journal of the
American Academy of Child and Adolescent Psychiatry, 35, 1523–1531.
Fombonne, E., Du, M. C., Cans, C., & Grandjean, H. (1997). Autism and associated medical
disorders in a French epidemiological survey. Journal of the American Academy of Child and
Adolescent Psychiatry, 36, 1561–1569.
Honda, H., Shimizu, Y., Misumi, K., Niimi, M., & Ohashi, Y. (1996). Cumulative incidence
and prevalence of childhood autism in children in Japan [see comments]. British Journal
of Psychiatry, 169, 228–235.
Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217–250.
Klin, A., Volkmar, F. R., Sparrow, S. S., Cicchetti, D. V., & Rourke, B. P. (1995). Validity and
neuropsychological characterization of Asperger syndrome: convergence with nonver-
bal learning disabilities syndrome. Journal of Child Psychology and Psychiatry, 36,
Lewy, A. L. & Dawson, G. (1992). Social stimulation and joint attention in young autistic chil-
dren. Journal of Abnormal Child Psychology, 20, 555–566.
Lord, C., Rutter, M., & Le, C. A. (1994). Autism Diagnostic Interview-Revised: a revised ver-
sion of a diagnostic interview for caregivers of individuals with possible pervasive devel-
opmental disorders. Journal of Autism & Developmental Disorders, 24, 659–685.
Lord, C., Rutter, M., Goode, S., Heemsberen, J., Jordan, H., Mawhood, L., & Schopler, E.
(1989). Autism Diagnostic Observation Schedule: a standardized observation of com-
municative and social behavior. Journal of Autism & Developmental Disorders 19, 185–212.
Mayes, L., Volkmar, F., Hooks, M., & Cicchetti, D. (1993). Differentiating pervasive develop-
mental disorder not otherwise specified from autism and language disorders. Journal of
Autism & Developmental Disorders, 23, 79–90.
Mundy, P. (1995). Joint attention and social-emotional approach behavior in children with
autism. Development and Psychopathology, 7, 63–82.
Piven, J., Wzorek, M., Landa, R., Lainhart, J., Bolton, P., Chase, G. A., & Folstein (1994).
Personality characteristics of the parents of autistic individuals. Psychological Medicine, 24,
Prizant, B. M. & Wetherby, A. M. (1998). Understanding the continuum of discrete-trial tradi-
tional behavioral to social-pragmatic developmental approaches in communication
enhancement for young children with autism/PDD. [Review] [98 refs]. Semantics Speech
Language, 19, 329–352.
Risch, N., Spiker, D., Lotspeich, L., Nouri, N., & Hinds, D. (1999). A genomic screen of
autism: evidence for a multilocus etiology. American Journal of Human Genetics, 65,
Robertson, J., Tanguay, P., L’Ecuyer, S., Sims, A., & Waltrip, C. (1999). Domains of social com-
munication handicap in autism spectrum disorder. Journal of the American Academy of
Child and Adolescent Psychiatry, 38, 738–745.
Russell, P. A., Hosie, J. A., Gray, C. D., Scott, C., Hunter, N., Banks, J. S., & Macaulay, M. C.
(1998). The development of theory of mind in deaf children. Journal of Child Psychology
and Psychiatry, 39, 903–910.

Sigman, M. & Capps, L. (1997). Children with Autism - A Developmental Perspective. Harvard
University Press, Cambridge.
Sorce, J., Emde, R., Campos, J., & Klinnert, M. (1985). Maternal emotional signaling: its effect
on the visual cliff behavior of 1-year olds. Developmental Psychology, 21, 195–200.
Szatmari, P., Jones, M. B., Holden, J., Bryson, S., Mahoney, W., Tuff, L., MacLean, White, B.,
Bartolucci, G., Schutz, C., Robinson, P., & Hoult, L. (1996). High phenotypic correla-
tions among siblings with autism and pervasive developmental disorders. American
Journal of Medicine & Genetics, 67, 354–360.
Tanguay, P. E., Robertson, J., & Derrick, A. (1998). A dimensional classification of autism spec-
trum disorder by social communication domains. Journal of the American Academy of Child
& Adolescent Psychiatry, 37, 271–277.
Van-der, G. R., Buitelaar, J., Van-den, B. E., Bezemer, M., Njio, L., & Van, E. H. (1995). A con-
trolled multivariate chart review of multiple complex developmental disorder.
J.Am.Acad.Child Adolesc.Psychiatry, 34, 1096–1106.
Volkmar, F. R., Klin, A., Siegel, B., Szatmari, P., Lord, C., Campbell, M., Freeman, B. J.,
Cicchetti, D. V., Rutter, M., Kline, W., Buitelaar, J., Hattab, Y., Frombonne, E., Fuentes,
J., Werry, J. S. , Stone, W. L., Kerbeshian, J., Bregman, J. D., Loveland, K. A.,
Szymanski, L., & Towbin, K. (1994). Field trial for autistic disorder in DSM-IV. Am J
Psychiatry, 151, 1361–1367.
Yirmiya, N., Solomonica-Levi, D., Shulman, C., & Pilowsky, T. (1996). Theory of mind abili-
ties in individuals with autism, Down syndrome, and mental retardation of unknown eti-
ology: the role of age and intelligence. Journal of Child Psychology & Psychiatry & Allied
Disciplines, 37, 1003–1014.

Mailing Address
Peter E. Tanguay, M.D., Julia Robertson, M.D., Ann Derrick, A.R.N.P., M.S., C.S.
Department of Psychiatry and Behavioral Sciences
Division of Child and Adolescent Psychiatry
University of Louisville
Bingham Child Guidance Center
200 E. Chestnut St.
Louisville, KY 40202
The Core Deficit in Autism and Autism
Spectrum Disorders

Nancy J. Minshew, M.D.

Abstract. This paper reviews the evidence that autism is a disorder of complex information
processing with intact simple information processing. This constellation explains why these par-
ticular symptoms travel together as a syndrome, the failure of intact skills to predict higher order
skills, and the common “co-occurrence” of mental retardation with autism. Complex informa-
tion processing does not refer to a particular ability, but is a concept that identifies a common
characteristic shared by all deficits. This concept is important because it highlights the funda-
mental importance to educational and behavioral treatment of simplifying tasks to reduce infor-
mation processing demands. It also is equally applicable in characterizing structural and
functional abnormalities of the brain, aiding in making links across levels from genes to behav-
ior. The neural basis of this cognitive profile was investigated in eye movement studies, which
demonstrated that deficits are the result of the selective failure of the most advanced levels of
brain circuitry to develop.

Ever since autism was determined to be of neurobiologic origin theories regarding

the core deficits and their neural origin have abounded. In the 1960’s and into the
1980’s, theories hypothesized a single primary cognitive deficit as the underlying
cause of the behavior in autism. The prevalent theories proposed deficits in the
acquisition of information, i.e., in elementary sensory perceptual, attentional, and
memory abilities, and localized dysfunction of brainstem and hippocampus
(Minshew & Payton, 1988 for review). Because phenylketonuria, fetal rubella, and
tuberous sclerosis had been discovered as causes for some cases, it was also hypoth-
esized that other diseases would ultimately be discovered that were responsible for
the remaining cases. Subsequent research greatly altered these views. It is now gen-
erally but not entirely appreciated that the behavioral syndrome of autism is the
result of multiple primary deficits and that these deficits involve the processing of
information and are the result of the underdevelopment of the neural systems of the
forebrain and not regional dysfunction (Minshew, 1997 for review). Although 5–10%
of cases will be found to be the result of other diseases, the majority of cases are
thought to be the result of 4–6 or more abnormal genes coding for or regulating brain
development (Cook, 1998; Rutter et al, 1994). It is also likely that it will not be the



same 4–6 genes in each case and that there will be a large number of genes that can
lead to autism. The diseases that cause that 5–10% of cases are thought to produce
the same neurologic and cognitive abnormalities triggering the same pathophysiol-
ogy but by a different mechanism. This concept of the cause of autism is depicted in
Figure 1. The gene abnormalities produce a protein or proteins that alter brain devel-
opment (developmental neurobiology). The outcome of this is the structural and
functional abnormalities of the brain studied by researchers. These underlie the cog-
nitive and neurologic deficits that cause the abnormal behavior that is called autism,
autism spectrum disorder, or pervasive developmental disorder.

Cognitive & Neurologic Basis of Behavior

The purpose of this discussion is to consider the cognitive and neurologic basis of
behavior. The importance of this issue is multi-fold. First and foremost is that clarifi-
cation of the cognitive and neurologic basis of behavior empowers behavioral, phar-
macologic and academic intervention. The efficacy of behavioral and pharmacologic
intervention is greatly improved by understanding the motive or drive behind the
behavior. Interventions can then be designed to address the cause of the behavior in
the individual. Improving the understanding of the cognitive and neurologic basis of
behavior also leads to improvements in diagnostic criteria, as evidenced by the evo-
lution of the diagnostic criteria for autism in the last three versions of the Diagnostic
and Statistical Manual for Mental Disorders (DSM). The improved characterization of
the deficits will also stimulate the development of new assessment instruments to
guide the evaluation of behavior problems. Third, the definition of the cognitive and
neurologic deficits identifies the abnormalities that must be explained by the genetic
and developmental neurobiologic mechanisms; these mechanisms hold the key to the
development of biologic interventions aimed at restoring brain development. In order
for that goal to be achieved, each of the levels involved in the cause of this disorder
(see Figure 1) and their inter-relationships must be precisely and accurately defined or
researchers will identify the wrong gene or wrong developmental brain mechanism.
The end result of imprecision is that we will be “on the right street but in the wrong
garage” in the search for a definitive treatment for autism.

A Neurologic Overview

Neurologists typically begin their assessment of a disease process by asking

where in the brain are the abnormalities and what is the mode of onset and progres-
sion. The first step is a quick orientation based on basic principles about structural-
functional relationships in the brain. The first distinction is between gray and white
matter dysfunction and cortical and subcortical dysfunction. Autism is characterized
by mental retardation, deficits in language, abnormalities in complex behavior, and
seizures. This constellation is classic for dysfunction of cerebral cortical gray matter.
The absence of deafness and blindness eliminates primary sensory cortex, leaving
association cortex. The absence of long tract signs (alterations in tone and reflexes)
eliminates white matter involvement. Thus, a brief neurologic appraisal of this clini-
cal syndrome points to deficits in higher order cognitive abilities and to association
cortex. The mode of onset is developmental.

The Pattern of Deficits & Intact Abilities

The next step involves the definition of all the deficits and intact abilities and the
determination of the pattern these follow. This pattern is critical because it identifies
a basic or defining characteristic of the brain abnormality in autism. If correctly iden-
tified, this pattern or common denominator will explain why this cluster of symptoms
occurs or “travels together.”
A number of patterns have been proposed in autism over the years. Some of
these are still debated in one arena or other. The issues debated have included: a sin-
gle primary cognitive-single neural system deficit versus multiple primary co-existing
cognitive-multiple neural systems deficit, auditory and not visual information pro-
cessing deficit, information acquisition deficit versus information processing deficit,
and simple versus complex information processing deficit(s).
These issues could be addressed with a comprehensive neuropsychologic test bat-
tery. Such a battery requires individuals who have a Full Scale and Verbal IQ score
of 70 or higher. Thus, this study spanned nearly 10 years because of the difficulty at
the time of identifying verbal individuals with autism. The addition of the Asperger’s
category and the emphasis on early intervention programs has since greatly
improved the identification of autism in verbal individuals. Table 1 contains the tests
and results used in the discriminant function analysis. Tables 2 and 3 provide the
results of the discriminant function analysis for 33 individually age, IQ, gender, and
race matched pairs of autistic and normal control adolescents and young adults.
Table 2 contains the cognitive domains that did not reveal deficits. The first two
domains on the list were the attention and sensory perception domains. Neither
domain demonstrated differences between autistic and control groups, and thus
deficits in these areas were not the cause of autism. This does not mean that some
autistic children do not have attention deficit disorder with or without hyperactivity
or sensory symptoms, just that these are not causing their autism. The absence of
deficits in attention, sensory perception, and associative memory means that infor-

Psychometric Data Used for Discriminant Analysis


Tests Entered into Prediction Equations M SD M SD

Attention Domain
WAIS-R Digit Span 9.88 3.81 10.52 2.46
Serial Digit Learning-Correct Responses 16.52 8.17 17.42 7.91
Continuous Performance Testb 0.34 0.62 0.23 0.66
Letter Cancellation-Omissions 1.09 1.63 0.45 1.00
Number Cancellation-Omissions 3.27 4.03 4.39 5.38
Sensory Perception Domain
Luria-Nebraska Tactile Scale:
Simple Touch Errors 0.29 0.55 0.17 0.48
Stereognosis Errors 0.46 0.59 0.21 0.42
Sharp-Dull Discrimination Errors 0.88 0.80 0.58 0.72
Position Sense Errors 0.00 0.00 0.08 0.41
Finger Position Errors 0.67 1.27 0.46 1.02
*Halstead-Reitan: Fingertip Number
Writing-Errors 5.38 4.30 2.79 2.84
Motor Domain
Finger Tapping-Dominant Hand 44.27 13.78 45.19 16.24
DVMI-Total Pointsc 15.42 32.43 22.18 31.69
*Grooved Pegboard-Dominant Hand-Seconds 86.73 18.30 70.67 16.03
*Trail Making A-Time in Seconds 31.52 15.81 20.45 7.99
Simple Language Domain
WAIS-R Vocabulary 9.45 3.02 9.70 2.26
K-TEA Reading Decoding 97.48 13.60 102.79 10.19
Controlled Oral Word Association (FAS)D 36.00 13.31 34.00 16.18
K-TEA Spelling 102.58 16.93 100.91 11.50
Woodcock Reading Mastery-Word Attack 107.24 11.55 103.52 15.53

TABLE 1 (Continued)


Complex Language Domain

*Woodcock Reading Mastery-Passage
Comprehension 92.27 15.04 104.27 14.34
*K-TEA Reading Comprehension 91.36 14.43 103.06 12.45
*Test of Language Competence
Metaphoric Expression 6.85 3.25 9.42 3.70
*Binet Verbal Absurdities-Raw Score 9.30 3.64 12.48 3.97
Token Test-Number Correct 18.03 2.19 18.42 5.19
Simple Memory Domain
Paired Associate Learning-Number Correct 42.55 23.13 48.76 24.21
3 Word Short Term Memorye 3.24 3.04 2.9`1 3.15
Maze Recall-Correct/Incorrect 0.42 0.61 0.52 0.57
CVLT A List-Trial 1 Number Correct 4.50 3.90 6.30 3.90
Complex Memory Domain
Paired-Associates-Delayed-Recall 16.00 7.46 17.45 6.13
CVLT A List-Long Delay 7.00 5.49 9.00 5.55
*Nonverbal Selective Reminding-CLTR 19.94 15.09 37.39 16.09
*WMS-R Logical Memory-Delayed Recall 5.58 5.79 8.45 6.02
*Rey-Osterrieth Figure-Delayed Recall 16.83 8.58 21.94 7.49
Reasoning Domain
Trail Making B-time in Seconds 65.48 37.19 52.42 23.31
Halstead Category Test-Errors 46.24 28.71 40.73 22.46
Wisconsin Card Sorting
Test-Perseverative Rrrors 16.45 15.48 13.27 11.13
*Binet Picture Absurdities-Raw Score 20.00 11.46 27.52 6.12
t-test results are provided to identify domains not entered into the equations but still demonstrating a sig-
nificantly different performance between groups; bMean reaction time Correct Responses; cDevelopmental
Test of Visual-Motor Integration; dNumber of Words; eNumber of Correct Sequences; *Tests with statisti-
cally significant results

Discriminant Function Analysis: Domains Without Deficits


Attention Letter Cancellation; Number Cancellation 66.7 .33

Sensory Finger Tip Writing; Luria-Nebraska 64.6 .29
Perception Sharp/Dull Tactile Scale item
Simple K-TEA Reading; K-TEA Spelling 71.2 .422
Language WRMT-R Attack; Controlled Oral
Word Association
Simple CVLT Trial 1 65.2 .30
Visuo-Spatial WAIS-R Block Design 56.1 .12
Kappa below .40 indicates poor agreement beyound chance
Significant Kappa reflects superior performance by autistic subjects

Discriminant Function Analysis1: Domains With Deficits


Motor Grooved Pegboard; Trail Making A 75.8 .52

Complex Language K-TEA Reading Comprehension; Verbal 72.7 .45
Absurdities; Token Test
Complex Memory Nonverbal Selective Reminding-Consistent 77.3 .55
Long Term Retrieval; WMS-R Story
Recall-Delayed Recall; Rey-Osterrieth
Reasoning 20 Questions; Picture Absurdities; Trail 75.8 .52
Making B
Based on 33 individually matched pairs of autistic and control subjects (Neuropsychologic Functioning
in Autism: Profile of a Complex Information Processing Disorder, JINS, 3:303-316, 1997)

mation acquisition is intact. They also did not have deficits in visuospatial process-
ing. The remaining domain, simple language, is essentially what is called formal lan-
guage or the capacity to speak freely, spell, and read words or sounds without
comprehension. It is notable that the autism group did significantly better than age,
IQ, and gender matched controls. Thus, their formal language abilities were superior
to their age and IQ. This higher expressive language ability will cause listeners to
over-estimate the language abilities of autistic individuals. This will exacerbate their
language comprehension deficit, which is revealed in Table 3.
Table 3 contains the domains that revealed deficits in the autism group. The first
of these is the motor domain and the tests responsible were those involving complex
motor actions or motor sequences as opposed to isolated motor movements such as
finger tapping. This domain was included not only to complete the survey of major

cognitive domains but also because dysfunction in higher brain regions often pro-
duces problems with complex motor sequences. This is generally called motor
apraxia. This motor deficit has been found by many other investigators and motor
skills deficits are now being recognized as an integral part of autism and the other dis-
orders in this category. In young children, motor apraxia causes the problems with
operating mechanical devices such as door knobs and wind up toys, whereas button
pressing is not a problem. Later, motor apraxia is responsible for difficulty holding
pencils, cutting with scissors, and tying shoe laces. In school age children, it is
responsible for the problems with handwriting, either slowness or sloppiness. In the
gross motor area, it is responsible for the lack of coordination in sports and con-
tributes to the inability of many to ride bicycles or rollerblades. It is likely also
responsible for the stilted quality to facial expressions.
The second domain revealing deficits in the autism group is that of complex lan-
guage, i.e., the interpretative aspects of language. These include reading compre-
hension, story comprehension, comprehension of idioms and metaphors, verbal
inference making, and comprehension of complex sentence structure. The latter is
particularly important because it is the language of everyday life. It typically involves
sentences like: before you do this, I want you to do that and then do thus and so.
These sentences place particularly heavy demands on information processing
because they require processing of each segment and then a second stage of pro-
cessing to determine the meaning of each bit to the next bit. Understanding the gist
of such a sentence requires yet a third level of processing. A similar process occurs
with the understanding of a story. The combination of superior formal language abil-
ities and inferior comprehension produces a wide gap between the listener’s estimate
of the autistic individual’s language comprehension and his or her actual compre-
hension. The failure to understand what the person with autism understands is a
major contributor to their dysfunction in many settings.
The next affected domain is that of memory for complex information.
Complexity can result from increasing amounts of simple information or increasing
inherent complexity of the information. In essence, individuals with autism have dif-
ficulty with recall of complex material, because they fail to make use of cognitive
organizing strategies or to benefit from the meaning of the material. Secondly, this
pertained to both visual and auditory information. The ramification of this deficit is
that they remember less from the material presented to them than age and IQ
matched peers; this has also been shown to reduce the amount of information they
remember from recently experienced events. Thus, this memory impairment is likely
to contribute to the social, language, and problem solving deficits. Knowledge of this
impairment can be used to improve learning. Memory and learning can be improved
by reducing the amount of material presented (smaller chunks), preprocessing the
information (give the bottom line rather than exemplars that require deduction of the
bottom line), and increasing the processing time. Visual presentation of information
often accomplishes all of these, and likely explains why they benefit from such adap-
tations. Visual material is constantly present for reference and re-reading, to guide

The last domain revealing deficits is that of abstract reasoning. This has long been
known to be an area of deficit in autism. The notable finding of this study is the dis-
sociation between intact performance on rule-learning tasks and deficits on concept
formation tasks. Rule-learning tasks are ones in which there is a rule to solve the prob-
lem and the task is to discover the rule. Although the individuals in this study typi-
cally identified the rule correctly, they had difficulty changing rules when the context
changed. Changing the rule to fit changing contexts adds to the information process-
ing demands of the task, and is the basis of generalization. Concept formation tasks
have no set solutions but require the individual to create one. Concept formation tasks
are essentially problem solving in novel situations. This pattern is consistent with the
behavior typical of autism, which is typically rule-dependent, lacking in flexibility, fail-
ing to consider the implications of context, and inability to cope with novel situations.
These were the findings by domain and the next issue was to determine the pat-
tern or common denominator to the deficits and to the intact abilities. Some features
of the pattern were relatively obvious. Information acquisition and visuospatial pro-
cessing were intact. Second, there were multiple co-existing deficits in information
processing abilities in both the visual and auditory modalities. The challenge was to
determine what the significance was of having intact abilities within the same
domains as deficits. That is, elementary motor, associative memory, formal language
and rule-learning abilities were intact in the presence of deficits in skilled motor,
complex memory, interpretative language, and concept formation abilities. The com-
mon denominator of the intact abilities was that they involved the more basic abili-
ties in each domain and the deficits selectively involved the highest order abilities
within the same domains. Thus, it appeared that the deficits were in abilities with the
highest information processing or computational demands. The intact abilities cor-
responded to abilities that were deficient in the prototypic simple information pro-
cessing disorder, i.e., selective language impairment. Thus, the neuropsychologic
profile in high functioning autistic individuals was the predictable converse of that
seen in selective language impairment, which attested to its validity. Thus, the pat-
tern in autism was that of selective deficits in complex information processing with
intact simple information processing. Complex information processing provided an
explanation for why the symptoms of autism “travel” together and also for the pat-
tern of deficits within domains.

Implications of the Pattern

The implications of this pattern were several. First, it differentiates autism from
general mental retardation in which both simple and complex abilities are more
evenly affected. This pattern also distinguishes autism from dyslexia and the other
learning disabilities. The traditional learning disabilities involve the simple abilities
but spare the complex abilities, which often provide top-down compensation for the
deficits. The pattern also distinguishes autism from the nonverbal learning disability
syndrome, which has deficits in visuospatial and arithmetic abilities. Third, this

pattern answers the frequently asked educational question of why they can do some
things well but do other things poorly. This is not so different from dyslexia, in which
there are intact and deficient skills; it is just that there is little familiarity with the cog-
nitive pattern associated with autism.
This pattern also explains some common clinical phenomena in autism. It has
long been known that as IQ scores decline across the autistic population, there is a
disproportionate or selective decline in social, communication and abstraction abil-
ities. It also explains the dissociation between IQ scores and Vineland scores, which
measure adaptive behavior or real life behavior. The corollary of this pattern is that
intact abilities cannot be used to predict the status of higher order abilities. And,
most importantly, it explains the common co-occurrence of mental retardation and
autism. Mental retardation is the predictable consequence of a progressive reduction
in complex information processing abilities, i.e., the capacity to attach meaning to

Generalizability of the Pattern

The demonstration of this pattern necessitated the assessment of verbal individ-

uals with autism. However, this pattern is equally apparent in lower ability autistic
individuals as revealed by the studies of Rapin and colleagues (1996). The particu-
lar tests that are failed will change such that lower ability autistic individuals may
have deficiencies on tests that non-mentally retarded autistic individuals do well on,
because their overall level of function is lower. If the concept of the pattern is
retained, though, it can be applied across the ability range with ability and age
appropriate tests; if the pattern is defined in terms of specific tests, then it cannot be
applied to younger or lower ability subjects. Also, if only one domain is assessed or
only a few tests are given within each domain, then the pattern will obviously not
be discernable.

Complex Information Processing As A Construct

Complex information processing is not a specific ability but a term for a class of
abilities that place high computational demands on the brain. It is a conceptual con-
struct. Deficits in specific abilities such as theory of mind and executive function that
are commonly discussed in connection with autism all fall under this general con-
struct. The value of this construct is that it emphasizes the need to evaluate tasks autis-
tic individuals cannot do in terms of the computational demands on the brain. This
approach provides guidelines for modifying the demands of tasks that individuals are
unable to do. However, the scientific value is that this term is also used in the neuro-
physiology to characterize the late or cognitive potentials. This construct also makes
it possible to think in terms of developmental processes in the brain involved in the
emergence of the intricate circuitry of the forebrain that underlies these abilities. Thus,

the complex information processing construct makes it easy to relate or link findings
across several levels of the pathophysiology of autism. Such links are critical if the
cause of autism is to be understood.

Neural Basis of Behavior

A second issue of great importance in understanding autism is the demonstra-

tion of the brain abnormalities that are responsible for the cognitive deficits under-
lying behavior. There have been a great many theories, many of them very
appealing, but few with any supporting data. One particular debate has been over
the cognitive and neural basis of the abnormal shifting attention behavior in autism.
Courchesne and colleagues have contended that this deficit relates to abnormalities
at the attentional level and dysfunction of the cerebellum. However, Ozonoff and
many others have demonstrated deficits in the executive control over attention and
intact basic attentional abilities. Executive abilities reside in the frontal lobes of the
brain, not the cerebellum.
Resolution of this debate depends on decomposing tasks into their component
cognitive processes and examining each individually. This has been done using three
different methods, all with the same results. Ozonoff et al (1994) decomposed the
Wisconsin Card Sorting Test into its individual components and found that autistic
individuals had no difficulty with simple shifts of attention but did have trouble with
the cognitive abilities of flexibility and with inhibition of prepotent responses, which
are related to frontal lobe function. In a detailed examination of shifts of attention,
Pasavalqua et al (1998) found no deficits in shifting attention abilities in autism but
did find deficits in the executive control of attention. In a review of attention research
in autism, Burack and colleagues (1997) reported that deficits in shifting attention or
orienting pertained to difficulties processing the directional cues and not the basic
capacity for shifting attention.
Saccadic eye movements provide an excellent method for addressing these con-
troversies. Visually guided saccades involve simple shifts of attention which reflect
cerebellar input. Volitional saccades to antisaccade and oculomotor delayed response
tasks assess the executive control over attention and are subserved by the frontal
lobes. The value of these methods is that localization is exceedingly well documented,
the tasks are similar to real life behavior, and the data are a rigorous measure of the
abilities and brain regions assessed. In the first such study in autism, deficits were doc-
umented in high functioning autistic individuals on the volitional saccade tasks but no
deficits were found on the visually guided saccade task (Minshew, Luna, & Sweeney,
1999). This study provided rigorous evidence that the difficulty in life with shifting
attention in autism relates to deficits in higher order control over attention by higher
brain regions and not to basic deficits in attention subserved by the cerebellum. These
data are also important for having demonstrated that there are selective abnormalities
in the highest order circuitry of the brain and that lower levels of circuitry are intact.
This pattern is analogous to that seen in the study of the neuropsychologic profile.

In summary, these studies provide evidence that autism is a selective disorder of

complex information processing abilities with intact simple information processing
abilities. The common denominator of deficits in autism is the high demands placed
on information processing or computation by the brain. The complex information
processing construct explains why these particular symptoms travel together as a syn-
drome, the failure of IQ scores and intact abilities to predict the status of higher order
abilities, the common co-occurrence of mental retardation in autism, and the differ-
ence between autism and general mental retardation. The validity of this characteri-
zation of cognitive functioning in autism is supported by its reciprocal relationship
with the neuropsychologic profile described for the prototypic simple information
processing disorder. The presence of this same dissociation between deficient com-
plex and intact simple abilities in the motor domain further attests to the validity of
this construct. The relationship of deficits in autism to their computational demands
on the brain is helpful clinically in understanding and analyzing behavioral and aca-
demic difficulties.
The saccadic eye movement studies investigating the neural basis of cognition
and behavior in autism provide evidence that these difficulties are the result of the
selective failure of the most advanced levels of brain circuitry to develop. The find-
ings of these studies, like the findings using other methods, indicate that difficulties
with shifting focus of attention relate to executive dysfunction and the frontal lobes,
and not to a deficit in the basic attentional capacity to shift attention.
The complex information processing construct does not refer to a particular abil-
ity but is a construct that identifies a common characteristic shared by all the deficits
in autism. In addition to the contributions it makes to the clinical understanding of
this disorder, this construct also provides a connection to brain function and struc-
ture. These links across levels are key to defining the sequence between genes and
behavior that will empower all intervention.


Burack, J. A., Enns, J. T., Stauder, J. E. A., Mottron, L., Randolph, B. (1997). Attention and
autism: behavioral and electrophysiological evidence. In: Cohen D. J., Volkmar, F. R.,
eds. Handbook of autism and pervasive developmental disorders. New York: John
Wiley & Sons, 226–247.
Boucher, J. (1981). Memory for recent events in autistic children. Journal of Autism and
Developmental Disorders, 11, 293–302.
Cook, E. H., Jr. (1998). Genetics of autism. Mental Retardation and Developmental Disabilities
Research Reviews, 4, 113–120.
Minshew, N. J. (1997). Autism & the pervasive developmental disorders: the clinical syn-
drome. In Behavior Belongs in the Brain: Neurobehavioral Syndromes, B. K. Shapiro,
P. J. Accardo, & A. J. Capute, eds. pp. 49–68, York Press, Baltimore, Maryland.
Minshew, N. J., Payton, J. B. (1988). New perspectives in autism: Part 1: The clinical spectrum
of infantile autism. Current Problems in Pediatrics, 18:561–610.
Minshew, N. J., Luna, B., & Sweeney, J. A. (1999). Oculomotor evidence for neocortical sys-
tems but not cerebellar dysfunction in autism. Neurology, 52:917–922.

Ozonoff, S., Strayer, D. L., McMahon, W. M., et al. (1994). Executive function abilities in
autism and Tourette syndrome: An information processing approach. Journal of Child
Psychology and Psychiatry, 32:1081–1105.
Pascualvaca, D. M., Fantie, B. D., Papgeogiou, M., Mirsky, A. F. (1998). Attentional capacities
in children with autism: is there a general deficit in shifting focus? Journal of Autism and
Developmental Disorders, 28:467–478.
Rapin, I. (1996). Preschool children with inadequate communication: Developmental language disorder,
autism, low IQ. London: Mac Keith Press.
Rutter, M., Bailey, A., Bolton, P. & Le Couteur, A. (1994). Autism and known medical condi-
tions: myth and substance. Journal of Child Psychology and Psychiatry, 35, 311–322.

Mailing Address
Nancy J. Minshew, M.D.
Associate Professor of Psychiatry and Neurology
University of Pittsburgh School of Medicine
Department of Psychiatry
Western Psychiatric Institute and Clinic
3811 O’Hara Street
Suite 430
Bellefield Towers Bldg.
Pittsburgh, Pennsylvania 15213
Neural Mechanisms In Autism

Andrew W. Zimmerman, M.D. and Barry Gordon, M.D., Ph.D.

Abstract. Autism is a heterogeneous disorder and can be associated with many coexisting
genetic and acquired diagnoses. The underlying defects that lead to it continue to elude us.
Various autistic syndromes share disruptions in several neural pathways that develop and func-
tion at different levels of organization within the central nervous system. Brain development
proceeds through cascades of different processes that tend to rely on and interact with previously
established ones. While certain degrees of individuality in maturation may occur normally, there
are critical thresholds for disturbing orderly, synchronized development. As a result of evolution,
the increased complexity of the human brain may have increased its vulnerability to fetomater-
nal immune reactions, potentially resulting in incompatibility that disturbs recently developed
functions, such as language and social skills.
Artificial neural networks, as in computational modeling, can be a useful theoretical
approach to explain how under- or over-innervated networks in persons with autism may affect
their abilities to discriminate and generalize, as one sees in autistic regression.

Neural Mechanisms in Autism1

Currently there is no reliable evidence as to exactly what are the neural bases for
autism. There are no accepted genetic markers, even though there are several can-
didates (Ingram JL, et al., 2000, Bailey, Palferman, Harvey, & Le Couteur, 1998;
Folstein, Bisson, Santangelo, & Piven, 1998; Szatmari, Zwaigenbaum, & MacLean,
1998). It is also increasingly clear that autism is a heterogeneous disorder, even if it
is genetic. There are no objective tests in vivo that are specific for the condition. There
have been no structural, metabolic, or neuropathologic abnormalities that have been
reliably linked to autistic features. There is no accepted animal model of the condi-
tion, although infant monkeys with selective brain lesions (Bachevalier, 1991;
Bachevalier & Merjanian, 1994) show behavioral features suggestive of autism.

Acknowledgments: Preparation by Andrew W. Zimmerman was supported in part by the National Alliance
for Autism Research and the East Tennessee Chapter, Autism Society of America. Barry Gordon was sup-
ported in part by the New York Community Trust—Hodgson Fund, the Hodgson Family, the Benjamin A.
Miller Family Fund, and by gifts made in memory of Bernard Gordon.
The authors thank Martha Bridge Denckla, M.D., for her helpful discussions regarding neural path-
ways in autism, and James Marcum, who contributed to the artwork (Figure 1). Susan L. Connors, M.D.,
gave helpful advice.
Condensed with permission from a chapter in Pasguale Accardo (Ed.), Autism: Clinical and research
issues. Towson, MD: York Press. (2000).


F IGURE 1. Putative neural networks in autism. Solid arrows indicate direction of

neual transmission, based on current information; large striped arrow depicts uncer-
tain cerebellar effects. Large scissors depict cellular “lesions” in important structures
that might contribute to network dysfunction or a “disconnection syndrome.” Small
scissors show potential sites for disconnection within networks and an example of
“functional’ (correct: globus pallidus to thalamus) or “dysfunctional” (aberrant) repair
that might occur from bypassing the globus pallidus. In schizophrenia, a disconnec-
tion is thought to occur between the dorsolateral prefrontal area and the anterior cin-
gulate cortex (Benes, F. M. 1993. Relationship of cingulate cortex to schizophrenia.
In The Neurobiology of Cingulate Cortex and Limbic Thalamus, eds. B. A. Vogt and M.
Gabriel. Boston: Birkhauser.).

Neurobiologic Studies

Although no neuropathologic features have been found yet to be characteristic of

autism, a number of abnormalities have been reported. Bauman and Kemper (1994)
found consistent neuronal changes (“too many, too small”) in the hippocampus,
amygdala, and other areas of the limbic system, as well as decreased Purkinje cells in
the lateral cerebellum. More recently, Bailey, et. al., (1998) reported cerebellar, neo-
cortical, and olivary (but not limbic) changes. These findings may reflect “develop-
mental curtailment” of the cellular connections in the developing cortex (neuropil)
that affects information processing and representational memory (in the hippocam-
pal complex); recognition of facial gestures and cross-modal memory (amygdala);

and shifting attention, language processing, and motor function (cerebellum). In spite
of the limitations of traditional methods inherent in light microscopic studies of
autism so far (Rapin & Katzman, 1998), and the paucity of postmortem tissue avail-
able for study, these studies have fostered a new era in neurobiological research in
autism by other investigators who are using recently developed genetic, neurochem-
ical, and morphological techniques.

Diverse Causes for Similar Defects of

“Higher Cortical Functions”

Autism can be caused by a number of different insults and etiologies. These causes
may be as diverse as viral infections, dysmorphic syndromes, or genetic abnormalities
of intracellular metabolism. In any individual, these would produce a fairly unique
pattern at the neural level, even though their behavioral outcomes are more similar.
There are some aspects of function in which it is possible to make a fairly direct
correlation to neuroanatomy, particularly in the fully developed organism. The
elementary sensory and motor systems are the best examples. However, higher cog-
nitive abilities, by their very nature, are the product of a number of different under-
lying mental functions. Each of these functions may have very complex relationships
to neural structures. These mental functions may not even be products of structures
per se, as much as of their internal dynamics or the dynamics of other systems and
structures. Moreover, many of the functions considered to be higher abilities are actu-
ally chains of abilities that unfold over time. The higher functions considered to be
most important were unlikely to have sprung up full-blown in phylogeny. More likely,
they have been cobbled together out of refined and rearranged combinations of other
functions. Therefore, such functions may not have very direct brain correlations. The
situation is even more complicated in the case of developmental disorders. Normal
mental development proceeds through a cascade of many different processes, which
tend to bootstrap each other. The interruption, or just simple delay, in any part of this
sequence, can and often does have major effects on the final components and their
assembly into a functional whole.
The genetic deficits of autism may be expressed in peculiar patterns that can be
related to neurobiologic organization, but not to the functional organization of the
nervous system. They also may be expressed at different times in the developing
nervous system. Some of these effects may be visible at the time they occur, whereas
some may take a long period of subsequent development to be expressed. It is still a
reasonable strategy to look as early in development as possible for clues to the neu-
robiologic problems. A recent example is the study by Teitelbaum, Teitelbaum, Nye,
Fryman, & Maurer (1998), which showed that movement disorders could be retro-
spectively detected in autistic children as early as ages 4 to 6 months. Of particular
relevance to the issues raised here is that the movement disorders were expressed in
different movements, and in different ways, among the different children.

Neural Networks

Neuropathological findings in the limbic system, cerebellum, and frontal cortex

in autism suggest that disorders in these structures may be important contributors to
the autism deficit. Variations in clinical expression among autism spectrum disorders
may relate to different types of effects-as well as their distribution-in related struc-
tures, such as the basal ganglia (important for motor planning) and prefrontal cortex
(motivation, executive functions) (see Figure 1). These regions may be dysfunctional
by themselves (e.g., following closed-head injury, stroke, or encephalitis) or may be
come disconnected from their interactive partners within networks due to their fail-
ure to develop, modify, or prune their connections during the development of the
neuropil (Zilbovicius et. al., 1995). For example, the basal ganglia (caudate and
globus pallidus) and thalamus are essential subcortical integrating way-stations in net-
works with prefrontal and anterior cingulate cortex. Abnormalities in subcortical
neurotransmission to or from the prefrontal cortex are likely to contribute to execu-
tive dysfunction, disinhibition and irritability, and apathy and inertia (Denckla &
Reiss, 1997).
The capacity for repair of, or compensation for, brain lesions (plasticity) is max-
imal during the early years of development ( Jacobson, 1991). Therapeutic programs
in autism may take advantage of this potential for repair (Greenspan & Wieder, 1997;
Lovaas, 1987). Although its biological basis is poorly understood, clinically effective
repair may depend on the regulation of multiple neurotransmitters, growth, and
other trophic factors in the brain while training programs are taking place. (Repair
also may occur to some degree with or without training.) Compensation for defective
way-station processes (e.g., hippocampal or cerebellar), or disconnection within net-
works, probably depends on correct forms of rerouting (“functional plasticity,” see
Fig. 1). Plasticity is functional if it compensates for a disconnection between way-
stations (e.g., between globus pallidus and thalamus). “Dysfunctional plasticity”
(e.g., from caudate to thalamus) may reduce efficiency of the repair or even negate
its effects if the new route bypasses critical parts of the network (e.g., globus pallidus).

Evolution, the Immune System and Fetal-Maternal Interaction

Several concepts from human evolution may offer alternative insights into the
origins of autism. Human language functions probably emerged in the past
50–100,000 years, although their neural substrates were evolving for at least several
million years before that (Pinker, S. J., 1995; Smith & Szathmáry, 1999).
Chimpanzees, our closest primate relatives from whom our primate ancestors
diverged 4 to 10 million years ago, are able to perceive and respond to human lan-
guages up to the level of human 3 year-olds. However, they are unable to use gram-
mar or complex constructions that children begin to acquire rapidly by this age.
Despite their capacity for communication, chimpanzees do not approach humans in
their ability to understand the mental states of others through language (Tomasello,

1999). Although non-human primates’ social skills and organization are advanced
compared to other mammals, humans are (arguably) advanced by comparison. The
evolution of brain networks (and their interactions) that made possible the expansion
of language (with syntax) and social skills, as well as mental capacity, in humans most
likely have been cobbled together over millions of years (Calvin and Bickerton,
2000). However, the pace of their evolution seems to have increased rapidly with the
appearance of art, communication and tool use in the past 50,000 years. These net-
works were certainly well developed by the advent of agriculture (10,000 years) and
recorded history (5,000 years ago; Diamond, 1992).
Two important driving forces behind evolution of the brain and its fourfold
increase in cortical surface area, in humans compared to chimpanzees, are likely to
include both immune adaptation and mate selection. Bacteria, parasites and viruses
are still the most abundant (and rapidly evolving) organisms on Earth. Despite our
multicellular complexity, humans are slowly evolving and static in comparison to
these infectious organisms (Allman, 1999). Our ancestors survived in this immuno-
logically challenging environment because they developed complex immune sys-
tems that were able to react effectively, by both innate and adaptive immune
mechanisms, to a wide array of organisms. This immune diversity continues today
to be directed, in large part, by the major histocompatibility complex (MHC; rep-
resented by human lymphocyte antigens or HLA, individual tissue types that are
represented on most cell surfaces), the most polymorphic of our genes, located on
the short arm of chromosome 6 and expressed in most of our cell types. Along with
their obvious benefits for diversifying our immune reactivity, a mismatch of HLA
types also leads us to readily reject “foreign” tissue transplants. One’s HLA types
must be carefully matched in order for an organ transplant to survive (an identical
twin or a sibling match is usually most successful). An enduring mystery, however,
is that during pregnancy, the developing fetus itself is able to survive as a “foreign
tissue transplant” to the mother (due to the fetus’s different HLA types that are
inherited from the father). Two possible explanations for this paradoxical immune
tolerance include decreased expression of HLA types by the placenta, and the
appearance of HLA-G, a special type of HLA produced by the fetus and placenta,
that serves to down-regulate maternal immune responses to the “fetal transplant”
(Hunt, et al., 2000). In addition to fetal reactions, maternal responses may include
antibodies against the fetus (Warren, et al.,1990), as well as interactions of maternal
genes with dietary and environmental factors that act as teratogens to the fetal brain
( Johnson, 1999).
HLA types are also important for mate selection, a form of natural selection that
drives evolution. They may function to increase dissimilarity between mates through
mutual attraction to olfactory cues for differing HLA types, and thereby increase
immune diversity within couples and survivability among their children. In humans,
mate choice appears to be mediated by HLA (Ober, et al., 1997), although
pheromones have not been demonstrated to mediate the process. Darwin (1871)
found that mate selection is an important epigenetic mechanism for natural selec-
tion that rapidly amplifies traits over time, such as the peacock’s tail and the finch’s

beak. Miller (2000) has suggested that the human drive to select mates may under-
lie many of our gender-related behaviors as well as our enthusiasm for the arts, cre-
ativity and cultural development. Mate selection and HLA diversity, therefore, may
be driving (and perhaps, accelerating) our recent (and continuing) evolution of brain
networks that support our most “human” attributes (language and social skills). By
extension, variants of HLA (Warren, et al., 1996) and other fetal-maternal epigenetic
mechanisms, as evolutionary adaptations, also may give rise to autism in human
beings. Autism, in its various forms, can therefore be seen as expressions of human
Several prenatal complications of pregnancy appear more commonly, in retro-
spect, when the child later develops autism. These findings have not been well
defined or explained, and may also occur in other neurodevelopmental disorders
(Nelson, 1991). The increasing incidence of autism with late birth order in some fam-
ilies (Tsai & Stewart, 1983) may reflect maternal-fetal immune incompatibility and
production of maternal antibodies (to paternal antigens and against the fetus). In
addition, an increased incidence of autoimmune disorders in the mothers and close
relatives of autistic children suggests that abnormal immune reactivity may predis-
pose the fetus to maternal “toxic” immune effects (Comi, et al., 1999). HLA associa-
tions occur commonly in autoimmune disorders, and if confirmed in children with
autism (or their mothers), may signify that both types of disorders may share genetic
mechanisms related to HLA or neighboring genes. Maternal reactivity may result
from paternal (as well as maternal) genetic components through the fetus, which has
been described recently in maternal preeclampsia (Esplin, et al., 2001).
Maternal-fetal immune interactions might produce effects in the developing
human brain that are novel to humans, and specific to unique networks that are
“recently evolving.” HLA antigens may specify network formation during critical
periods of brain development. Based on the recent work of Shatz (Huh, et al.,
2000), the MHC antigens in mice are important for cellular contact recognition
(and activity-dependent synaptic plasticity) in the developing visual system, as well
as other areas: they are especially rich in the limbic system and may be tied to the
activity of glutamate and other neurotransmitters in the developing brain. If it can
be shown that human brain networks that have recently evolved depend upon the
expression of specific HLA types, then the expression of these antigens on cell sur-
faces might be inappropriately down-regulated due to maternal-fetal intolerance
during critical periods of their development in utero. Such reactions would initiate
a cascade of downstream effects that would continue postnatally, and selectively
disrupt “recently evolved” networks and functions that support language and social

Fundamental Cognitive And Behavioral Deficits In Autism

There have been a number of attempts to tease apart the cognitive and behav-
ioral deficits that occur in autism, and to hopefully identify some as more funda-

mental than others (for reviews, see Bailey, Phillips & Rutter, 1996; Happe & Frith,
1996; Rapin 1997; Rumsey, 1996; and Litrownik & McInnis, 1982). A recent exam-
ple has been attention to the lack of theory of mind in autism. Theory of mind was the
term used by Premak & Woodruff (1978) to describe an individual’s understanding of
the motives, knowledge and beliefs of others. Frith, Baron-Cohen and others (Baron-
Cohen, 1995; Baron-Cohen & Swettenham, 1997; Frith & Happe, 1994) have noted
that autistic individuals do not seem to have such understanding, nor are they able
to develop it. Consequently, these authors have posited that a deficit in the primitive
functions that form the basis for having theory of mind could be a major cause of the
difficulties in autism. However, just what constitutes a theory of mind, who has it, and
whether it is truly impaired in autism has been debated (see, for example, Povinelli
& Preuss, 1995).
A different, perhaps more fundamental deficit should be entertained as being
present in many persons with autism, particularly if they are low-functioning: a
deficit in the ability to selectively manipulate sensory representations, concepts, and
thoughts themselves (although these may also be deficient). In basic terms, this is a
problem with the ability to imagine. However, it is not a deficit in simple visual
imagery; there is self- reported evidence that high-functioning persons with autism
not only have visual imaginations but rely upon them (Grandin, 1997). Instead, what
is referred to here is the ability to select elements of mental states and manipulate
them. Normally, humans are able to focus on different aspects of an object or expe-
rience, and even seem to break these aspects away from the original experience and
manipulate them separately. A person can see a red cup and separate out its redness
from its shape. Persons with autism, however, are notorious for not being able to do
this. They are notorious for context-dependence and for apparently focusing on the
“wrong” features of everyday objects.
There is some evidence that, in normal individuals, this ability to select features
from otherwise unitary representations is dependent upon the prefrontal cortex
(Thompson-Schill, Esposito, Aguirre, & Farrah, 1997). A deficit in such functions
would certainly fit with many of the other noted behavioral characteristics or persons
with autism: their rigidity, repetitive behavior, and perseveration; their lack of sym-
bolic play; and more elaborately, a theory of mind. (Deacon (1997) and others have
suggested that this type of mental manipulation, which is essentially symbolic, is one
of the important mental prerequisites of humanness.
Within the subgroups of autism, much still needs to be explained. Each of the
sub-groups (described in Zimmerman & Gordon, 2000) is known to be associated
with at least two paradoxes. One is that autistic individuals often have dispro-
portionate mental abilities and skills, in addition to their obvious disabilities. The
other paradox is that, despite the clear cognitive and behavioral abnormalities in
each of the autistic categories, the underlying neural pathology still seems to resist
a consistent description. Going back and reclassifying the pathology of autism into
clinical subcategories does not yet result in a more coherent picture. According to
published studies, even within these clinical subcategories, reported abnormalities
may be present in some individuals and absent (or different) in others. The inability

to find beneficial effects of any categorization scheme may simply reflect how con-
flated these categories have been in reported studies, and in the impossibility of
reconstructing them from the published accounts. It is also very possible that some
heterogencity-in mental functions as well as in neuropathology-will prove to be a fun-
damental characteristic of each category of autism.

Behavioral Heterogeneity

One of the most striking features of autism is that it is often accompanied by

relative strengths in some areas of cognition, in addition to disabilities in others
(Happe & Frith, 1996; O’Connor & Hermelin, 1989). Such patterns are well known
in developmental disorders. In Williams syndrome, speech, surface language abil-
ities, and (at times) musical ability, are typically far superior to visual-spatial abili-
ties and to general cognitive abilities (Capirci, Sabbadini & Volterra, 1996;
Tager-Flusberg, Boshart, & Baron-Cohen, 1998). Many of the developmental syn-
dromes of mental retardation have relative preservation of visual-perceptual abil-
ity (Pulsifer, 1996). However, supranormal islands of ability are much rarer in other
conditions compared to autism spectrum disorders (Happe & Frith, 1996). It has
been claimed that 10% of the autistic population has “special abilities” (Rimland &
Fein, 1988). The supranormal skills that have been described in both autistics and
in individuals with other diagnoses include lightning calculation, calendar skills, list
learning (Mottron, Belleville, Stip, & Morasse, 1998), visual memory, hyperlexia,
puzzle construction, drawing ability, musical memory, and playing by ear and
improvisation. (For more complete lists, see Happe & Frith, 1996; O’Connor &
Hermelin, 1989.) Regardless of the exact proportion having such abilities, the
overabundance of such skills demands some explanation and might even shed
some light on the nature and neurobiology of autism itself, as Frith (1989),
O’Connor (1989), and others have suggested.
Not all of the apparently superior skills that have been reported are difficult
to explain. Restricted attentional focus, repetitiveness, and the lack of compet-
ing thoughts or abilities (Frith, 1989), can certainly account for many apparent
abilities. A recent study of recent atypical memory abilities in one individual
(Mottron et. al., 1998) is perhaps an example of how superior performance in one
area may be accounted for, in some instances, by actual cognitive deficiencies in
other areas.
However, there are other examples of apparently superior ability that seem to
arise spontaneously (e.g., Selfe, 1977) and do not seem to be easily explained by the
absence of normal mental impediments. These often seem to involve implicit learn-
ing of rules and patterns (Hermelin & O’Connor, 1986). They also often seem to be
remarkable circumscribed. An individual who can do lightning calculations of dates
may not even be able to multiply numbers (Happe & Frith, 1996). It may not be
unreasonable to ask that any unified account of the neural basis for autism account
for these abilities as well as autism’s documented disabilities.

Neural Network Theories

It may be possible to unify both the behavioral herterogeneity-the abnormalities

and the supernormalities-as well as the possible neural heterogeneity. To do so
requires a digression into neural networks theories. (It should be noted that Cohen,
1994 raised many of the same hypotheses proposed here.) Neural networks theories
of cognitive processes posit that many mental operations are carried out through
successive sets (layers) of neuronal processing elements. (For a brief overview, see
Gordon, 1997.) With the proper input and training criteria, and the proper learning
of rules, such networks have proven to be extremely adept at embodying rules and
patterns that are implicit in the data presented to them. However, the accuracy of
this extraction is very dependent on the number of processing elements in the active
learning layer (Baum & Hausler, 1989). If there are too few elements then the net-
work does not learn with very good accuracy: it, in fact, tends to over-generalize. If
there are too many elements, then the network learns each specific situation pre-
sented to it and doesn’t generalize enough. If some number of working elements
leads to adequate performance, a somewhat greater number can result in truly supe-
rior performance in learning implicit rules and patterns, as long as it avoids becom-
ing too specific.
This observation might be tied in to normal development, and to the abnormal
development(s) that occur in autism, in the following way: the normal development
of higher cerebral functions in a child’s cortex appears to be driven by at least two
major influences. One is predetermined connections; the other is activity and use.
It has often been noted that the number of genes coding for the brain and neural
tissue (~50,000) are insufficient to specify all the connections of the mature brain.
Thus, the development of these connections must be guided in part by experience.
Edelman (1987) and Intrator and Edelman (1997) have suggested that whether an
uncommitted area develops connections with one region or another is based on the
outcome of a competition for use. The developing child’s brain normally has sev-
eral primary sensory inputs, including vision, audition, and touch. These inputs are
hardwired and fairly compelling. Such sensory inputs will do all they can to recruit
whatever upstream neuronal processing resources are not yet committed.
Normally, the multiple influences on a child lead to a balance of forces, with the
normal balance of lower and higher processing abilities (and neuroanatomic maps)
as a result. The amount of neural tissue that is devoted to each higher function
therefore represents a tradeoff between several forces: an attempt to optimize pro-
cessing, the practical limits on optimization (because of lack of enough experience
and training time), and competition with other functions for those same neuronal
processing elements.
What if a developing brain had all those same forces at work, but for some
reason some processing systems were impaired or delayed in their development?
What if the systems in question were those involved in speech perception and
speech production? Specific genetic deficits in speech production have been tenta-
tively identified. It is conceivable that there are other deficits or combinations of

deficits with more widespread effects on both speech production and speech per-
ception. If the systems related to speech perception and speech production were
developmentally impaired, then many higher abilities dependent upon appropriate
auditory input and output would never develop properly. Whatever cerebral tissue
would have been devoted to those higher functions would then be free to be incor-
porated into other processes (assuming the tissue itself was not too badly affected
by the same defects). If vision were intact, then visual-related abilities would be
expected to appropriate extra cerebral tissue. The result would be a child’s brain
that was not capable of all of the normal functions of a child, but that was capable
of performing some functions superlatively well. The brain would not be capable
of those abilities that are related to speech and language capability, such as a long-
term component of working memory (the part normally dependent upon an artic-
ulatory loop), and perhaps even such higher functions as the “inner voice” aspects
of consciousness. It would, however, be extraordinarily good at wordless visual
perception and analysis. Neuropathologically, such a brain might have only a few,
apparently nonspecific, abnormalities. It would not have to have fewer neurons
than normal. Autistic brains are, if anything, average or larger-than-average in size
(Courchesne, Muller & Saitoh, 1999; Lainhart et al., 1997). It might be possible to
detect additional territory devoted to visual-related functions, but perhaps not with
current behavioral tasks and instrumentation. Autism may therefore represent dis-
orders of activity-dependent plasticity during brain development that occur at sev-
eral different levels: gene, synapse, neuron, network, and neuronal group.

Hypothesis of Activity-Dependent Plasticity

In broader outline, the hypothesis is this: Either because of genetics or external

influences, several regions or neuronal networks of the developing brain are dam-
aged or delayed in their development. Regions involved in social connection and
those involved in speech and language seem to be particularly susceptible. (It is not
too speculative to imagine that they have a functional linkage and perhaps, therefore,
a genetic one as well.) There are two consequences of this primary pathology.
Functions that require these inputs cannot develop fully. Functions that were not
dependent upon these impaired routes can develop normally, and might well
develop supranormally. They would develop supranormally if these functions were
normally kept constrained by a competition for neural resources from the functions
that were now impaired (with the competition being either in functional space or per-
haps just through simple anatomic proximity).
This hypothesis has several testable consequences. There will be forme frustes of
autistic disorder—in speech and language and in socialization—representing less
extreme forms of the autistic pathology. These types of deficits should be familial.
The domino effect on functions should be predictable after research establishes a
better understanding of what functions depend upon other functions, in both

development and in operation. Finally it should be possible to identify some in vivo

correlates of the extra neural tissue that has been adopted for processing (e.g., vision)
in these individuals.
This hypothesis does not explain the primary cause or causes of the deficits. It
would, however, help to explain why persons with autism tend to have the patterns of
disabilities and abilities that they do, and why their neuropathology (in the broadest
sense) has been so variable from individual to individual. It might also suggest ways
in which functional retraining can try to ameliorate some of their disabilities or take
advantage of their particular strengths.


The next stages of investigation of neural mechanisms in autism spectrum disor-

ders should first focus on the selection of subjects and clinical definition of subsets.
Although well-studied animal models are desirable, high-functioning subjects with
autism are more likely to reveal the essential abnormalities in this very “human” dis-
order. Multiple investigative techniques, from cellular to neurochemical to cognitive
neurophysiology, and quantitative and functional neuroimaging, will help to define
the neural networks that contribute to autism.


Allman, J.M. (1999). Evolving Brains. New York: Scientific American Library.
Bachevalier, J. (1991). An Animal Model for Childhood Autism. In C.A. Tamminga and S. C.
Schulz (Eds.), Advances in Neuropsychiatry and Psychopharmacology. Pp.129–140. New York:
Raven Press.
Bachevalier, F., & Merjanian, P. (1994). The Contribution of Medial Temporal Lobe Structures in
Infantile Autism: A Neurobehavioral Study in Primates. In M. L. Bauman and T. L. Kemper
(Eds.), The Neurobiology of Autism. Baltimore: Johns Hopkins University Press.
Bailey, A., Phillips, W., & Rutter, M. (1996). “Autism: Towards an Integration of Clinical,
Genetic, Neuropsychological, and Neurobiological Perspectives.” Journal of Child
Psychology and Psychiatry, 37. 89–126.
Bailey, A., Palferman, S., Heavey, L., & Le Couteur, A. (1998). “Autism: The Phenotype in
Relatives.” Journal of Autism and Developmental Disorders, 37. 369–392.
Bailey, A., Luthert, P., Dean, A., harding, B., Janota, I., Montgomery, M., Rutter, M., & Lantos,
P. (1998). “A Clinicopathological Study of Autism.” Brain, 121. 889–905.
Bailey, C. H., Bartsch, D., & Kandel, E. R. (1996). “Toward a Molecular Definition of Long-
term Memory Storage.” Proceedings of the National Academy of Sciences of the United States
of America, 93. 13445–13452.
Baron-Cohen, S. (1995). Mindblindness: An Essay on Autism and Theory of Mind. Cambridge,
MA: MIT Press.
Baron-Cohen, S., & Swettenham, J. (1997). Theory of Mind in Autism: Its Relationship to
Executive Function and Central Coherence. In D. J. Cohen, & F. R. Volkmar (Eds.),
Handbook of Autism and Pervasive Developmental Disorders. New York, NY: John Wiley
& Sons.

Baum, E. B., & Hausler, D. (1989). “What Sized Net Gives Valid Generalization?” Neural
Computation, I. 151–160.
Bauman, M. L., & Kemper, T. L. (1994). Neuroanatomic Observations of the Brain in Autism. In
M. L. Bauman, & T. L. Kemper (Eds.), The Neurobiology of Autism. Baltimore, MD: Johns
Hopkins University Press.
Benes, F. M. (1993). Relationship of Cingulate Cortex to Schizophrenia. In B. A. Vogt & M. Gabriel
(Eds.), The Neurobiology of Cingulate Cortex and Limbic Thalamus. 581–605. Boston:
Calvin, W.H., & Bickerton, D. (2000). Lingua Ex Machina. Reconciling Darwin and Chomsky with
the Human Brain. Cambridge, MA: MIT Press.
Capirci, O., Sabbadini, L., & Volterra, V. (1996). “Language Development in Williams
Syndrome: A Case Study.” Cognitive Neuropsycology 13. 1017–1040.
Cohen, I. L. (1994). “An Artificial Neural Network Analogue of Learning in Autism.” Biological
Psychiatry, 36. 5–20.
Comi, A.M., Zimmerman, A.W., Frye, V.H., Law, P.A., & Peeden, J.N. (1999). “Familial
Clustering of Autoimmune Disorders and Evaluation of Medical Risk Factors in
Autism.” Journal of Child Neurology, 14. 388–394.
Courchesne, E., Muller, R. A., & Saitoh, O. (1999). “Brain Weight in Autism: Normal in the
Majority of Cases, Megalencephalic in Rare Cases.” Neurology 52. 1057–1059.
Darwin, C. (1871). The Descent of Man, and Selection in Relation to Sex. (2 vols.). London: John
Murray (Reprinted in 1981 by Princeton University Press).
Deacon, T. (1997). The Symbolic Species: The Co-evolution of Language and the Brain. New York:
W. W. Norton.
Denckla M. B., & Reiss, A. L. (1997). Prefrontal-Subcortical Circuits in Developmental
Disorders. In N. A. Krasnegor, G. R. Lyon, & P. S. Goldman-Rakic (Eds.), Development
of the Prefrontal Cortex: Evolution, Neurobiology, and Behavior. Baltimore, MD:
Paul H. Brookes.
Diamond, J. (1992). The Third Chimpanzee. The Evolution and Future of the Human Animal. New
York: HarperCollins.
Edelman, G. M. (1987). Neural Darwinism: The Theory of Neuronal Group Selection. New York:
Basic Books.
Esplin, M.S., Fausett, M. B., Fraser, A., Kerber, R., Mineau, G., Carrillo, J., & Varner, M.W.
(2001). “Paternal and Maternal Components of the Predisposition to Preeclampsia.” New
England Journal of Medicine, 344. 867–872.
Folstein, S. E., Bisson, E., Santangelo, S. L., & Piven, J. (1998). “Finding Specific Genes that
Cause Autism: a Combination of Approaches Will be Needed to Maximize Power.”
Journal of Autism and Developmental Disorders, 28. 439–445.
Frith, U., (1989). Autism: Explaining the Enigma. Oxford: Blackwell.
Frith, U. and Happe, F. (1994). “Autism: Beyond “Theory of Mind.” Cognition, 50. 115–132.
Gordon, B. (1997). Models of Naming. In H. Goodglasss & A. Wingfield (Eds.), Anomia. 31–64.
San Diego, CA: Academic Press.
Grandin, T. (1997). A Personal Perspective on Autism. In D. J. Cohen and F. R. Volkmar (Eds.),
Handbook of Autism and Pervasive Developmental Disorders. New York, NY: John Wiley
and Sons.
Greenspan, S. I., & Wieder, S. (1997). “Developmental Patterns and Outcomes in Infants and
Children with Disorders of Relating and Communication: A Chart Review of 200 Cases
of Children with Autistic Spectrum Diagnoses.” The Journal of Developmental and Learning
Disorders, 1. 87–141.
Happe, F., & Frith, U. (1996). “The Neuropsychology of Autism.” Brain, 119. 1377–1400.
Hermelin, B., & O’Connor, N. (1986). “Idiot Savant Calendrical Calculators: Rules and
Regularities.” Psychological Medicine, 16, 885–893.

Huh, G.S., Boulanger, L.M., Du, H., Riquelme, P.A., Brotz, T.M., & Shatz, C.J. (2000).
“Functional Requirement for Class I MHC in CNS Development and Plasticity.” Science,
290. 2155–2159.
Hunt, J.S., Petroff, M.G., Morales, P., Sedlmayr, P., Geraghty, D.E., & Ober, C. (2000).
“HLA-G in Reproduction: Studies on the Maternal-Fetal Interface.” Human Immunology,
61. 1113–1117.
Ingram, J.L., Stodgell, C.J., Hyman, S.L., Figlewica, D.A., Weitkamp, L.R., & Rodier, P.M.
(2000). Discover of Allelic Variants of HOXA1 and HOXB1: Genetic Susceptibility to
Autism Spectrum Disorders. Teratology, 62. 393–405.
Intrator, N., & Edelman, S. (1997). “Competitive Learning in Biological and Artificial Neural
Computation.” Trends in Cognitive Sciences, 1. 268–272.
Jacobson, M. (1991). Developmental Neurobiology (3rd Ed.) New York, NY: Plenum Press.
Johnson, W.G. (1999). “DNA Polymorphism-Diet-Cofactor-Development Hypothesis and the
Gene-Teratogen Model for Schizophrenia and Other Developmental Disorders.”
American Journal of Medical Genetics (Neuropsychiatric Genetics), 88. 311–323.
Lainhart, J. E., Piven, J., Wzorek, M., Landa, R., Santagelo, S. L., Coon, H., & Folstein, S. E.
(1997). “Macrocephaly in Children and Adults with Autism.” Journal of American Academy
of Children and Adolescent Psychiatry, 36. 282–290.
Litrownik, A. J., McInnis, E. T. (1982). Cognitive and Perceptual Deficits in Autistic Children: A
Model of Information Processing, Critical Review, and Suggestions for the Future. In
J. J. Steffen and P. Karoly (Eds.), Autism and Sever Psychopathology, Vol. 2. Advances in
Child Behavioral Analysis and Therapy. 103–155. Lexington, KY: Lexington Books.
Lovaas, O. I. (1987). “Behavioral Treatment and Normal Educational and Intellectual
Functioning in Young Autistic Children.” Journal of Autism and Developmental Disorders, 9.
Miller, G.R. (2000). The Mating Mind: How Sexual Choice Shaped the Evolution of Human Nature.
New York: Doubleday.
Mottron, L., Belleville, S., Stip, E., & Morasse, K. (1998). “Atypical Memory Performance in
an Autistic Savant.” Memory, 6. 593–607.
Nelson, K.B. (1991). “Prenatal and Perinatal Factors in the Etiology of Autism.” Pediatrics
(Supplement), 87. 761–766.
Ober, C., Weitkamp, L.R., Cox, N., Dytch, H., Kostyu, D., & Elias, S. (1997). “HLA and Mate
Choice in Humans.” American Journal of Human Genetics, 61. 497–504.
O’Connor, N., & Hermelin, B. (1989). The Memory Structure of Autistic Idiot-Savant
Mnemonists. British Journal of Psychology, 80. 97–111.
Pinker, S. (1995). Facts About Human Language Relevant to its Evolution. In J. P. Changeux and J.
Chavaillon (Eds.), Origins of the Human Brain. New York: Oxford.
Povinelli, D., & Preuss, T. (1995). “Theory of Mind: Evolutionary History of a Cognitive
Specialization.” Trends in Neurosciences, 18. 418–424.
Prior, M. R. (1987). “Biological and Neuropsychological Approaches to Childhood Autism.”
British Journal of Psychiatry, 150. 8–17.
Pulfiser, M. B. (1996). “The Neuropsychology of Mental Retardation.” Journal of the International
Neuropsychological Society, 2. 159–176.
Rapin, I. (1997). “Autism.” New England Journal of Medicine, 337. 97–104.
Rapin, I., and Katzman, R. (1998). “Neurobiology of Autism.” Annals of Neurology, 43. 7–14.
Rimland, b., and Fein, D. (1998). Special Talents of Autistic Savants. In L. Obler and D. Fein
(Eds.), The Exceptional Brain: Neuropsychology of Talent and Special Abilities. 474–492. New
York: Guilford Publications.
Rumsey, J. M. (1996). Neuroimaging Studies of Autism. In G. R. Lyons & J. M. Rumsey (Eds.),
Neuroimaging: A Window to the Neurological Foundations of Learning and Behavior in Children.
119–146. Baltimore, MD: Paul H. Brooks.
Selfe, L. (1977). Nadia. New York: Harcourt Brace Jovanovich.

Smith, J. M., and Szthmáry, E. (1999). The Origins of Life From the Birth of Life to the Origin of
Language. New York: Oxford.
Szatmari, P., Jones, M. B., Zwaigenbaum, L., and MacLean, J. E. (1998). “Genetics of Autism:
Overview and New Directions.” Journal of Autism and Developmental Disorders, 28. 351–368.
Tager-Flushberg, H., Boshart, J., & Baron-Cohen, S. 1998. “Reading the Window to the Soul:
Evidence of Domain-Specific Sparing in Williams Syndrome.” Journal of Cognitive
Neuroscience, 10. 631–639.
Teitelbaum, P., Teitelbaum, O., Nye, J., Fryman, J., & Maurer, R. G. (1998). “Movement
Analysis in Infancy May be Useful for Early Diagnosis of Autism.” Proceeding of the
National Academy of Sciences of the United States of America, 95 (23). 13982–13987.
Thompson-Schill, S., D’Esposito, M., Aguirre, G. K., & Farrah, M. J. (1997). “Role of the Left
Inferior Prefrontal Cortex in Retrieval of Semantic Knowledge: A Reevaluation.”
Proceedings of the National Academy of Sciences of the United States of America, 94. 14792–14797.
Tomasello, M. (1999). The Cultural Origins of Human Cognition. Cambridge, MA: Harvard
University Press.
Tsai, L.Y., & Stewart, M.A. (1983). “Etiological Implication of Maternal Age and Birth Order
in Infantile Autism.” Journal of Autism and Developmental Disorders, 13. 57–65.
Warren, R.P., Cole, P., Odell, J.D., Pingree, C.B., Warren, W. L., White, E., Yonk, J., and Singh,
V.K. (1990). “Detection of Maternal Antibodies in Infantile Autism.” Journal of the
American Academy of Child and Adolescent Psychiatry, 29. 873–877.
Warren, R.P., Odell, J.D., Warren, W.L., Burger, R.A., Maciulis, A., Daniles, W.W., & Torres,
A.R. (1996). Strong Association of the Third Hypervariable Region of HLA-DR( With
Autism. Journal of Neuroimmunology, 67. 97–102.
Zibovicius, M., Garreau, B., Samson, Y., Remy, P., Barth, C., Syrota, A., & Lelord, G. 1995.
“Delayed Maturation of the Frontal Cortex in Childhood Autism.” American Journal of
Psychiatry 152. 248–252.
Zimmerman, A. W., and Gordon, B. (2000). In P. Accardo (Ed.), Autism: Clinical and Research
Issues. Towson, MD: York Press.

Mailing Address:
Andrew W. Zimmerman, M.D. and Barry Gordon, M.D., Ph.D.
Kennedy Krieger Institute
707 North Broadway
Baltimore, MD 21205
To Contributors

Manuscript Preparation:
Text: Double spaced; typed on heavy bond paper; 1.5 inch margins on all four sides. Note:
All materials submitted must be double spaced-block quotations, text, footnotes, references,
extracts from case material, and tables. Pages must be number sequentially.
Abstract: All papers and case presentations should begin with a brief, double spaced descrip-
tion of the material, approximately 150 words.
Art: (includes charts, tables, figures, drawings, photographs) is to be submitted to the pub-
lisher camera ready.
Dimensions of artwork: 4.25 inches wide by 7 inches high.
Cost of alterations to artwork to be borne by author.
Tables: should be double spaced, rules kept to a minimum-head, foot, saddle-no vertical
Footnotes: Numbered sequentially and double spaced: should appear at the foot of the page
where cited; footnote numbers should be typed one space above line without punctuation or
parentheses. A corresponding number should appear in the text. Note: Initial footnotes refer-
ring to the title of the paper, acknowledgements, bios, or the author do not carry a number.
Author’s Address and Affiliation should appear following the reference list at the end of the
References: Full reference to all works cited in the text should be given in the list of refer-
ences at the end of the paper. Authors listed alphabetically and their works chronologically.
When the author has published several works in the same year, the date is followed by a, b, c,
etc. All authors are listed in a mutiauthor reference. Include only works cited in the text.
References with the same first author and different coauthors are listed alphabetically accord-
ing to the second, third, etc., coauthor, i.e., Jones, Brown, O’Malley (1989) would appear
before Jones, O’Malley, Brown (1972). See the Publication Manual of the American Psychological
Association for further information.
Books: Author. (date). Title. Place: Publisher. (Original date of publication) [if different
from cited date]. When referring to the writings of Sigmund Freud, cite only the Standard
Chapter in Edited Volume: Author (date). Title of chapter. In J. Smith & B. Jones (Eds.),
Title of Book (pp. 111–135). Place: Publisher.
Journal Articles: Author. (date). Title of paper. Journal Title in Full, volume number, inclu-
sive page numbers.
Text citations: Provide author(s) name(s) followed by date. Jones (1918), or ( Jones 1918).
Multiple text citations are listed in alphabetical order (i.e., Abraham, 1966; Jones; 1950).
Use of et al.: Six or fewer must list every author in the text at first mention, thereafter et al.
is used. More than six authors is et al., at first and subsequent mentions.
Direct quotation (less than 40 words) are run in text: Jones & Smith (1918, p. 45) or ( Jones &
Smith, 1918, p. 45)
Block quotation (40 words or more): [ Jones & Smith, 1918, p. 45]
Floor Time Techniques and the DIR Model
With Stanley I. Greenspan, M.D. and Serena Wieder, Ph.D.
This 13 tape series, nearly 20 hours in length, features Stanley I. Greenspan, M.D. and
Serena Wieder, Ph.D. demonstrating the Developmental, Individual Difference, Relationship-
Based (DIR) model and Floor Time strategies for different types of children and families with
special needs. Included with the videotapes is a guide highlighting what to look for in each
of the tapes.
ICDL Members $495.00 each set plus $35.00 shipping and handling = $530.00
Non-members $550.00 each set plus $35.00 shipping and handling = $585.00

Clinical Practice Guidelines
Redefining the Standards of Care for Infants,
Children, and Families with Special Needs
For professional, parents and others concerned with improving the care of children with devel-
opmental and learning disorders. To view and/or download the Guidelines, visit
ICDL Members $28.00 each plus $7.50 shipping & handling = $35.50
Non-Members $35.00 each plus $7.50 shipping & handling = $42.50

The Functional Emotional Assessment Scale (FEAS)

For Infancy and Early Childhood: Clinical and Research Applications
Written By, Stanley I. Greenspan, MD, Georgia DeGangi, Ph.D., OTR & Serena Wieder, Ph.D.
A systematic in depth approach to assessing emotional functioning during infancy and early
childhood. It enables clinicians, educators and caregivers to assess the child’s functional,
emotional, developmental level and create a treatment plan based on the child’s individual
profile and measure their progress. The FEAS not only delineates the emotional functioning
of the infant and child but captures the richness of the interactions between the child and his
or her caregivers.
ICDL Members $32.00 plus shipping and handling
Non-Members $39.50 plus shipping and handling
Additional Protocol Booklets - $5.00 each plus shipping and handling

Please allow 6-8 weeks for delivery
Membership information and order forms can be downloaded from our website: or call: 301-656-2667 or write to:
The Interdisciplinary Council on
Developmental and Learning Disorders
4938 Hampden Lane, Suite 800
Bethesda, MD 20814