PLEASE NOTE: THIS DISSERTATION IS SEVERAL YEARS OLD. THE INFORMATION AND TECHNIQUES DESCRIBED HAVE BEEN REFINED AND IMPROVED AND NEW INFORMATION ABOUT THE VESTIBULAR SYSTEM, GUT-IMMUNE-ENDOCRINE LINKS AND OTHER ISSUES MAY BE OUTDATED. PLEASE REFER TO THE NEWEST BOOK, "Brain Training: New Hope for Children with Developmental Delays" FOR THE LATEST INFORMATION.
Michelle L. MacAlpine, "Microdevelopment in Autism: The Power of Play"
PHD Dissertation, University of Texas at Dallas, 1998.
This study investigated the small steps of development, called microdevelopment, that occurred in two groups of autistic children during the process of recovery. One group of children was treated by the author, and a second group of children was treated by parents who had been trained during a short workshop. After a mean of 8 months (range 5-15 months), the children in both groups made similar progress on three outcome measures: the Childhood Autism Rating Scale (CARS), a Parent Survey of Abnormal Behaviors, and a clinical assessment of developmental stage of play. The group treated by the author was reevaluated three months after the first follow-up, at which time 100% of the children were scoring in the non-autistic range on the CARS.
A microdevelopmental approach focuses on development from within, which is driven by the learner, rather than a teacher. In this study, microdevelopmental methods were used to facilitate the children's development from the point at which their development had apparently arrested. The children were given rich, stage appropriate objects and activities to explore in a free play situation in which an adult was available to provide meaning and to offer similar objects to those that interested the child.
Microdevelopmental approaches have been recommended as a means to facilitate problem solving and developmental learning by Granott (1991, 1992, 1995, 1998a, (1998b), who used stage appropriate activities to facilitate the learning process in adults who were exploring novel objects. Granott found that microdevelopment was characterized by oscillations and variability. She described a pattern of regressions to lower levels of exploration and understanding that preceded higher levels of understanding. The microdevelopment of children in this study was characterized by these same attributes.
Autism has been associated with a plethora of conditions, including metabolic disorders, neuropathies, genetic disorders, immunological disorders, and sensory impairments. A recent explanation of autism suggests that it is the result of impaired hippocampal functioning (DeLong, 1992). The hippocampal system in the brain is a central processing area involved in learning. It develops meaning from sensory information, creates pragmatic categories, and compares across contexts (Eichenbaum et al, 1996; McLelland et al. 1995). It is particularly active when exposed to novelty, a rich environment, and during problem solving tasks (Thompson et al, 1996; Zhu et al, 1997). The attributes of microdevelopmental methods are particularly well suited to these functions, emphasizing a rich environment, problem solving, categories, and meaning.
In this study, many of the participants had additional neurological, sensory, or immunological impairments and were severely autistic. In spite of this, within a relatively short time (5 to 11 months), the children had been able to achieve a score in the non-autistic range on the CARS, progress to higher stages of developmental play, and lose autistic behaviors. This suggests that the microdevelopmental methods used in this study may have provided the necessary catalyst so that development could resume.
It is commonly suggested that autistic children will remain autistic throughout their lives (Schopler, 1983), and that methods currently available to treat autism can produce only modest decreases in autistic behaviors (Gilberg, 1991; Gonzalez et al, 1993; Lord and Venter, 1992; Shapiro & Hertzig, 1995). Yet, there are reports in the literature of recovery from autism (Lovaas, 1987; Zappella, 1990) as well as case history reports of such recovery (Grandin, 1992; Kaufman, 1979).
The authors of these reports credit different reasons for the children's recovery. Lovaas (1987) credited years of intensive skill learning. Zappella (1990) credited the ability of family members to learn how to give meaning to their child's behaviors. Grandin (1992) credited dedicated teachers and family members. Kaufman (1979) credited an attitude of acceptance and an ability to share meaning. As diverse as these reports appear at first glance, they do have one thing in common: in each case, there was a focus on some type of learning.
Definitions
The term, "autism" is often used interchangeably to describe any of the autistic spectrum disorders. These disorders include Infantile Autism, Atypical Autism, Asperger's Disorder, Landau-Kleffner, Rett's Disorder, Childhood Disintegrative Disorder, and Pervasive Developmental Delay. Children with these disorders share characteristic autistic behaviors, which are described in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, [DSM-IV] (1994). The criteria require that a child meet six criteria with at least two from 1-4, at least from 5- 8 and at least one from 9-12.
1. Marked impairment in the use of eye gaze, facial expression, body postures and gestures to regulate social interaction
2. Failure to develop peer relationships appropriate to developmental level
3. A lack of spontaneous seeking to share enjoyment, interests, or achievements with others
4. A lack of social or emotional reciprocity
5. Delay or lack of development of spoken language (without attempts to communicate by gesture)
6. Inability to initiate and sustain conversation ( in verbal individuals)
7. Stereotyped and repetitive use of language or idiosyncratic language
8. Lack of varied, spontaneous, make-believe play or social imitative play appropriate to developmental level
9. Encompassing preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal either in intensity or focus
10. Apparently inflexible adherence to specific, non-functional routines or rituals
11. Stereotyped and repetitive motor mannerisms
12. Persistent preoccupation with parts of objects
Developmental delay in either language, social interaction, or play appears prior to 3 years of age. When at least 6 of these criteria are met, a diagnosis of autism is given. If the criteria are met but there is no significant language delay, a diagnosis of Asperger's is used. When there are slow-sleep, spike and waveform seizures present in addition to autistic behaviors, a diagnosis of Landau-Kleffner is used. If there has been a deceleration of head growth in infancy, impairment in gait, coordination, and hand wringing in addition to autistic behaviors, a diagnosis of Rett's Disorder is appropriate. If normal development was disrupted by an insult to the developing brain, after which the child loses previously acquired skills and develops autistic behaviors, a diagnosis of Childhood Disintegrative Disorder is used. If a child does not meet all 6 criteria for a diagnosis of autism but still has a significant number of autistic behaviors, a diagnosis of Pervasive Developmental Disorder is used.
Unfortunately, there is often disagreement between professionals (psychologists, pediatricians, speech pathologists, occupational therapists) in the diagnosis of these disorders (Bishop, 1989). For example, a child who demonstrates autistic behaviors may be given a diagnosis of "Autism" by an American psychologist, "Asperger's Syndrome" by a British psychologist, "Semantic-Pragmatic Disorder" by a speech pathologist, and "Pervasive Developmental Delay Not Otherwise Specified" by a pediatrician (Bishop, 1989). In this dissertation, the term "autism" will be used to represent any of the autistic spectrum disorders, but not Semantic-Pragmatic Disorder or Pervasive Developmental Delay.
Traditional approaches to treating autism
Traditional treatments for treating autism have been directed at reducing autistic behaviors by treating a related pathology of the immune, endocrine, or sensory systems. Examples of this approach include the use of immune globulin (Gupta, 1996), dialysable lymphocyte extract (Fudenberg, 1996), secretin (Horvath et al, 1998), B6 and magnesium (Pfeiffer, 1995; Rimland,1996; Rimland, Callaway & Dreyfus, 1978), diets that eliminate gluten and casein (Lucarelli et al, 1995), ascorbic acid therapy (Dolske et al, 1993), Auditory Integration Therapy [AIT] (Rimland & Edelson, 1994), Auditory Training therapy (Bettison, 1996), Sensory Integration Therapy (Ayers, 1978), and Music Therapy (Wimpory et al, 1995). These approaches appear to have benefits for a portion of the autistic population.
Drugs are also frequently used to treat autistic behaviors. Common drug therapies include buspirone (McCormick, 1997), Ritalin (Quintana et al, 1995), clomipramine (Brasic et al, 1994), clonidine (Jaselskis et al, 1992), Risperidone (Findling, Maxwell, & Wiznitzer, 1997), and corticosteroids (Stefanatos, Grover & Geller, 1995). Although some drugs had positive effects in early studies, subsequent double-blind studies or research reviews failed to confirm more than a minimal effect for Haloperidol (Chabrol, Bonnet, & Roge, 1996), fenfluramine (Sherman et al, 1989; Varley & Holm, 1990), and naltrexone (Kolmen et al, 1997; Willemsen-Swinkels, Buitelaar & Van Engeland, 1996) or immuloglobulin (Plioplys, 1998). Drug treatments also have the risk of adverse side effects, such as those associated with clomipramine (Sanchez et al, 1996), fenfluramine (Varley & Holm, 1990), and risperidone (Horrigan & Barnhill, 1997). After a review of the literature, Chabrol, Bonnet & Roge (1996) concluded that although drug therapies can show statistically significant effects for certain behaviors, in general they have not been effective in changing either the course or the symptoms of autism.
Other traditional methods for treating autistic behaviors are based on rote skill learning of skills in a contingent punishment and reward system. The most popular of these methods is an approach called "behavior modification", or "behaviorism".
Followers of the behaviorist approach contend that humans and animals learn through stimulus and response associations, in which a particular event (stimulus) has been associated with a reward or punishment. Behaviorists use a learning method called "operant conditioning", in which a person is rewarded for correct responses and punished for incorrect responses.
I.O. Lovaas incorporated operant conditioning into the treatment of autism in the 1970's. The Lovaas method consists of intensive, repetitive training or skill drills (called "discrete trials") in which correct responses are rewarded and incorrect responses are punished (Lovaas et al, 1973).
The behaviorist approach came under severe attack in the late 1980's due to the use of painful techniques like electric shock, slapping, cold water, ammonia, physical restraint, forced exercise, or tickling, as punishment for behaviors as innocuous as vocalizations, mouthing objects, toe walking, or hand play (Matson & Taras, 1989). However, the behaviorist approach is still the most common type of therapy available today for autistic children, and is actively promoted under the title "Early Intervention Project" (Gresham & MacMillan, 1998).
Lovaas originally gained great influence in this field after the publication of two papers on the results of his behavioral therapy (Lovaas, 1973, 1987). In the first paper (Lovaas, 1973), Lovaas examined the records of 20 children who had been enrolled in an intensive behavior therapy program for autistic children over a period of 12-14 months. The discrete trials used during therapy were designed to address five areas: reducing echolalia and self- stimulation, and increasing appropriate verbalizations, social nonverbal acts, and appropriate play. Lovaas was able to demonstrate a therapeutic effect for his techniques: a 10% reduction in self-stimulation behaviors, a 3-10% increase in appropriate verbal responses and social nonverbal acts, and a 20%increase in the incidence of appropriate play.
Lovaas published a follow-up review of 13 of these children several years later (Lovaas, 1978). In his follow-up study, Lovaas divided his results into two categories: children who had been discharged to parents trained to continue their behavioral therapy (Group A), and children who were discharged to either untrained parents or an institution (Group B). The children who were released to trained parents (Group A) sustained treatment gains, but the children in Group B not only lost treatment gains, in some areas they had regressed to a point worse than they were prior to therapy. Lovaas (1978) acknowledged:
"The treatment gains were often situation specific: that is, there was often limited generalization from the treatment environment to the outside. Follow-up measures taken from one to four years after treatment showed large differences depending upon the post treatment environment... Most often, the treatment proceeded slowly, and required a massive effort which one could only achieve with a large, enthusiastic, and well-trained staff."( p. 369)
Subsequent to these findings, Lovaas began to incorporate a program of intensive parent training into therapy. In 1987 he published another paper in which he analyzed the progress of 19 autistic preschoolers who were in an intensive behavioral therapy program (40 hours or more per week for at least two years), and whose parents had received extensive training. For more than 4000 hours, therapists used discrete trials and operant conditioning to promote language, play, social behavior, and reduce rituals and aggression. By first grade, these children had achieved higher IQ scores and 9 children (47%) were in regular school classrooms.
That same year (1987), Lovaas also published a case study of two autistic siblings who had recovered from autism using the Lovaas therapy methods. These papers succeeded in propelling the Lovaas method to new heights of popularity, in spite of continued criticism directed at him for using unusual and inadequate outcome measures (Schopler, Short & Mesibov, 1989; Shapiro & Hertzig, 1995), and for problems replicating the results (Gresham & MacMillan, 1998; Shapiro and Hertzig, 1995).
In 1993, McEachin, Smith, and Lovaas published a follow-up study on the children from his controversial 1987 paper. In this study, the authors concluded that the children In his original study had retained IQ scores in the normal range, and 47% of the children were still in regular school classrooms. The authors added two parent interview scales as outcome measures, the Vineland Adaptive Behavior Scales (Sparrow, Balla & Ciccetti, 1984) and the Personality Inventory (Wirt et al, 1977), in order to assess the degree of abnormal behaviors and psychological disorders in these children. These measures showed mild to moderate levels of maladaptive behaviors present in the original group of treated children. Among the nine children who achieved the best outcome in the original Lovaas study (the 47% who were classified as normal in the 1987 paper), one-third (three) children had Maladaptive Behavior Scores at clinically significant levels.
Other authors have also published long-term reports on the efficacy of behavioral modification techniques to treat autistic children. Browning (1971) published a long-term study on five children enrolled in an intense behavior modification program for three years that included parent training and family therapy. Browning described the program as an individually tailored program in which "all activities were planned and designed to extinguish deviant behavior and train age-appropriate responses" (p. 320). At the end of the study, Browning drew some rather discouraging conclusions:
"At best, with the behavior modification programming utilized here, we were able to train these children to perform as variably tractable, but severely retarded and socially dependent children... It appeared as if these children could be trained to perform specific responses, but the acquisition of this behavior did not facilitate subsequent learning of related behavior... Each response had to be acquired separately, and it appeared that in shaping behavior, acquisition of the previous response did not facilitate learning of later ones in the same response hierarchy.... It took thousands of trials to teach new behavior and to extinguish unacceptable behavior; the children did not acquire a normal response repertoire expected of a child their age... The results indicated that in general the children were trainable, but only with an expensive, intense, long-term program. At the conclusion of the program, the basic learning deficits still remained, i.e. the retardation, failure to alert and maintain attention, tendency to lapse into isolation and stereotypic behavior, and dependency on structured programs."(p. 326).
In a more recent study on the effects of intensive behavioral treatment for autism, Smith et al (1997) assessed the outcomes of retarded autistic preschoolers. The authors found that children in the experimental group obtained a higher mean IQ and had more expressive speech than the control group, but the children remained quite delayed.
Perhaps in response to these problems, the TEACCH (Treatment and Education of Autistic and related Communication Handicapped) program was born (Lansing and Schopler, 1978). Working in the school systems of North Carolina, Eric Schopler and Gary Mesibov added Montessori routines, sensory activities, music, and sign language, to the basic philosophy of operant conditioning and created the TEACCH method. Today, this method is used in special education classrooms throughout the world.
The effectiveness of the TEACCH program was reviewed in 1989 in a paper published by Mesibov, Schopler, Schaffer, and Michal. In this paper, the progress of a group of 59 autistic children enrolled in the TEACCH program over a seven-year period was reviewed. The authors used a relatively new measure of outcome called the Childhood Autism Rating Scale, or CARS (Schopler, Reichler & Renner, 1986). This scale consists of 14 areas of impairment (social relationships, imitation, emotional response, body use, object use, adaptation to change, visual response, listening response, taste/smell/touch response, fear, verbal communication, nonverbal communication, activity level, and intelligence) in addition to a general impression rating. The rater marks each area according to degree of impairment, from normal (1) to severely abnormal (4), to obtain a composite score.
Mesibov et al (1989) evaluated the children at baseline (mean age of 8 years) and again in adolescence, and found that the mean decline in CARS scores during this period was less than 3 points (from 38.47 to 35.54). The general trend was towards improvements in severity of symptoms, and 8% of children originally diagnosed as autistic were no longer autistic at follow-up. The authors concluded that, "significant positive changes do occur for autistic clients in the TEACCH program as they pass through adolescence... There is no certainty the same progress would occur without these appropriate services, and, in fact, the literature suggests it might not." (Lansing & Schopler, 1978, p. 540)
Autistic children treated in other educational settings have also demonstrated modest improvements over time. A study by Rutter and Bartak (1973) compared the behavioral, scholastic, and social progress of 50 children across three educational settings over a period of three and a half years. The educational settings primarily varied in classroom structure and academic emphasis, with some children required to attend a structured traditional program, and others attending less structured programs. Rutter and Bartak reported, "the children generally had made considerable social, behavioral and scholastic progress. Moreover, a substantial number of children had been able to leave the autistic units and receive education in a more ordinary environment". Significant improvement in deviant behavior was noted in 32% of the children.
In addition to the long-term studies reviewed above, there exist a few in which the type or extent of treatment was not specified, but improvements over time were noted. For example, Ballaban-Gil et al (1996) studied the behavioral, language, and social outcomes in a population of 99 autistic adolescents and adults who had been initially examined in childhood. The authors used a telephone interview with parents as their outcome measure. Parents answered questions about the child's intelligence, behavioral difficulties, behavioral change, socialization abnormalities, living arrangements, expressive language, and language comprehension. At follow-up, 31% had few behavior problems, and 4% had socially appropriate behavior.
Another example is a study by Gonzalez et al (1993), who reviewed the progress over five years of 30 young autistic children who had been hospitalized for autism or Pervasive Developmental Delay in early childhood. Ten patients were using haloperidol during the time of the study. Assessment measures included clinical diagnoses, a parent interview, a patient interview, and a review of school reports and past psychological evaluations. At follow up, the authors reported a statistically significant decrease in severity of illness. Normal play appeared in 26% of the children, 13% had peer friendships, and 26% had no stereotyped body movements (self-stimulation behaviors).
Unfortunately, operant conditioning, the most common learning method used to treat autism, requires a staggering investment of time, effort, and financial resources to produce positive results. Lovaas (1987) described his methods as requiring early intervention carried out during all or most of the child's waking hours, addressing all significant behaviors in all of the child's environments, by all significant persons, for many years. Lovaas estimated the cost of such a program in 1987 as $20,000 per year. In addition, behavioral methods continue to be controversial. For example, Gresham and MacMillan (1998) recently reviewed the claims of the Early Intervention Project (which relies on operant conditioning methods of learning) and recommended an "attitude of healthy skepticism before proceeding to an unqualified endorsement of the EIP as a treatment for autism".
Impairments associated with autism
A wide variety of metabolic, immunological, genetic, and sensory impairments have been associated with autism. Metabolic errors related to autism include errors of carbohydrate metabolism (Coleman & Blass, 1985), errors of peptide metabolism (Reichelt, et al, 1986; Shattock & Lowdon, 1991), purine and pyrimidine disorders (Simmonds et al, 1997), and malabsorbtion (Goodwin, Cowen, & Goodwin, 1971).
Immunological impairments include congenital viral infections (Chess, 1977; Chugani et al, 1990; Ivarsson et al, 1990; Stubbs, 1978), impaired immune responses (Singh, 1996; Warren et al., 1996; Warren & Singh, 1996); abnormal immune response to neuronal proteins (Singh, et al, 1997; Singh et al 1993; Weizman et al, 1982), elevation of serotonin in dysregulated immune systems (Warren & Singh, 1996), and food allergies (Lucarelli, et al, 1995; Waring & Ngong, 1993).
Genetic defects or markers include micro deletions on the X chromosome (Petit, et al., 1996), Fragile X Syndrome (Mazzocco et al, 1997; Reiss & Freund, 1992), and other chromosomal disorders and genetic syndromes (Howlin, Wing, & Gould, 1995; Landgren, Gillberg, & Stromland, 1992).
Sensory impairments that have been associated with autism include visual impairments (Brown, Hobson, & Lee, 1997), impairments related to fetal alcohol exposure (Aronson, Hagberg, & Gilberg, 1997), hearing impairments (Konstantareas & Homatidis, 1987; Jure, Rapin, & Tuchman, 1991; Klin, 1993; Smith et al, 1988), and impairments in sensory coding (Barry, & James, 1988; Garreau et al, 1984; O'Neill & Jones, 1997).
Autism has been also related to various neuropathies, including increased volume of cerebellum (Piven et al, 1997), hyperplasia of vermal lobules VI and VII (Courchesne, Townsend & Saitoh, 1994), hypoplasia of vermal lobules I-V and VI-VII (Ciesielski, et al, 1997; Courchesne, Townsend, & Saitoh, 1994), parietal lobe abnormalities (Novick, Kurtzberg & Vaughn, 1979; Courchesne, Press, & Yeung-Courchesne, 1993), temporal lobe dysfunction or abnormalities (Hoon,& Reiss, 1992), reduced size of midbrain and pons (Hashimoto et al, 1991), hippocampal sclerosis (DeLong,& Heinz, 1997), Chiari I malformation (Brill, Guierrex, & Mishkin, 1997), Bourneville's Tuberous Sclerosis (Reich et al, 1997), seizures (Rossi et al, 1995; Volkmar & Nelson, 1990), traumatic brain injury (Weir. & Salisbury, 1980), and reduced dendritic branching in the hippocampus (Raymond, Bauman & Kemper, 1996).
The hippocampal system and autism
Recently, functioning of the brain's learning center, the hippocampal system, has also been associated with autism (De Long, 1992; De Long & Heinz, 1997; Boucher & Warrington; Hoon and Reiss, 1992; Raymond, Buaman & Kemper, 1996). De Long (1992) proposed that autism is a developmental syndrome of hippocampal dysfunction. The hippocampus is necessary for learning semantic and pragmatic meaning (Gabrieli, Cohen & Corkin, 1988), discovering relationships between stimuli (Eichenbaum, 1997; McClelland, McNaughton & O'Reilly, 1995 ), and using context to solve problems (Bibbig, Wennekers, & Palm, 1994; Gluck et al, 1997; Nadel and Moscovitch, 1998; Thompson, Moyer & Disterhoff, 1996; Vargha-Khadem et al, 1997). It slowly discerns patterns and relationships across contexts (Eichenbaum et al, 1996; McLelland et al. 1995). It is particularly active when exposed to novelty, a rich environment, and during problem solving tasks (Thompson et al, 1996; Zhu et al, 1997).
The attributes of microdevelopmental methods are particularly well suited to these functions, emphasizing a rich environment, problem solving, categories, and meaning. If autism is indeed related in some way to a deficit in learning meaning and discerning patterns and relationships, then it may be possible to promote this type of learning in autistic children and reduce their autistic behaviors.
The development of play in autism
Abnormal or absent play is one of the distinguishing criteria for autism. Some authors have suggested that assessment of play skills may be useful in the clinical diagnosis of autism (Doherty & Rosenfeld, 1984; Stone et al, 1990). The development of play is thought to reflect a sequential growth of cognitive skills (Belsky and Most, 1981; Cohen, 1987; Fein, 1975; Piaget, 1952). This would suggest that play is an important area to assess when considering the efficacy of an intervention for autism.
Several authors have proposed various schemas that describe hierarchical stages in the development of play that reflect a growth in cognitive abilities. These systems describe the development of play behaviors related to symbolic functioning (Le Normand, 1986; Piaget, 1952; Smilansky, 1968), pretense and substitution (Belsky and Most, 1981; Fein, 1975), or social interaction (Parten, 1932). For example, Piaget's (1952) hierarchy included imitation, practice play, symbolic play, and games with rules; Smilansky used the levels of functional play, constructive play, dramatic play, and games with rules. Belsky and Most's (1981) hierarchy included mouthing, manipulation, functional use of an object, relational use of two objects, functional-relational use, enactive naming, pretending with self, pretending with other, sequence of pretend, and substitution.
All of these systems pertain to normal children and use hierarchies related to age. It is unknown if they are representative of autistic children. An alternative system has been described by MacAlpine (1998) which is not related to age and which was developed through observations of free play in autistic and developmentally delayed children. MacAlpine's system consists of four sequential levels of play (sensory play, exploration of change, social representational play, and role play), with a series of increasingly complex stages within each level. These stages represent a child's ability to explore and understand objects and experiences that are increasingly more complex.
CHAPTER 2: METHOD
Participants
Two separate groups of autistic children participated in this study: 12 children in the United States (Group A), and 6 children in Italy (Group B). All children in Group A were referred to the author by a parent, state agency, or speech pathologist. All children in Group B were referred to the author by a national parent support group for severely handicapped children.
In Group A, there were 10 males and 2 females. All but one of the children was under the age of eight; one was an adult. The adult was 22 years of age, with a genetic disorder and related metabolic disorders. This participant had been in a variety of therapies for several years prior to referral. One child had also had a genetic disorder known to be associated with autism, and one child became autistic after a seizure episode in the second year of life. Nine children in this group were considered severely autistic and the other three children were moderately autistic according to their scores on the Childhood Autism Rating Scale (CARS). All but two children had been evaluated prior to referral by an independent agency and had a diagnosis in the autistic spectrum at the time of referral. The two that were not diagnosed by an independent agency were referred by a parent who was concerned about their children's bizarre (autistic) behaviors. Two children were participating concurrently in speech therapy.
In the Italian group (Group B), there were six children, four males and one female; all were under the age of seven years. All children in this group either had an independent diagnosis in the autistic spectrum (five children), or had a disorder associated with autism and autistic behaviors (1 child). In Group B, one child was severely and multiply handicapped. This child did not have the ability to independently sit, or grasp objects, and had low vision and severe metabolic and immunological problems. One child in this group had a neuropathy (arachnoid cyst), and one child had seizure disorder concurrent with an immunological disorder. CARS scores for four of the children were in the "severely autistic" range, and scores for two children were in the "moderately autistic" range. In this group, one child was participating concurrently in family therapy.
Assessment Measures, Scoring and Coding
Children were assessed at baseline evaluation and follow-up on three measures: 1. the Childhood Autism Rating Scale (CARS), 2. a 42-item Parent Survey of Abnormal Behaviors, and 3. an assessment of developmental play stage.
1. Childhood Autism Rating Scale, CARS (Schopler et al, 1988), is a 15 item clinical survey measure which scores severity of impairment in the following areas: relating to people, imitation ability, emotional response, body use, object use, adaptation to change, visual response, listening response, taste /touch/ smell response, nervousness, verbal communication, nonverbal communication, activity level, level of intellectual response, and general impressions of autism. Each item is scored from 1-4, with 1 representing normal behavior and 4 representing profoundly impaired behavior. Scores below 30 are considered non-autistic, scores between 30-37 are considered to represent mild to moderate autism, and scores between 37-60 represent severe autism. CARS reliability was calculated by the mean agreement of scores from the author and a second rater. There was a mean agreement of 94% between raters (mean total point difference 1.8).
The Parent Survey of Abnormal Behaviors was a report measure in which parents were asked to rate the observed frequency of 42 autistic behaviors and deficits (see Appendix 1). These 42 behaviors were chosen as representing the most commonly appearing abnormal behaviors in the background reports of autistic children and in the autistic literature. The measure focused on language abnormalities, pragmatic problems, abnormalities in non-verbal pragmatics, perseverative behaviors, abnormal social development, abnormal play, abnormal affective development, self-stimulation behaviors, aggressive behaviors, sensory abnormalities, and motor planning abnormalities.
MacAlpine's (1998) developmental play stages were used to assess each participant's developmental stage of play. Play events were coded if they exceeded 3 seconds of interaction. The highest stage of play observed in any play event was coded. The highest stage of play for each was obtained by determining the highest play stage observed in a play event at baseline and follow-up play sessions. Two raters coded the highest stage of play at baseline and follow-up evaluations; there was a mean agreement of 92% for all coded play stages and a mean agreement of 100% for highest play stage at baseline and follow-up sessions. An explanation of each of the fifteen stages of developmental play follow.
Stage 1: Exploration of Sensory Properties
This stage is marked by the exploration of an object's properties, particularly the look, feeling and taste. When a child explores a novel object at this stage, they taste, lick, smell, mouth, or stroke it. If an object moves, the child is content to watch the movement without attempting to stop or alter the movement. The children in this stage explore only one sensory property at a time. If they are mouthing an object, they do not feel or smell it. Sensory exploration is repetitive. For example, a child will explore objects at first only with his eyes or mouth. One child would walk back and forth in front of a set of vertical window blinds, watching how the visual pattern on the corner of the wall or window changed as he moved.
Stage 2. Exploration of Sequential Sensory Properties
In this stage, children appear to be able to expand their sensory exploration to include multiple properties. For example, they mouth or watch an object one moment, then rub or flick it, or mouth it before observing it again.
Stage 3: Exploration of Single Change
This stage is marked by the gradual abandonment of licking, tasting, or mouthing, although children occasionally brush or touch an object with their lips. Play at this stage is marked by an interest in toys that move or pop up, objects that make a simple sound, spatial relationships (putting things in holes, pulling out drawers, tracking toys), and simple changes (popping bubbles, color changes). Play is highly repetitive and centered on a single change such as turning a switch on and off, or turning a handle on a jack-in-the-box so that it pops up. Children are content to explore a toy's movement by altering speed or pressure, or by arresting a movement or changing direction, rather than exploring the mechanism that made the object move.
There is some play with baby dolls and toy animals in this stage, but children treat them as objects rather than representations of people. For example, they may place a bottle in a baby doll's mouth or put it to bed, but they do not bring the doll to their chest, rock the doll, or act nurturing in any way. Play in this stage is highly imitative and repetitive. Children begin to play with vehicle sounds and animal sounds while holding or moving an object. Experimentation with spatial relationships appears in this stage. A favorite toy at this stage is the mailbox, in which children can place mail in the slot, watch it disappear, and retrieve it repeatedly.
Stage 4. Exploration of Multiple or Sequential Change
Play in this stage is exploratory rather than imitative or simple. Children change or act on a toy in more than one way or perform a sequence of changes. Several changes are made in sequence, and actions are linked together. For example, a child might shovel sand into a bucket for a minute, then shovel it into a sand wheel, then into water.
Stage 5: Exploration of Complex Change
Children in this stage are interested in how things go together to create something totally different, and why something works. They begin to seek out activities like cooking, making simple crafts, and playing with sand or clay. Objects that changed in different ways is fascinating to the child at this stage (e.g. electric pianos, toys that changed color in different temperatures of water, beating and cooking eggs, etc). Play is creative and complex.
Stage 6: Exploratory interest in social representational toys
In this stage, children begin to briefly approach dolls or a dollhouse. They often pick up a doll and then drop it, or carry a doll with them without playing with it. They do not respond to a doll if someone makes it "talk", or use the doll in any type of scenario play. Children also explore doll furniture at this stage by touching or moving it. This appears to be a stage of curiosity about representations of people.
Stage 7: Responsive play with social representational toys
In this level, children begin to respond to dolls or toy animals as if they were people. For example, a child would feed the doll when the doll "asked" for food, or use a simple "yes" or "no" or gesture when a doll was manipulated as if it were speaking to them. Multiple actions are strung together (e.g. doll eats, goes to sleep, wakes to play). Occasionally, a child uses third person narration rather than at this stage, such as "he could go in the truck" or "he is going to the store", or a combination of first and third person narration, such as, "go to bed, the little boy said".
Stage 8: Simple representational play
At this level, children begin to use a doll or stuffed animal to represent self. At first, this play is very simple. The child uses a doll or stuffed animal as if the doll were "speaking". At this level, play is highly imitative. The children use toys to recreate familiar conversations and actions that have been heard or seen at home, on TV, or from play scenarios observed with siblings. Play is primarily egocentric, with the child only having one doll or toy at a time "speak", and the doll usually represents self. There is only one scenario at a time in this stage, such as eating dinner or sorting animals at the farm. Scenarios are not linked together (eating dinner, then going to the park).
Stage 9: Simple sequential representational play
Children give several dolls the same name at this level, and this is usually their own name or the name of a familiar person. The children play with only one doll at a time, dropping one to pick up another, however they begin to link more than one scenario together in their play (going to the park, going to the store, eating dinner). Scenarios are stereotypical and lack creativity.
Stage 10: Creative Social Representational Play
At this level, children use several different scenarios linked together. Dolls have different names in this stage. The children use dolls in sequence (dropping one doll to play with another doll) rather than interactively.
Stage 11: Complex or Creative Social Representational Play
At this level, children use reciprocal conversations between more than one doll or animal, speaking for several dolls or toys as if they were separate and different people. Imaginary characters are manipulated in complex and rich scenarios (pirates, knights, astronauts, cowboys, etc.), and play is varied and creative. If there is a play partner during this pretend play, the children direct the partner with constant demands to comply with their own scenario. The children seem unable at this stage to adapt their play scenarios to another partner's adjustments or ideas. One child directs and commands the script, and resists any suggestions or redirections from a partner, but may take turns with the partner in directing the script.
Stage 12: Creative Social Representational Play (with partner)
In this stage, children are able to adapt to a play partner's scenario and adjust their own scenario in response. There are efforts at collaboration and repeated directions addressed to play partners (e.g. "let's pretend that...", or "just suppose...", etc.). This collaboration often changes the script's direction.
Stage 13: Simple role-play
In this stage, children use costumes or simple props to support a pretend scenario about a simple action (e.g. playing basketball with a real basketball) or person (e.g. putting on a costume and acting like a fireman). Children at this stage have little information about how people act, or about how various events occur. Scripts are repetitive and imitative. For example, a child pretended to be a fireman, put out a fire, and then ended the game. Another child pretended to be a policeman pulling over a speeding car, and then end the game. Scripts are also highly familiar; children act out familiar fairy tales or familiar family rituals.
Stage 14: Sequential Role-play
In this level, children are able to play the role of several different characters in a familiar story. They have difficulty, however, creating novel or unfamiliar roles or playing the role of unfamiliar or novel characters.
Stage 15: Creative Role Play
In this level, children begin to create longer and more complex scripts and scenarios. They are able to incorporate novel events into their play that had not been previously seen or experienced. Children at this stage can invent entirely new characters and scripts on their own.
Analysis of Developmental Stage of Play
Initial and follow-up sessions were coded for highest developmental play stage observed. for all children in both groups. For six children in Group A, play events (engagement followed by disengagement with an object) that lasted longer than three seconds were coded from video tapes of play sessions. A microanalysis of the progress through the play stages was charted for each of these children.
Procedure
Intervention Method for Group A
In Group A, children came with a parent to the author's home for twice per week play sessions over a mean period of 8 months (range 5-12). A large number of toys and objects from all play stages were placed within easy reach on the floor and on tables or ledges around the room. Toys were chosen as exemplars of categories (e.g. balls, objects that blew air, bumpy objects, objects that had a spiral movement, etc.). Each toy shared some feature in common with at least two other toys in the room, so that there were multiple exemplars within each category.
If a child explored an object, one of the category related objects was offered. The similarities between the objects were explained in simple words and sign language. Signs were used to mark important words in the sound stream (nouns, verbs, spatial relationships). The movement of the gestures provided a means to direct the child's attention to the object or relationship and provided a means to visually cue and pantomime an object or movement.
The child had control over the type of object to explore as well as the manner and duration of this exploration. The author acted as a facilitator and performed the following functions: narrating the child's actions when appropriate, providing other toys within a category, offering new (stage appropriate) toys if a child became bored or disorganized, and demonstrating features of a toy for the child.
If a child's play became repetitive or ritualistic, a small change or variety was introduced into the ritual. For example, the child's movements were interrupted by a barrier, by removing something in the ritual, or by altering slightly the orientation of the toy or the direction of movement of part of an object. A more detailed explanation of these methods are included in Appendix 2).
Sessions were intermittently videotaped when changes in play skills were observed, or when there had been several play sessions in a row that had not been videotaped. Sessions were video taped about once per month and coded by two raters for play stages observed.
Intervention Method for Group B
In Group B, parents came with their child to an evaluation session in which the author observed the developmental stage of play for each child. Evaluations were video taped and CARS scores were later performed from these videos by the author and another independent judge. After the evaluations, the parents came to a 4 hour training seminar (presented in English and translated into Italian) in which the author presented the theory, rationale, and methods used in the therapy. Parents were allowed to ask questions and participate in discussions with the author and other parents during this seminar. Parents were given an Italian translation of the author's therapy manual, which contained detailed information about play stages and appropriate toys and activities to use in each stage.
After the author left, there was minimal or no contact between the author and families of the children until follow-up 8 months later, as the parents lived in different cities in Italy.
Validity Issues
It is important to consider whether the children's progress to recovery demonstrated in this study can reasonably be considered a result of the category learning provided by the intervention, rather than some unknown cause. Was there sufficient external, internal, and construct validity associated with these results to support considering replacing traditional approaches to the treatment of autism with these methods?
External validity is assessed by the ability to generalize across persons, settings, and time. The ability to generalize across persons was enhanced by including children in both groups who had other disorders known to be associated with autism (metabolic disorders, genetic disorders, neuropathy, and sensory impairment). These individuals have typically been excluded from studies, and their presence, based on the literature, would make it more difficult to achieve a strong treatment effect.
The ability to generalize across settings was also enhanced by the inclusion of a second group in Italy, who had only minimal training. Any treatment effects that were similar across groups could not be attributed to a particular setting, since each parent in the Italy group treated their own children and these parents had minimal to no contact with each other.
The effects could not reasonably be attributed to maturation or to time, in therapy, since the autistic literature indicates that such improvements are unlikely (Ballaban-Gil, 1996; Eaves and Ho, 1996; Lovaas et al, 1973; Mesibov et al, 1989; Smith et al, 1997). Kazdin (1998) suggests that ethical problems may arise in the use of control groups when treating an impaired population in need of intervention. In these cases, Kazdin suggests that it is possible to compare the treatment effect to comparison groups that are treated with traditional methods. In order to determine if the changes noted in this study were comparable to, or better than, changes reported in the literature for traditional therapies, comparison groups were obtained by reviewing the literature. Comparison groups were sought in studies that reported long-term or follow-up effects of treatments for autism using either the CARS or changes in one of the 42 behaviors covered by the Parent Survey of Abnormal Behaviors.
Two studies (Eaves and Ho, 1996; Mesibov et al, 1989) reported changes in CARS scores, and four studies (Ballaban-Gil, 1996; Lovaas et al, 1973; Smith et al, 1997; Zappella, 1990) reported changes in one or more of the behaviors from the parent survey (echolalia, aloofness, and self-stimulation). These studies were used to compare improvements seen in Group A and Group B in this study.
Internal validity is assessed by the ability to determine a causal relationship between variables. Can a cause and effect relationship be reasonably suggested between the treatment methods and the changes in the outcome measures? Two arguments suggest that this is possible.
First, the children in both groups had received a diagnosis of autism or autistic disorder prior to their referral for intervention. All but two of these children had participated in other types of therapy prior to this study, yet were still considered autistic at the time of referral. Only two of the children in the American group and one of the children in the Italian group were participating in other types of therapy during the study.
One child was participating in a behavioral therapy program, one was receiving psychotherapy, and one was receiving pragmaticism speech therapy. These children had been participating in these programs prior to the study and their diagnosis had not changed by the time of referral.
Second, the autistic literature indicates that recovery from autistic disorders is not expected using alternative methods (Schopler, 1983; Mesibov et al, 1989; Eaves and Ho, 1996). A demonstration of recovery, therefore, would not be expected to result from traditional therapies or from maturity. The validity of the assessment of recovery was also enhanced by the use of multiple outcome measures and dual raters.
The CARS is a well-established objective measure of diagnosis (Van Bourgondien et al, 1992). A criterion-related validity study (Eaves & Milner, 1993) indicated that the CARS was capable of identifying 98% of the autistic subjects; the authors of the CARS (Schopler et al, 1988) reported a criterion-related validity correlation between the CARS and clinical ratings of r= .84 (p<.001). The authors of the CARS reported a mean agreement between CARS scores produced by a group of clinical directors and a group of other professionals as 92% (Schopler et al, 1988, p. 6).
CHAPTER 3: RESULTS
Results of the Microanalyses
All play events longer than three seconds were coded if the interaction was visible on the sample. The highest and the lowest play stages observed were recorded by session. The results of these analyses are represented in Figures 1 through 6. In general, the coded play events demonstrated that there was a general trend to higher and more complex stages of play, but embedded within that trend were movements up and down between several play stages. The children all played at Stages 3 and 4 for several months; this long period preceded a move into social representational play in all but two children (Child 3 and 4), who had begun intervention at lower stages of play than the other four children.
Microdevelopmental patterns across sessions, within sessions, and within play events demonstrated oscillations between play stages and variability in the nature of these oscillations. Children typically explored objects that were novel initially at lower levels than they explored familiar objects, then progressed to more complex forms of play as they began to become familiar with the object. There was an extensive period of lower level (Stages 3-5) play, which lasted several months and which preceded movement into social representational play.
Microdevelopmental oscillations within and across sessions
Regressions typically occurred prior to a movement to higher play stages. This can be seen in Figure 1. See Child 1 in weeks 10-12. The difference between the highest and lowest play stages observed within a session spanned many play stages. These differences were greatest at the end of a movement to a higher stage of play. Examples of this appear in week 10 in Child 1, week 13 in Child 2, week 16 in Child 3.
Extensive periods of lower stage play
The charts of all of the children demonstrated an extensive period in which the highest play stage observed was in Stages 3 to 5. For example, Child 1 spend 18 weeks in these stages, Child 2 spent 9 weeks, Child 3 spent 40 weeks, Child 4 spent 30 weeks, Child 5 spent 6 weeks, and Child 6 spent 23 weeks in these stages. The mean time spent in these stages for each child was 21 weeks, or 5 months. In four of the children (Child 1, 2, 5, and 6), this extensive period in Stages 3 and 4 directly preceded a move into higher stages of play (social representational, Stage 6 or above).
Stage related behaviors
Preferred objects and toys were highly related to the stage of play for all six children. When a child was playing in Stages 3 - 5, there was a clear preference for toys that represented one of three categories: 1. movement, 2. spatial relationships, or 3. transformations. The children played with toys and objects in these categories nearly exclusively during sessions where Stage 3 and 4 play was the highest play stage observed.
Moving toys were explored in most Stage 4 play episodes. Toys in this category included racetrack toys, toy skaters on ramps, electric tracking toys, marble runs, ball runs, pound ball toys, flying toys, and toy trains.
Children also explored spatial relationships in Stage 3 and 4 play. Toys from this category included objects with doors and drawers, lock boxes and purses, mail box toys, containers with holes, ejecting toys (bow and arrow, cannons), and toys that popped up. Play that focused on transformations in Stages 3 and 4 could be divided into two categories: exploration of sound transformations, and exploration of shape transformations. In Stage 3, some children preferred simple musical toys, such as bells, xylophones, drums, or music boxes. These children demonstrated little interest in moving objects until they had explored sound for several sessions (for example, Child 5 during weeks 1-3). In Stage 3 play, the children typically ignored toys that had multiple changes (electric pianos, pushing the frets on a guitar while strumming, etc.). However, the reverse was true of Stage 4, in which children rejected toys with only simple changes (bells, drums, etc.) in favor of pianos, guitars, and other toys that offered different types of sounds.
Toys that offered transformations of shape included play dough, Hoberman spheres, and toys that stretched and recovered their shape. In Stage 3, the children preferred simple toys in this category such as slinkys, stretching dolls, and Hoberman spheres. In Stage 4, the children abandoned these type toys in preference for play dough, clays, and cooking in the kitchen.
The children appeared to abandon exploration of a concept only after extensive exploration of that concept over time. If the highest stage of play exhibited by a child was Stage 3 or 4, they generally accepted toys that moved, changed spatial relationship, or changed sound or form. However, they rejected other types of toys (a doll, puppet, simple craft, cooking experiment, game, etc.) in these stages. Toys that were extensively investigated in Stage 3 and 4 were simple; play was repetitive.
Toys appropriate to Stage 3 or 4 were generally rejected when a child was capable of at least Stage 5 play. Objects that were explored extensively were always stage appropriate.
Play with baby dolls (but never other types of dolls) appeared in Stage 3 and Stage 4, but the play at this stage was qualitatively different from higher level social representational play that appeared later. In Stage 3 and 4, children appeared to be exploring the spatial relationships of putting a blanket on a baby, a baby in a bed, or a bottle to a baby's mouth, rather than in the doll as a representation of a person. The child would repeat this action several times with little or no variation in much the same way as they placed mail in a mailbox, opened and closed drawers, or placed objects in containers. No child in this stage ever appeared to nurture the doll, bring the doll to their own body, rock the doll, or comfort the doll. One child placed a bottle in a baby doll's mouth without even looking at it.
In Stage 5, suggestions to help in the kitchen (making cookies, cutting fruit, or play with dough) were accepted. Prior to this stage, children wandered out of the kitchen or observed passively for a brief period.
In Stage 6 and above, an interest in the dollhouse was nearly universal and usually preceded interest in animals, knights, pirates, spacemen, divers, cowboys, or puppets. Play in Stages 6 and 7 was nearly exclusively limited to dollhouse objects, although play at Stage 8 and above tended to span all of the representational toys in the room.
The length of engagement with any particular toy tended to be fairly brief (8 minutes 21 seconds was the longest duration of play for any sample), however play with similar objects related to the same concept (movement, changing properties, etc,) usually lasted an entire session. Objects that were novel and unfamiliar were often rejected, picked up and dropped, or played with for only a few seconds. Many play events for objects that were novel and unfamiliar could not be coded because the duration of interaction was less than 3 seconds, the minimum set for coding a play episode. As a child became more familiar with an object, play was extended.
To summarize, throughout the period of development, the children showed microdevelopment from lower to higher stages of developmental play. Through the processes of oscillation and regression, they progressed to higher levels.
Results from the Macroanalyses
The length of time in intervention for each participant in Group A ranged from 5 to 12 months from baseline to the first follow-up; the children who failed to score in the non-autistic range at this follow-up had an additional CARS assessment 3 months after the first follow-up assessment. (see Table 1)
Table 1. Length of Time in Intervention: Group A
Child |
A |
B |
C |
C |
E |
F |
G |
H |
I |
J |
K |
L |
| Total Months in Intervention Follow-up 1 |
12 |
12 |
6 |
6 |
8 |
10 |
10 |
7 |
5 |
6 |
12 |
6 |
| Total Months in Intervention Follow-up 2 |
10 |
9 |
15 |
Changes in CARS Scores
In Group A, the mean score on the CARS prior to intervention was 42.08, and the mean CARS score at first follow-up was 25.79, a drop of 17.12 points. The mean decline in CARS scores per month for this group was 2.05 points. In Group A, 9 children scored in the non-autistic range at the first follow-up 5-12 months later. The 4 children who were still in the autistic range were in the moderately autistic range at this time. One of these children moved, and a second follow-up was performed three months later for the other three children who remained in intervention. All of these children scored in the non-autistic range on the CARS (see figure 3).
In Group B, the mean CARS score prior to therapy was 41.83, and at follow-up the mean CARS score was 28, a drop of 13.83 points. In Group B, three children scored in the non-autistic range at follow-up; one of these had initially scored in the severely autistic range, 1 had scored in the moderately autistic range, and 1 had scored in the mildly autistic range (see figure 3). The mean decline in CARS scores per month for this group was 1.78 points.
A Wilcoxin signed rank test was performed to determine if changes in autism severity ratings (severe, moderate, non-autistic) between baseline and follow-up were significant for either group. The Wilcoxin test is a non-parametric test for before/after studies. The alpha value of .05 was used for all statistical tests. The changes in severity rating for Group A at the first follow-up was significant, T (N= 12) = 3.06, p = .002. The difference in severity ratings for children in Group B at follow-up was also significantly significant, T (N =6) = 2.20, p = .027.
Informal follow-up of released children
Published results using CARS as an outcome measure
Two published follow-up studies were found that used the CARS (Eaves and Ho, 1996; Mesibov et al, 1989). The mean decline in severity on CARS scores reported by Eaves and Ho (1996) for 76 children who participated in special education, speech/language therapy, and behavioral training for an average of 4 years was 3 points (no standard deviation reported). The mean decline reported by Mesibov et al (1989) for children who participated in behavioral therapy (TEACCH program) over a mean of 7.2 years was 2.93 points (no standard deviation reported). The mean decline for Group A on the CARS was 15.87 points at the first follow-up (standard deviation 7.01) and 17.12 points at the second follow-up (standard deviation = 7.24), and for Group B was 13.83 points (standard deviation 5.56) (see figure 9).
Published results using autistic behaviors as outcome measures
Four studies were found that reported changes in three autistic behaviors included in the 42-item parent study of abnormal behaviors. These were echolalia, aloofness, and self-stimulation. The comparison studies were: Lovaas et al (1973); Smith et al (1997); Ballaban-Gil et al (1996); and Zappella, 1990. Table 2 demonstrates the difference in percentage of participants who demonstrated any improvements in these target behaviors.
CHAPTER 4: DISCUSSION
The results of the three outcome measures in this study indicated that a microdevelopmental approach was capable of producing developmental progress in the autistic children in this study. In spite of the inclusion of children in this study who had associated genetic, neurological, sensory, or metabolic impairments, all of the children in Group A had achieved scores in the non-autistic range on the CARS by the time of the second follow-up (three months after the first follow-up). The results achieved by Group B demonstrated that a short workshop was sufficient for teaching parents to apply these principles effectively.
Achieving a non-autistic score on the CARS did not necessarily mean that the children were completely without impairment, as development continued to progress even within this category. Children who scored in the upper half of the non-autistic range had abnormal ratings in 27% of the behaviors on the Parent Survey of Abnormal Behaviors. Children who scored in the lower half of this range had only 7% abnormal ratings on this survey. These children appeared very similar to peers, with the exception of language delays. However, all of the children were no longer aloof, had begun to communicate their needs and interests to others, no longer insisted on sameness or rituals, and had lost their flat emotional affect.
It appeared that relapse was unlikely in these children. The second follow-up for Group A demonstrated that the children who were reassessed had continued to make progress toward recovery. The three children who were seen informally one year after release had also continued to make progress, achieving higher play stages and improving social and language skills.
Age was not a meaningful construct for understanding the developmental patterns of these children. The children were comparable in play abilities and social skills with younger children. The children who were operating at Stage 1 or 2 (sensory stages) at the first evaluation appeared much like infants, in spite of the fact that they were school age. The children who were operating at Stage 3 or 4 seemed to play like toddlers.
Initially, the children had difficulty understanding the meaning of their experiences. It appeared as if the concepts of objects, movements, spatial relationships, and transformations, were novel concepts for them. Microdevelopmental methods facilitated their ability to obtain meaning about objects and events. The inclusion of sign language also seemed to facilitate the child's ability to understand meaning, since the gestures marked the significant words in the sound stream (nouns, verbs, spatial relationships) and were essentially pantomimes. Signs appeared to help the children connect the words with the objects or actions to which they referred, in order that the meaning of these things could be established.
When the children were aided in obtaining meaning from their play experiences, development progressed. What they learned in one stage became the foundation for higher, more complex forms of learning at higher stages. As development progressed, children lost their autistic behaviors.
In the last few decades, researchers have sought a neurological explanation for autism. There have been many neuroimaging studies of autistic adults using MRI (magnetic resonance imaging), PET (positron emission tomography) or SPECT (single photon emission computed tomography). Findings in these studies have been variable. In a review of the literature, Rumsey (1996) concluded that the neuroimaging research has failed to demonstrate consistently or convincingly that there is abnormal brain metabolism or blood flow in autism.
There have been only a few anatomical studies from autopsies on the brains of autistic individuals. Abnormalities in all cases were limited to limbic areas (the hippocampal complex, subiculum, entorhinal cortex, amygdala, mamillary body, and medial septal nucleus) and the cerebellar circuits (Rumsey, 1996). Bauman & Kemper (1985) and Ritvo et al (1986) found reduced Purkinje cell counts in both hemispheres and vermis of the cerebellum.
Anatomical studies that have used MRI have had variable results, including a reduced number of cells (hypoplasia) in cerebellar lobules VI and VII (Courchesne et al, 1988; Murakami et al, 1989; Piven et al, 1992), an increased number of cells (hyperplasia) in these areas (Courchesne et al, 1994), or normal cerebellar lobules (Garber and Ritvo, 1992; Hashimoto et al, 1992; Hashimoto et al, 1993; Hsu et al, 1991). Some studies have reported smaller brain stems (Hashimoto et al, 1992, 1993) in autistic and mentally retarded individuals, a thinning of the corpus callosum (Courchesne et al, 1993; Courchesne, Saitoh et al, 1994), and normal frontal lobes (Courchesne et al, 1993), or altered symmetry of frontal lobe volume (Hashimoto et al, 1989).
One explanation for the variation in these findings is the presence of confounds related to small sample size, the inclusion of individuals with co-morbid conditions such as epilepsy, or mental retardation, the inclusion of individuals on medication, age related changes, and the use of non-normal (children to adults or autistic to mentally retarded) control groups (Rumsey, 1996). Another explanation is that there are a variety of factors that may contribute to the development of autism, and that these factors are not evenly distributed in small samples.
There must be some way, however, to explain the fact that autism has a similar clinical pattern of behaviors and deficits across individuals with different abnormalities. One explanation may be that various neuropathies (or sensory impairments, metabolic or immunological disorders) may interfere in different ways and at different stages of information processing in the brain. If an individual is unable to understand the information that their senses provide, then it would be difficult for learning and cognitive development to proceed.
All sensory information is eventually routed to the hippocampal formation, the center of learning and memory. If sensory information is distorted or absent, it may affect the ability of the hippocampal formation to adequately process information. Indeed, De Long (1992) has suggested that impaired hippocampal functioning may explain the clinical, pathological, and neuropsychological patterns of autism.
Pathologies of the hippocampal formation have been discovered in autistic children with a history of seizures (De Long, 1997) and also in children without seizures (Raymond, Bauman & Kemper, 1996). The hippocampal pathologies reported in children without epilepsy were consisted with arrested development, rather than malformation (Raymond, Bauman & Kemper, 1996).
The hippocampal area is necessary for constructing semantic and pragmatic meaning about the world and using contexts to solve problems (Bibbig, Wennekers, & Palm, 1994; De Long, 1992; Eichenbaum, 1997; Gluck et al, 1997; Hoon and Reiss; 1992, Nadel & Moscovith, 1998; Raymond, Bauman, and Kemper, 1996; Vargha-Khadem et al, 1997). These abilities are impaired both in autistic children and individuals with hippocampal damage (De Long, 1992; Gabrieli et al, 1988; Martin et al, 1997; McClelland et al, 1995). The hippocampal formation is also involved in the reconstruction of memories (Arnold, 1984; DeLong, 1992). Boucher and Warrington (1976) demonstrated that the pattern of deficits in memory function in autistic children was similar to that of amnesic adults with hippocampal damage. In addition, animals that have had their hippocampus removed (ablated) have difficulty screening out irrelevant stimuli or responding appropriately to novel stimuli, show rigid and stereotypical behavior, perseverate, and increase motor activity, deficits which also apply to autistic children (DeLong, 1992).
De Long's (1992) theory that impaired hippocampal functioning is related to autistic behaviors proved a useful guide for the design of the intervention method used in this study. Since the brain is plastic and capable of reorganization, activities that stimulate the hippocampal area or require its activation may have an effect on autistic behavior. Indeed, animal studies have provided evidence that learning (an activity that requires the hippocampus) can create significant changes in the hippocampus (Kempermann et al, 1997; Rosenzweig & Bennet, 1996; Mohammed et al, 1990; Schrott, 1997; Thompson, Moyer, & Disterhoff, 1996), and have a positive effect on behavior (Kolb & Whishaw, 1998).
The hippocampal system slowly discerns patterns through repeated exposure to items that are similar or related (McClelland, McNaughton & O'Reilly, 1995). In this study, all objects in the playrooms were related by category to several other objects, which provided the children with opportunities to explore their similarities and relationships. The children, by directing the activities, were able to spend as much time as necessary investigating each category.
The hippocampus is preferentially activated by novelty (Zhu et al, 1997). In this study, novelty was provided by the rich and varied assortment of toys, games, and objects in the playrooms, as well as by novel toys that were introduced throughout the study. Novelty is one of the principles for promoting microdevelopment specified by Granott (1998a). A second principle that guides Granott's methods is to create a fun learning environment. This was produced by allowing the children to control the situation, by providing interesting objects and activities for them to explore, and by treating them with dignity and kindness at all times. Since hippocampal activity is disrupted by stress (Reagan & McEwen, 1997; Vicedomini et al, 1986), respecting this principle may be of particular importance, since autistic children endure high levels of stress in general due to their inability to communicate or understand others.
Changes in the hippocampus are related to learning, rather than just memory or performance of skills (Thompson, Moyer & Disterhoff, 1996). Therefore, this study provided objects and activities that invited exploration and investigation (toys with moving parts, electric tracking toys, toys that changed in unusual ways, etc.) rather than activities that involved skill drills. It also avoided the use of rewards, which have been demonstrated to lower a child's motivation to participate in a task and to learn independently (Kohn, 1993).
Granott's (1998a) general principles for promoting microdevelopment dictate that a learner should have independent control in a fun, novel, and interesting environment in which there are sufficient stage-appropriate activities to promote problem solving and discovery of concepts. These principles, when combined with sufficient category-related objects, pragmatic explanations of actions and events, and a focus on meaning, establish a learning environment that may help activate the hippocampal area of the brain.
Conclusion
All of the autistic children in this study made dramatic progress in development, regardless of the severity of their initial diagnosis or the concurrence of related problems such as genetic disorders, immunological disorders, seizures, neuropathies, or sensory impairments. Parents with only a few hours of training were able to treat their own children as effectively as the author.
This study suggests that autism may be related a child's inability to understand the meaning of what they observe and experience during the period in which children typically acquire these concepts. When the children acquired concepts related to objects, transformations, movement, and spatial relationships, they progressed to higher stages of developmental play and lost their autistic behaviors.
A focus on the process of microdevelopment is a very different approach than those currently used to treat autism. While family therapy focuses on social interactions, this method allows the child to first develop an understanding of basic concepts related to objects, space, movement, and transformation of properties, before exploring the complexities of social relationships. While drug therapy focuses on altering brain chemistry to reduce behaviors like aggression or hyperactivity, this approach helped a child acquire meaning, which reduced the need for the behaviors (Arwood, 1991).
Microdevelopmental methods offer a different approach to treating autistic children than traditional methods based on punishment and reward systems. Behavioral methods emphasize adult control and focus on the goal of compliance. Lovaas (1973) described his philosophy in this way:
"Throughout, there was an emphasis on making the child look as normal as possible, rewarding him for normal behaviors and punishing his psychotic behavior, teaching him to please his parents and us, to be grateful for what we would do for him, to be afraid of us when we were angry, and pleased when we were happy. Adults were in control." (p. 135-136)
Microdevelopmental methods, on the other hand, offer a different approach. Children are in control of their own development. The adult acts as a facilitator and guide, to help when tasks become frustrating, and to provide appropriate experiences, and to offer objects that allow the child to develop concepts and solve problems. The adult and the child act as partners in the learning process. In this environment, learning and development appeared to flourish, and autistic behaviors disappeared.
APPENDIX 1: PARENT SURVEY OF ABNORMAL BEHAVIORS
Please score each behavior below according to this scale:
1= normal behavior in this area
2= does this rarely or occasionally
3= does this often
4= usually or always does this
ABNORMAL BEHAVIOR ----------BEFORE PROGRAM----------- NOW
Hyperactive or hypoactive
Avoids eye contact
Avoids people
Ignores people (seems aloof)
Does not turn to new sounds
Does not turn to name called
Does not play with dolls, puppets, etc.
Does not explore toys
Explores objects by mouthing
Uses self stimulation
Does not play with children
Does not play with siblings
Insists on rituals
Insists on sameness
Has aphasic speech
Does not look at person spoken to whom they are speaking
Does not use "I" or "me"
Does not point
Does not turn to person speaking
Does not smile at people
Has blank facial expression
Screams frequently
Bites self
Head bangs
Hits self
Hits or bites others
Refuses most food
Has difficulty using hands
Has difficulty pulling or pushing
Has difficulty undertanding others
Has difficulty turn taking
Is hypersensitive to being touched
Is hypersensitive to some noises
Rocks, swings, or jumps excessively
Has difficulty tracking moving objects
Does not enjoy social interaction
Can not reach across midline of body
Tends to be aggressive
APPENDIX 2: USING MICRODEVELOPMENTAL METHODS WITH AUTISTIC OR DEVELOPMENTALLY DELAYED CHILDREN
1. POINTING AND NON-VERBAL COMMUNICATION
Demonstrating the act of pointing a finger
The act of pointing is an important part of non-verbal communication; it directs someone's attention to something, and it can serve as a request for help. Children with developmental delays often do not develop this important method of non-verbal communication without assistance.
If a child takes an adult's hand and places it on an object (often to request their help), then the adult can respond by taking the child's hand and molding it into the pointed finger gesture immediately before helping them. For example, if a child takes an adult's hand and places it on a door handle (to request that the adult open the door), the adult can place their hand over the child's hand, extending the child's first finger and pointing to the door, then immediately opening the door. The act of pointing should be immediately be followed by the action that the child desired, so that they can connect this gesture with obtaining their goal.
Some children have developed the "proto", or early, pointing gesture, in which they extend their entire hand, rather than just one finger. Since adults rarely understand this gesture, you can help the child learn the traditional pointing gesture by placing your hand over theirs when they use a proto point and gently remold the hand into the proper position while you are acknowledging their intent (i.e. "look at the ...?, "get the ...?", "give me the....?", etc.).
Demonstrating the meaning of pointing
Often, a child does not use a point because they have no idea what it means. These children do not respond to another person's pointing by directing their gaze in the direction of the point. Here are two easy techniques to help a child learn the meaning of pointing. The first is to wait until the child requests something, then point to it before you give it to them. The second is to wait until the child is looking for something, then point to the object and the run their hand down your pointing arm, down the pointed finger, until they discover the object at the end of the point.
A child often does not understand that the extended arm is targeting their gaze in the direction of an object. One game that helps this problem is to use water guns to shoot lines of water on a dry sidewalk or patio. The water track provides a visual marker of the direction of the pointed water gun. You can also let the child use water guns to shoot at targets. Occasionally, I place the child's first finger along the barrel of the water gun, then wrap their hands around the trigger. This action creates a motor memory for the hand gesture of the pointed finger, which will then be related to the visual water track of the pointed gun.
2. COMMUNICATING WITH EYE GAZE AND MUTUAL GAZE
Using mirrors
One of the most effective methods of helping a child tolerate and understand the gaze of others is to place mirrors around the house at their eye level. Children usually can tolerate the gaze of others in a mirror, even when they can not tolerate direct gaze. If they can watch others in a mirror, they will begin to learn important information about their own face as well as how it is the same or different from other faces. This is a safe way to help a child begin to associate visual cues (facial expressions and body language) with emotions and social interactions.
Bringing objects to mouth or face
If a child is interested in an object, you can occasionally bring the object close to your mouth when you name the object. The face (movements of the lips and tongue) carry important information that helps us understand language. (It is difficult to understand a conversation from across a room unless we can see the faces of the speakers).
Interrupting an action
If a child is completely aloof and rarely uses spontaneous speech directed to another person, I will OCCASIONALLY frustrate them in a very SMALL way by interrupting some action in order to obtain their attention. For example, I might hold in my hand a toy that they were reaching for, or I might stop or interrupt some action that they were watching or performing (leaning across a slide to stop their descent, or placing a hand on a tracking toy to interrupt its movement). Then, I take their hand and touch it to my face and model an appropriate reaction ("Move, Michelle", or "give it to me"), and then I immediately respond as if they had made the request themselves ("I will move", "here it is").
3. LANGUAGE
Connecting words to referents
The first goal in helping children develop language is to help them learn that a particular sound refers to a particular thing (its referent). The easiest way to do this is to obtain a set of objects that all share something in common. Then, give the same label to each item in the category of objects. For example, I might give a child a rubber ball, a baseball, a ball with flashing lights, a ball that makes noise, a bumpy ball, a string ball, etc. Each time the child began to explore one of these items, I would say, "ball". Since each item is labeled, "ball", and since each item differs in size, texture, color, and sound, then the label must refer to the thing that they all have in common (a round shape). The use of categories of items in this way helps the child determine the quality that is being labeled by a word.
Encouraging responsive language
Responsive language refers to words (or linguistic signs) that are used to respond to another person. This might be imitating a new word, responding to a question or choice, or signaling agreement or disagreement (using "yes" or "no", shaking the head or nodding). Responsive language is important because it lets us know that the child is acknowledging other people (even if for only a moment) and is responding to them.
Encouraging spontaneous language
Spontaneous language is the use of language to communicate with others or wants to use language to express a thought. Children do not use spontaneous language as simple repetition or labeling, but to express themselves. To use spontaneous language, a child must have words for things that they want to express ("more", "eat /drink", "no", "on/ off", or the names of favorite foods, toys, are often acquired first). So, first we must facilitate a child's acquisition of words. Then, they must discover that words have power to make people do things. We must show them how to use words in this way, and then make sure that they have many successful experiences with obtaining what they want through language. These experiences must be pleasurable, not forced, or the child will associate conflict and anger with language.
There are at least four effective ways to help a child develop an interest in spontaneous language. First, we must make every attempt to guess what the child means when they attempt to communicate with language. If they gesture, model an appropriate word as you give them what they want. If they use gaze to communicate a desire for something, give them an appropriate word and gesture and then give them the desired object. Do NOT hold something hostage to your demands for language. If you do this, they may respond with a word, but it will have no more effect on producing real language than a dog barking for a biscuit.
The second method involves demonstrating the power and fun of language. I like to play games with the children in which I become their "puppet", manipulated by words. These games require that you first introduce words for movements. To do this, I act out a particular word, for example "jump". I say, "Michelle jump" and then I jump around. I say, "mom jump", and mom jumps around the room, "Barney jump", and a toy Barney is made to jump around the room, etc. When this is done, I then ask the child, "Michelle jump?" If the child answers, "jump", I begin to jump. I then say, "stop", and then I stop jumping. I turn to the child and ask, "jump or stop?" , and I wait for their answer. If they do not give me an answer, I model an appropriate choice ("jump?") and begin to do this. I vary this game with different movements and activities, always watching to see if the child is interested or bored. If they lose interest, I immediately stop the game. If the child is interested in this type of play, I tell them that I am going to walk, and they must tell me when to stop. I begin to walk slowly towards an obstacle or wall, repeating the word, "walk, walk, walk" as I go. If the child does not say, "stop", then I slam into the wall or over the obstacle like a clown, making a great joke of the fall. "Oh, you did not tell me to stop, and I walked into the wall." We repeat this game as long as the child is having fun. You can vary it by asking them to tell you to go back, side, jump, fall, stop, etc. It is a wonderful and joyful game, but must be played only when the child is capable of using words in this manner and only while they are enjoying themselves.
The third method is to offer a choice between two items; pause for a moment to wait for a response, then use their non-verbal cues to decide which item they want before offering that item. ("Do you want milk or juice? Milk?......Juice?........ Oh, you want milk. Here's milk.") This offers the opportunity for language and a model of an appropriate way to use it. If your guess is incorrect, then respond to their non-verbal cues with language. ("Oh, you did not want milk. You want juice. No milk, juice.") Begin with a choice between two items; make sure that you have both items visible, and when you talk about one, move the other item out of the way. Once the child is comfortable with a choice between two items, offer a choice between three items.
When a child is comfortable with using language to choose between items that they can see, begin to introduce choices with just words rather than objects. Children with language delays typically choose the last item that you mention, so make that item an unlikely or bizarre choice (e.g."Do you want cookies or salt?"). If the child answers with the last item ("salt"), then offer them the salt. When they look surprised or annoyed, reply, "Oh, you wanted cookies. Not salt, cookies. You want cookies." Then, immediately give them the cookies. Children who have difficulty processing language often retain only the most recent thing that they hear. Placing a bizarre item in this position will force a child to consider more than just the most recent thing heard. Soon, a child may decide that the correct answer is always the first thing mentioned. At this point, you can vary the position of the bizarre item (first or last place in the choice) so that they will be forced to actually process the meaning of the choice.
The fourth method to promote spontaneous language is to model a sentence and pause for a moment to allow the child to "fill in the gap" with an appropriate word or phrase. For example, you might say, "The doll is sitting on the ...........chair; here is her.......... cup, so she can ............ drink". If the child does not fill in the word, fill it in for her and then continue. You can also pretend to not understand a child, and ask a leading question (" You want to go to the ......?).
In general, offering opportunities for spontaneous language, providing a model for appropriate words to use, and reducing frustration by giving them opportunities to use language but not forcing them to do so, will facilitate the development of spontaneous language.
Encouraging combination words
When a child has referents for single words, you can then link these familiar words with one unfamiliar word to help the child learn combinations. In the same way that single words and concepts were introduced, you can introduce combination words. If the child understands the word, "ball", then you can introduce "blue ball", "light ball", "baseball", "bumpy ball", etc. Or, you can introduce "blue ball", "blue car", "blue cloth", etc. Remember to simplify and use familiar things to help learn something new.
Facilitating refinements in language
When a child has developed phrase language, the next step is often the addition of refinements such as definite and indefinite articles ("the", "some", "any", "all", "this", "that", etc.), verb tenses, and adverbs and adjectives. These refinements are more complex in nature than nouns and verbs that represent specific and concrete objects or actions. Verb tenses imply, for example, the ability to use language to connect events in the past and present, to imagine and to remember. These abilities are late to appear in children with developmental delays. Refinements like definite and indefinite articles, relative pronouns, opposites, etc., depend on understanding a set of flexible relationships rather than a concrete relationship. After all, a chair is a chair whether it is close to you, whether there are many chairs, whether it was at the table yesterday, or tomorrow. "This chair", however, means either the chair that is closest to the speaker, or the chair to which the speaker is referring. "Will run" refers not to an action that we can see, but an action that exists in our imagination through faith. These concepts are much more difficult to obtain than concrete concepts, since they represent flexible relationships.
You can facilitate development of this type of refinements in the same way that word referents were related to their concrete meanings. A child must have many experiences with clear examples of a word's meaning in order to be able to use a word. When a word has a relative meaning, you must provide many clear and specific examples of this relationship. Let us consider introducing the words "this" and "that".
You must be aware at all times of what determines the use of a word, and to introduce one of these uses to the child at a time in the simplest way. In this example, the words, "this" and "that" can refer to a relative location to the speaker, or to something that a speaker is indicating through pointing or eye gaze. The simplest choice between these two meanings is the relative location to a speaker, since you can demonstrate it more easily than you can demonstrate eye gaze or intent of a speaker.
To demonstrate the use of "this and that" as markers of relative placement, take two identical objects (perhaps apples) and place them on a table or the floor. Take the child and stand close to one of the apples, point to it, and say, "this apple". Next, point to the apple farther away and say, "that apple". Then, move and stand next to the other apple, and repeat the process. The apple that was closest to the child before is now the apple that is farthest away. The apple that had originally been "this apple" is now "that apple". This is a very clear example of the meaning of "this' and "that", but it is insufficient by itself to help the child obtain the meaning of the words. You must provide many such examples, varying the objects, the distance, the context, etc. until the only thing that remains constant across all of these examples for the child is the relative distance. Close objects were named "this", and farther objects were named "that".
Verb tenses are somewhat more difficult to introduce in this manner, since they depend on memory and imagination. In order to help make the meaning of verb tenses clear, I use pictures. It is best to begin with an activity that is just about to happen. For example, I might begin with a simple walk down the street. I will draw a picture of the child and I walking down the street, and say, "we will walk". Then, as we walk, I say, "we walk". After walking for a few seconds, I again point to the picture and say, "we walked".
You can use a similar technique to introduce verb tenses within the context of a routine, for which there is already a memory and expectation of an event. For example, if the father always leaves the house at the same time, wait until just before the father's departure and say, "daddy will leave". Then, as the father goes out the door say, "daddy is leaving". Finally, when the door closes, "daddy left". These experiences provide meaning for verb tenses. As always, it is easier to obtain meaning when there are many similar (but not identical) experiences with a word.
Addressing pragmatic problems using pictures or dolls
Pragmatic problems refers to difficulty understanding the meaning (particularly the assumed or unspoken meaning) of words and actions. Here is an example of the pragmatics of an action: a policeman raises his hand, and traffic stops. What is the connection between the raised hand and the movement of the traffic? It is the pragmatic knowledge that when a policeman raises his hand (palm outward), any person or vehicle moving towards his hand must stop. This is something that we must learn either by observing the relationship between the act and the reaction of others, or by having someone explain it to us. There is little hope that an alien to our culture would understand the meaning of the cue of a policeman's raised hand, and the expected response of arrested movement, without such observation or explanation. This is equally true of children with developmental delays. They have a huge void in understanding the pragmatic meaning of behaviors. Often, their pragmatic knowledge is missing even for such simple acts. I frequently see delayed children placing dolls crosswise on a bed to sleep, placing a baby doll in a refrigerator, or using cars to "fly" rather than drive down a pretend road. The knowledge of how people act, how cars move, and other such simple things, is missing.
Once we understand that detailed explanation is the only method by which some children can gain knowledge about expected behaviors and actions, then we take care to provide this information. We will explain in detail the purpose of each act. I might say, for example, "My hand is on this drawer handle so that I can pull on the handle and since the handle is connected to the drawer, the drawer will open when I pull on the handle." Even this type of detail may not be enough, so I demonstrate each part of the action. "This is a drawer, it has a handle. Handles are used to open drawers. Do you see how the handle is connected, to the drawer? The drawer and the handle are together, are one. If I pull on the handle, I can pull the drawer. Here, you do it. Do you feel the handle? Pull. See? The drawer opened when you pulled the handle". This is the level of detail and explanation that we use when giving pragmatic information. Do not imagine that the child has any type of pragmatic information. It is best to give them redundant information than to assume that they have knowledge that they do not possess.
One of the most effective methods of demonstrating the pragmatics of some action or word is to act it out. As in the previous example, it is best to use the child themselves to perform an act. If a child pulls on a handle, I know that they understand the pragmatics of force and opening a door. If I pull on the handle, they may only understand the visual relationship between the handle and my hand, and miss entirely the necessity of adding force and pulling on the handle. In fact, this is a very common problem.
A somewhat higher level of giving pragmatic meaning can be accomplished through using another person, doll, or puppet to perform an action. If I am anxious to provide pragmatic meaning about pouring a drink into a cup, for example, I can pour a drink, or use a doll to pour a drink, as the child watches. This is only one small bit of information, however. I must also give meaning to the action itself, not just perform the mechanics of the action. I must describe WHY I pour a drink (so that I can drink from a cup), and WHY I pour it into a cup rather than something else (because it is small and easier to hold a cup in my hands as I drink). I would also act out and describe the difference if I did not pour a drink into a cup but used a plate instead. I would also demonstrate all of the different types of cups that I could pour a drink into and then drink from it. In this way, I establish the cultural boundaries of drinking and demonstrate their meaning and purpose.
The most complex form of providing pragmatic information is also the most convenient, since it does not require props or acting. This method uses simple pictures or stick figure drawings to represent actions and events. This is very convenient, since paper and pencil can be carried anywhere and are available to provide greater meaning through visual representations rather than relying on language. When you draw a pragmatic explanation of something, make sure to make it detailed and to use the same level of explanation that you would use with the other methods. Pictures provide the opportunity to draw attention to facial expressions, body language, direction of eye gaze, and sequence of events.
I have found that video tapes and television also offer opportunities to provide pragmatic explanations. Watch a program with a child and use the events in the story to explain what is happening. Video tapes are particularly effective, since you can stop them at any point and explain something at your what people are doing and why they are doing it.
Responding to echoing
Echoing refers to a repetition of something that was heard. Echoing can be immediate, as when a child repeats a word or phrase, or delayed, as when a child repeats part of a conversation or television show later in a different context. Echoing is normal when it is immediate and is used to confirm a word label or concept. For example, if I offer a child a new toy and say, "weebo", then the child might respond, "weebo" to confirm that this item has been labeled. On the other hand, if I offer a child a new toy and they say, " come down the staircase now", then this is not imitative or normal echoing.
It is very difficult at times to recognize delayed echoing, since children often are able to match a particular learned phrase to various compatible contexts. Sometimes, they only sound a little quirky or odd, as when a child said, "well, what do you think about that?" Other times, it is mistaken for higher level language abilities, as in the children who have a few memorized phrases that they use in appropriate situations (i.e. "I can not find her", or "open this for me", or "we can do that later"). The difference between echoed language and spontaneous language use is the simple fact that spontaneous language is flexible and is consistent in terms of level of language. Echolalia is rigid (the same words or phrases used across situations), and inconsistent with the rest of the language used by the child.
Echolalia is used when a child has not learned the meaning of the words. If a child understands the meaning of "ball", for example, they can correctly apply this word to new objects that are round and that roll. If they are echoing, they may only apply the word "ball" to a particular type of ball, or to a specific ball. If a child understands the meaning of "more juice", then they could also use these terms in other contexts, such as "more milk", or "juice gone".
Since echolalia is a sign that the child is not able to apply the meaning of words, then it is important to apply the techniques for addressing pragmatic problems and the techniques to help language referents, in order to reduce echolalia. Some children are particularly prone to echolalia, and it seems that these children fall back on echolalia as a strategy when they are fatigued, or when the situation is more complex than they can understand. In order to help children at these times, apply the microdevelopmental principles of simplification and reducing complexity. Children who are devoting their attention to a more complex stage of play or to a more complex social or environmental situation will often resort to echolalia because there are few cognitive resources left for them to devote to processing language. If you reduce your language or other cognitive demands on the child and reduce stress levels, then echolalia often disappears.
Responding to chants and verbal perseveration
Chanting is the endless repetition of a word or phrase. It appears to be related to one of two situations; first when the child needs an acknowledgement that the meaning of their words was understood, or when they are trapped in some type of linguistic "loop". In the first situation, the child will repeat something many times until the adult repeats EXACTLY what the child said. When this happens, the child typically appears relieved and nods and says, "yes". In this situation, the child does not know that another person can understand their meaning unless that person repeats EXACTLY what they said.
For example, if a child chants, "I want to go to the store", and I reply, `OK", there is no verbal indication that my words are related to the child's request. It is a pragmatic problem. If I reply, "We will go later", the child still has no verbal match between their words and mine, so they continue to repeat their request. However, if I reply, "You want to go to the store? Yes, you can go to the store", then there is enough of a match between my words and the initial request that the child understands that I have received the meaning of the language. To help a child understand that meaning can be received without such an exact match between our words, first use the matched reply, then add some additional words like, "OK means that I understand that you want to go to the store. When I say we will go later, it means that you can go to the store later with me." This provides a connection between the matched response and a pragmatic response ("OK") or a response that has a limited match (".... will go..."). With many such experiences, the child will be able to discover that the pragmatic response ("OK") or the limited match response (....."will go".....) implies understanding.
Exploring gaps in knowledge
You can NEVER overestimate how deep, how wide, and how many gaps there are in the pragmatic knowledge of a child with delayed language development. Even in children with excellent language, you will find strange gaps in their knowledge of the most basic or simple categories of information. For example, a child who is doing well in first grade and who has excellent language may not have a clue about the nature of smoke or the word that is used to label it. You may discover that a child who has recovered normal language may still not understand that clothes are found in a closet, even if they have had dozens of experiences that associate these things. It appears that if a child has not been explicitly told about a relationship, or had the pragmatics of something specifically stated to them, they may miss learning this information. This type of learning is from context, and is often called "incidental". Children with developmental delays do not demonstrate good incidental learning until they have been provided with a firm foundation of pragmatic knowledge through explicit instruction and descriptions.
One additional caution must be raised. Children with developmental delays also lack the pragmatic knowledge of what to do if they do not understand something. They must be specifically taught to say "what is this?", or "what does that mean?", when they do not understand something, or when something is new to them. If you do not do this, they usually remain silent and you will have great difficulty stumbling by accident upon all of their pragmatic gaps.
"Walk them through it": Responses to non-compliance and refusal
Children who are resistant to directions, or who are non-compliant, often simply do not understand the request or demand. Always begin with the assumption that non-compliance or refusal is a sign of impaired understanding. Ask the child to tell you what they are supposed to do. If they do not know, then "walk them through it" by taking them through each step as you explain the entire sequence of actions that you expect. If the child repeats what you asked them to do, then ask them how they will accomplish your request. At this point, you can tell if they are simply echoing your words or agreeing with you, or if they really understand what to do. If they still refuse, ask if they would like you to "walk through it" with them. If they still refuse, ask them to suggest something. This is the point at which you can discover if their refusal or lack of compliance is related to a lack of understanding on their part or on yours.
4. RITUALS, PERSEVERATION, AND SAMENESS
Helping a child understand the purpose of actions vs. the actions themselves
When a child insists on preserving sameness, doing something in the same way, or when a child becomes quite angry about a meaningless change in a routine, sequence, or spatial arrangement, it is a sign that they are not able to separate meaningful from irrelevant information. This is a pragmatic problem related to problems understanding which parts of an action are necessary for accomplishing the goal of the action, and which parts are irrelevant to that goal. Children with developmental delays have difficulty understanding their world, and usually do not understand what is happening in their world. This is a frightening and isolating situation, and so they need some parts of their world to be predictable. When they repeat a ritual or preserve sameness, they know what will happen next, and they find great comfort in this knowledge. The rest of the world may be incomprehensible, but rituals are familiar and safe. The less they understand about the world, the greater their dependence on familiar routines and rituals.
Adding novelty or small alterations to a ritual
It is possible to rather quickly help a child lose the need for ritual and familiarity. The method that I use to do this is to provide information and meaning to their world, and to add small amounts of variety to the familiar. As the ritual is changed in a very small way, I show the child that some (irrelevant) parts of a sequence, spatial pattern, or action can change without disrupting the ultimate goal. It is VERY important to only add a small bit of novelty each time to a ritual, or the child may not be able to understand or tolerate the changes. Although developmentally delayed children often become violent or uncontrollably upset when a small change is introduced into a familiar ritual or routine, show them that their ultimate goal is not destroyed. So, if they arrive at a location via a different route, they will still arrive. If a video program is begun at a different part of the tape, it does not change the familiarity.
As a child learns to tolerate very small changes, offer them something that is very similar (but still novel). You can change furniture placement in a room, greet them by entering through a different door, bring them to a place via a different route, offer a similar puzzle, tear different textured paper, give them a video with familiar characters but a different story, etc. Adapt the type and complexity of change that you offer to the child's ability to tolerate that change. As the child becomes more flexible, you can vary the type of change to a greater degree.
5. PLAY
Providing choices
One of the primary principles of microdevelopment is to provide the child with the ability to choose freely among a rich variety of possibilities. This is accomplished by placing within the child's easy reach a wide variety of toys and objects at all levels of play. In this way, the child can demonstrate by their exploration behavior the types of objects and actions that they understand or find meaningful.
Novelty is absolutely necessary in a learning situation. A child will not continue to explore or investigate objects that have become familiar or which offer no new information. It is important to offer a variety of toys and activities at higher and lower levels of play so that the child can continue to investigate favorite objects in more complex ways, and challenge themselves with novel objects. A child that becomes bored or that ceases to explore or investigate has ceased to learn.
Allowing the child to control the situation
Another major principle of microdevelopment is the autonomy of the learner. The child, not the adult, is given control of the learning situation. This means that the child determines what to do, how to do it, and for how long. The adult is there to offer choices, to show that something may be interesting (but NOT to insist that the child find it interesting). They are there to offer information about what the child explores, to act as a play partner and peer friend, to help suggest new ways to explore, and to offer appropriate activities and objects to explore. They are NOT there to teach, to direct, or to demand compliance. The journey of learning and development belongs to the child's alone, and the adult can only facilitate (or impede) their progress. If a child shows no interest in an activity or toy modeled by the adult, then the adult must accept this and try something else or wait until the child approaches something. If the adult is demonstrating something or explaining something and the child walks or turns away, then the adult must accept this and also move on. The adult is to NEVER force the child or insist that they participate, comply, watch, or do something.
Giving information to children in the sensory stage
Parents often find sensory exploration in an older child very disconcerting, and they often actively discourage it. However, it is very important that a child be allowed to explore and learn about their world at their own stage of development. If a child still has gaps in their understanding of the sensory nature of the world, they will be impaired in their ability to understand how those sensory properties change through action, temperature, or movement. Offer the child not only categories of objects that have some sensory property in common, but also give them information (and words to describe) the sensory experiences.
Giving information at the simple cause and effect stage
When children have sufficient knowledge about the sensory properties of objects, they will be able to understand how these sensory properties change through some type of manipulation (heating, freezing, painting, moving, etc.). This stage is particularly marked by a fascination with moving objects and simple musical instruments.
Giving information at the complex cause and effect stage
When children have a foundation of knowledge about how sensory properties (shape, color, texture, sound, etc.) change through simple manipulation, they are ready to explore more complex changes or sequences of change. This is the stage to provide simple crafts, sequences, something that changes in different or unexpected ways, and things that create different changes in different objects.
Using dolls, animals, or puppets to demonstrate representation
When a child has sufficient information about how objects change in predictable ways through various types of manipulation, they become very interested in more complex and unpredictable changes of the sort offered by people. This is the stage to provide puppets, dolls, figurines, doll houses, and action figures in which simple and familiar activities and words can be represented. This is also the stage in which books and pictures become a meaningful tool for information. Begin with realistic pictures or photographs of people in