My Age of Anxiety: Fear, Hope, Dread, and the Search for Peace of Mind

by Scott Stossel

  Securely attached children were able to create what Bowlby called “an internal working model” of their mothers’ love that they could carry out into the world with them throughout their lives—an internalized feeling of psychological security, a sense of being loved and of being safe in the world. But when the mothers had “insecure” or “ambivalent” attachment relationships with their toddlers—if the mothers were anxious and overprotective or emotionally cold and withdrawn—the children were more anxious and less adventurous; they would cling to their mothers and become very agitated in response to any separation.

  Over the next four decades, Bowlby and his colleagues would develop a typology of attachment styles. Secure attachment in childhood predicted low anxiety levels and a healthy degree of intimacy in adult relationships. Ambivalent attachment—which described those children who clung most anxiously, who displayed high levels of physiological arousal in novel situations, and who were much more concerned with monitoring their mothers’ whereabouts than in exploring the world—predicted high levels of anxiety in adulthood.e Avoidant attachment in a child—which described those kids who tended to withdraw from their mothers after separations—predicted a dislike of intimacy in adulthood.f

  The person most responsible for helping Bowlby develop this taxonomy of attachment styles was the psychologist Mary Ainsworth. In 1929, Ainsworth was a freshman at the University of Toronto plagued by feelings of inadequacy. That year she took a course in abnormal psychology from William Blatz, a psychologist whose security theory held that a young child’s sense of well-being derives from proximity to his parents and that the child’s ability to grow and develop depends on the constancy of his parents’ availability. Drawn to Blatz by her own abiding sense of insecurity, Ainsworth went on to do graduate work in psychology, eventually becoming, in 1939, a lecturer in the University of Toronto’s psychology department. But when her husband decided to attend graduate school in England, she had to find work in London. A friend directed her to an advertisement in The Times, placed by a psychoanalyst seeking research help on a project about the developmental effects of early childhood separation from the mother. Keen to understand her relationship with her own mother, who had been self-absorbed and distant, Ainsworth applied. John Bowlby hired her—and with that began the central partnership in the development of attachment theory.

  Ainsworth made two signature contributions to the field. The first was in the mid-1950s, when she accompanied her husband to Kampala, Uganda. In Kampala, Ainsworth identified twenty-eight unweaned babies from local villages and began observing them in their homes, studying attachment behavior in a natural environment. She kept meticulous records, tracking breast-feeding, toilet training, bathing, thumb sucking, sleeping arrangements, expressions of anger and anxiety, and displays of happiness and sadness, and she watched how the mothers interacted with the children. It was the most extensive naturalistic observation of this sort yet conducted.

  When Ainsworth first arrived in Uganda, she agreed with both the Freudians and the behaviorists that the emotional attachment babies invested in their mothers was a secondary association with feeding: mothers provided breast milk, which provided comfort, so babies came to associate that feeling of comfort with the mother; there was nothing inherent in the maternal relationship itself, distinct from the provision of food, that was psychologically significant. But as Ainsworth totted up her meticulous observations, she changed her mind. The Freudians and behaviorists were wrong, she concluded, and Bowlby was right. When the babies began to crawl on their own and to explore the world around them, they would repeatedly return to their mothers—either physically or by exchanging a reassuring glance and a smile—and appeared always to remain conscious of exactly where their mothers were. Describing what she observed when the babies first began to crawl, Ainsworth wrote that the mothers seemed to provide the “secure base” from which these excursions can be made without anxiety. The secure base would go on to become a crucial element of Bowlby’s attachment theory.

  Ainsworth noticed that whereas some babies clung fiercely to their mothers at almost all times and cried inconsolably when separated from them, others seemed indifferent, tolerating separations without evident distress. Did this mean that the uninterested babies loved their mothers less than the clinging babies and were somehow less attached to them? Or, as Ainsworth came to suspect, did this mean that the clinging babies were in fact the less securely attached ones?

  Ainsworth ultimately deemed seven of the twenty-eight Ugandan babies “insecurely attached.” She studied them carefully. What made them so anxious and clingy? For the most part, these insecure babies seemed to receive the same quantity of maternal care as the other ones; the insecure babies had not suffered inordinate or traumatic separations that would explain their anxiety. But as she looked more closely, Ainsworth began to notice things about the mothers of these insecure babies: some of them were “highly anxious,” distracted by their own preoccupations—often they had been deserted by their husbands or had disordered family lives. Still, she wasn’t able to point conclusively to specific maternal behaviors that generated separation anxiety or insecure attachment.

  In 1956, Ainsworth moved to the United States and began teaching at Johns Hopkins. Determined to find out whether attachment behaviors were culturally universal, she decided to contrive an experiment that could test this.

  Thus was born what Ainsworth called the strange situation experiment, which has been a staple of child development research ever since. The procedure was simple. A mother and child would be placed in an unfamiliar setting—a room with a lot of toys in it—and the baby would be free to explore. Then, with the mother still present, a stranger would enter the room. How would the baby react? Then the mother would exit the room, leaving the baby with the stranger. How would the baby react to that? Then the mother would come back. How would the baby respond to the reunion? This could be repeated without the stranger—the mother would leave the child alone in the room and then return a while later. All of this would be observed by researchers sitting behind a two-way mirror. Over the ensuing decades, thousands of repetitions of this experiment produced mountains of data.

  The experiments yielded some interesting insights. In the first phase of the experiment, the babies would explore the room and look at the toys while checking in frequently with the mothers—suggesting that babies’ psychological need to operate from a “secure base” is indeed universal across cultures. But babies varied a lot in how distressed they would become when separated from their mothers: about half of them cried after their mothers left the room, and some babies became severely distressed and had a hard time recovering. When their mothers returned, the distressed babies would both cling to and hit them, displaying both anger and anxiety. Ainsworth labeled these insecure babies “ambivalent” in their attachment. Even more fascinating to Ainsworth than the ambivalent babies were those she would come to label “avoidant” in their attachment style: these babies seemed completely indifferent to their mothers’ departures and rarely got perturbed. Superficially, they seemed quite healthy and well adjusted. But Ainsworth would come to believe—and a lot of research would eventually be produced to support the idea—that the independence and equanimity these avoidant babies displayed were in fact the product of a defense mechanism, an emotional numbing designed to cope with maternal rejection.

  As Ainsworth collected her data, the most telling fact to emerge was the powerful correlation between a mother’s parenting style and her child’s general level of anxiety. Mothers of the children identified by researchers as securely attached were quicker to respond to their children’s signals of distress, tended to hold and caress their children for longer, and derived more apparent pleasure from doing so than did the mothers of the ambivalent and avoidant children. (The mothers of securely attached children didn’t necessarily interact more with them, but they interacted better, being more affectionate and responsive.) The mothers of the avoidant children display
ed the most rejecting behavior; the mothers of the ambivalent children displayed the most anxiety and also by far the most unpredictability in their responses to their children—sometimes they were loving, sometimes rejecting, sometimes distracted. Ainsworth would later write that the predictability of maternal response helped dictate a child’s confidence and self-esteem later in life; those mothers who predictably responded to distress signals quickly and warmly had calmer, happier babies who became more confident, independent children.

  Over the next few decades, the connection between attachment style and psychological health was repeatedly confirmed by a host of different measures.g A series of influential longitudinal studies begun by researchers at the University of Minnesota in the 1970s have found that securely attached children are happier, more enthusiastic, and more persistent and focused when working on experimental tasks than anxiously attached children are and that they have better impulse control. On almost every test the researchers devised, the securely attached children did better than the ambivalently attached ones: they had higher self-esteem, stronger “ego resiliency,” and less anxiety and were more independent; they were even better liked by their teachers. They also displayed greater empathy for others—probably because the insecurely attached children were too self-preoccupied to be much attuned to anyone else. The securely attached children just seemed to enjoy life more: none of the ambivalently attached children smiled, laughed, or expressed delight at the same level as the securely attached children. Many of the ambivalently attached children tended to fall apart when subjected to even minor stress.

  These effects persisted for years, even decades. Teenagers who had been securely attached as toddlers had an easy time making friends—but those who had been ambivalently attached were overwhelmed by the anxiety of navigating social groups and often ended up friendless and alienated. Studies found that adults with mothers who had ambivalent attachment styles tended to procrastinate more, to have more difficulty concentrating, to be more easily distracted by concerns about their interpersonal relations, and—perhaps as a result of all this—to have lower average incomes than those with mothers who had either secure or avoidant attachment styles. Many studies from the last thirty years have suggested that insecure attachment as a baby and young child is highly predictive of emotional difficulty as an adult. A two-year-old girl with an ambivalent attachment to her mother is on average much more likely to become an adult whose romantic relations are plagued by jealousy, doubt, and anxiety; she will always be seeking—likely without success—the secure, stable relationship that she did not have with her mother. The daughter of an anxious and clingy mother will herself likely grow up to be an anxious and clingy mother.

  A mother who, due to adverse experiences during childhood, grows up to be anxiously attached is prone to seek care from her own child and thereby lead the child to become anxious, guilty, and perhaps phobic.

  —JOHN BOWLBY, A Secure Base (1988)

  During the postwar decades, neurochemical research would demonstrate that when an infant or an adult is stressed, a cascading series of chemical reactions in the brain produces anxiety and emotional distress; returning to a secure base (the mother or a spouse) releases endogenous opiates that make the individual relax and feel safe. Why should this be so?

  Back in the 1930s, John Bowlby, already absorbed in his studies of the mother-child bond, discovered the work of the early ethologists. Ethology, the scientific study of animal behavior, suggested that many of the attachment behaviors Bowlby had been observing in humans were universal to all mammals, and it supplied an evolutionary explanation for these behaviors.

  The evolutionarily adaptive benefit of early attachment behaviors is not hard to figure: holding offspring near helps a mother to keep them safe until they are fit to fend for themselves. Thus it was possible, Bowlby realized, to explain separation anxiety almost purely in terms of natural selection: there’s an adaptive value to psychological mechanisms that encourage mothers and children of whatever species to stick close to one another by producing distress when they are separated; those children most predisposed to cling to their mothers in times of distress may gain a Darwinian advantage over their peers.

  In yanking the sources of anxiety out of the realm of fantasy and into the world of ethology, Bowlby alienated his psychoanalytic colleagues.h When Bowlby first presented his emerging research findings in the early 1950s, he was attacked from two sides—by the psychoanalysts and the behaviorists. For the behaviorists, the mother-child bond had no inherent importance; its relevance to separation anxiety derived from the “secondary gains”—the provision of food, the soothing presence of the breast—that the child came to associate with his mother’s presence. For the behaviorists, an attachment wouldn’t even exist distinct from the specific needs—mainly for food—that the mother met. Bowlby disagreed. Attachment behaviors—and separation anxiety—were biologically hardwired into animals, including humans, independent of the association between food and mothers. In defense of this argument, Bowlby cited Konrad Lorenz’s influential 1935 paper “The Companion in the Bird’s World,” in which Lorenz had revealed that goslings could become attached to geese, and sometimes even to objects, that did not feed them.i

  Freudians argued that Bowlby’s reliance on animal models of behavior gave short shrift to the intrapsychic processes—such as the battle between the id and superego—that set the human mind apart from other animals’. Once, after Bowlby presented an early paper on separation anxiety to the British Psychoanalytic Society, the organization dedicated multiple subsequent sessions to the presentations of all the critics who wanted to “bash” him. There were calls to “excommunicate” him for his apostasy.

  As psychoanalytic criticism of Bowlby swelled, he received a bracing injection of support from the world of animal research when, in 1958, Harry Harlow, the president of the American Psychological Association and a psychologist at the University of Wisconsin, published an article in the American Psychologist called “The Nature of Love.” In it, Harlow described the series of experiments that are now a fixture of every introductory psychology course.

  The experiments came about by happenstance. Many of the rhesus monkeys in Harlow’s lab had been contracting fatal diseases, so he took sixty infant monkeys from their mothers a few hours after birth to be raised in a germ-free environment. It worked: the separated monkeys stayed free of disease, and their physical development seemed normal, even though they remained apart from their mothers. But Harlow observed some strange things about their behavior. For one, they clung desperately to the cloth diapers used to line the cage floors. Those monkeys who were placed in mesh cages without diapers seemed to struggle physically to survive; they did better if given a mesh cone covered in terry cloth.

  This gave Harlow an idea for how he might test a hypothesis that he, like Bowlby, had always found suspect: the notion, advanced by both the psychoanalysts and the behaviorists, that a baby becomes attached to his mother only because she feeds him. Even granting that the mother’s association with food may provide a “secondary-reinforcing agent” (in behaviorist terminology), Harlow didn’t think that those early feedings were enough to account for the maternal bond—the love and affection—that persisted for decades afterward. Could his separated rhesus monkeys, Harlow wondered, be used to research the origins of a child’s love of his mother? He decided to try.

  He separated eight rhesus babies from their mothers and placed each in its own cage, along with two contraptions he called surrogate mothers. One of the two mothers in each cage was made of wire mesh; the other was made of wood but was covered with terry cloth. In four of the cages, a rubber nipple offering milk was affixed to the mesh surrogate; in the other four, the nipple was affixed to the cloth-covered one. If the behaviorist supposition was correct and attachment was merely a by-product of association with feeding, then the infants should always have been drawn to the surrogate possessing the nipple.

  That’s not what happened
. Instead, all eight monkeys bonded with the cloth mother—and spent sixteen to eighteen hours a day clinging to it—even when the wire mother provided the feeding nipple. This was a devastating blow to the behaviorist theory of separation anxiety. If the monkeys were more likely to bond with a soft and cuddly object that didn’t feed them than with a wire one that did, hunger relief could not be the operative association in the bonding process, as the behaviorists had assumed.j

  By coincidence, Bowlby attended the American Psychological Association meeting in Monterey, California, where Harlow first presented his “Nature of Love” paper. Bowlby immediately recognized the relevance of Harlow’s work to his own, and the two men made common cause. In the ensuing years, other studies would replicate Harlow’s initial findings. For Bowlby, this was vindication, armor against the assaults of the Freudians and the behaviorists. “Thereafter,” Bowlby would later write, “nothing more was heard of the inherent implausibility of our hypotheses; and criticism became more constructive.”

  The Harlow study would prove even more relevant to Bowlby’s ideas about attachment relationships than either man knew at the time. In later years, the monkeys from Harlow’s initial study suffered lasting effects from the separation experiment. As intensely as the babies had seemed to bond with the inanimate cloth surrogates, this was clearly no replacement for a real mother-child relationship: for the remainder of their lives, these monkeys had trouble relating to their peers and exhibited abnormal social and sexual behavior. They were abusive, even murderous, parents. When presented with novelty or stress, they became much more anxious, inhibited, and agitated—which is exactly what Bowlby had observed in his studies of humans who had endured separations or difficult relations with their mothers. All of this was haunting confirmation of the long-term effects of early experiences with separation and attachment.k