During the first half of the twentieth century, childhood psychosis and childhood schizophrenia were common labels for children with dramatic behavioral disorders. They were used more or less interchangeably for decades whenever children’s behavior resembled symptoms observed in psychotic and schizophrenic adults. Research on these conditions in childhood was sporadic, but cases studies did appear in the medical literature. One 1940 summary concluded that schizophrenia before puberty was rare, more common in boys than girls, and had such a poor prognosis that recovery was the exception to the rule. Before autism existed as a clinical syndrome, children with autism were called schizophrenic or psychotic.
Complicating the fact that child psychosis or schizophrenia lacked clear diagnostic criteria was the reality that making diagnoses early in life was notoriously difficult. Before they developed language, children could obviously not respond to questions. Even after they could speak, their answers might be unreliable. That children were in the midst of rapid developmental change could inhibit adult symptoms from appearing clearly. No matter how confusing children’s behavior, childhood psychosis and schizophrenia were routine classifications. They were applied to individuals who would be placed on the autism spectrum today along with many who would not.
Little empirical or statistical data existed a century ago either to confirm or refute assertions about dramatic developmental disorders in childhood. Observers of childhood psychosis or schizophrenia were well aware of how little they knew, but they stubbornly pursued a few key questions. The first was about the perplexing relationship between mental retardation and mental illness. This reflected long-established concern about the heritability of feeblemindedness and eugenicists’ worries that feebleminded children would grow up to become mentally diseased and menacing adults. The second was about whether symptoms observed in children were signs of purely pediatric conditions or early warnings that full-blown adult psychiatric disorders lay ahead.
In the case studies and research that appeared before 1943, these two questions surfaced repeatedly. Evidence that some psychotic or schizophrenic children were mentally normal—with average IQs, for instance—while others were not prompted much debate. Could the pediatric onset of psychosis or schizophrenia coexist with normal cognition in some children and cause deterioration in others whose intellectual development had previously been normal? If so, should feebleminded and institutionalized children remain the focus of treatment for psychosis or schizophrenia?
Best practices for increasing diagnostic precision were another subject of much discussion. Should children be diagnosed only when they displayed behaviors that characterized adult mental illness, or did the fact of children’s immaturity call for a distinctive symptom profile? Should there be different profiles at different ages? It seemed plausible that magical thinking, belief in fantasies, even hallucinations might be perfectly normal for young children and still be reliable indicators of mental illness in adults. At what point during childhood should such signs start pointing toward pathology?
One core characteristic of psychosis and schizophrenia in childhood—inability to relate socially and emotionally to others—eventually became autism’s defining feature. At Bellevue Hospital in New York, Harry Bakwin observed bizarre behaviors in young children, compared them to research on cats and dogs, and concluded that the relational dimension of human development was by far the most essential. “Socialization is a basic mammalian pattern of behavior, and only psychotic persons are isolationists.”
Bakwin’s use of the term psychotic was no coincidence. Nor was it coincidental that virtually all descriptions of childhood psychosis and schizophrenia emphasized the radical separateness of these children. Observers looking for signs of childhood schizophrenia relied on noticing “the early withdrawal, autistic thinking, self-preoccupation, and socializing difficulties which stamped the child as different…” Psychotic and schizophrenic children inhabited inaccessible worlds.
Another Bellevue physician, neurologist and child psychiatrist Lauretta Bender, saw hundreds of cases of dramatic behavioral disturbance on the hospital’s Children’s Service during the 1930s and 1940s. After witnessing abrupt developmental regression and behaviors like whirling, head-banging, mutism, echolalia, and fixations on inanimate objects, she diagnosed almost all of them with childhood schizophrenia. Some of these children would be placed on the autism spectrum today, but in the 1930s and 1940s, that diagnosis did not exist.
In 1942, Bender described the condition of childhood schizophrenia as a “definite syndrome,” a “pathology at every level and in every field of integration within the functioning of the central nervous system” that undermined children’s growth and inhibited the development of precisely those features considered most human. In the mid-1950s, Bender estimated that 47% of all the children seen at Bellevue had been diagnosed with schizophrenia, including five times more boys than girls. Standard treatments included electric shock, anti-convulsant medications, major tranquilizers, even experimental courses of LSD.
From the 1930s through the 1950s, childhood schizophrenia remained an ill-defined but expansive category. In 1959, psychiatrist Hilde Mosse of the Lafargue Clinic in Harlem called it “a fashionable and much abused diagnosis.” Using it was scientifically and socially wrong most of the time, she insisted. Children in trouble for other reasons and children who lived in “a socially difficult milieu,” by which Mosse meant impoverished families and communities, were disproportionately diagnosed and then subjected to interventions that caused permanent harm.
During the late 1950s and 1960s, autism attracted increasing attention as a clinical syndrome in its own right and became a more familiar term. Consequently, childhood psychosis and schizophrenia disappeared by the 1970s, at least as viable diagnoses. With epidemiological research showing that the mental illness we still call schizophrenia was different than autism, these terms were discarded. At the time, they were embarrassing reminders of how little had been known in the past about child psychiatry and pediatric neurology. They remain important touchstones in the still uncertain story and history of autism.