Harry Harlow’s experiments with rhesus monkeys, conducted primarily at the University of Wisconsin from the late 1950s through the 1970s, demonstrated that infant primates form attachments based on comfort and physical contact rather than food. The work upended the prevailing belief that babies bond with their mothers simply because mothers feed them. It also ventured into darker territory, with isolation studies that revealed how devastating the absence of social contact can be during early development, and sparked ethical debates that helped reshape how scientists treat animal subjects.
The Wire Mother and the Cloth Mother
The most famous of Harlow’s experiments involved two artificial surrogate mothers placed in the cages of infant rhesus monkeys. One surrogate was made of bare wire mesh and held a bottle of milk. The other was wrapped in soft terrycloth but offered no food. If the dominant theory of the time was correct, that infants attach to whoever feeds them, the baby monkeys should have spent most of their time clinging to the wire surrogate with the bottle. They did not. The infants overwhelmingly preferred the cloth mother, spending the vast majority of their time pressed against the soft surface and visiting the wire surrogate only briefly to feed.
Harlow reported this finding in a landmark 1958 address to the American Psychological Association, titled “The Nature of Love,” and published it the following year. The conclusion was striking in its simplicity: contact comfort, the feeling of something soft and warm to hold onto, mattered more to infant monkeys than nourishment when it came to forming an attachment bond.
The preference was not just a passive habit. When Harlow introduced a frightening stimulus into the cage, such as a loud mechanical toy, the infants ran to the cloth mother, clutched her, and gradually calmed down. Monkeys raised with only a wire surrogate showed far more distress and did not use it as a refuge in the same way. The cloth-mother infants also used their surrogate as a kind of home base: they would cling to her, then venture out to explore a new object or room, returning when they felt uncertain. This “secure base” behavior looked remarkably like what researchers were beginning to observe in human children with their own mothers.
What Isolation Did to the Monkeys
Harlow and his colleagues went further than surrogate studies. In a series of experiments that grew progressively more severe, they raised infant monkeys in total social isolation, sometimes for as long as twelve months. These monkeys could not see, touch, or interact with any other living being. The results were devastating. Total social isolation for at least the first six months of life consistently produced severe deficits in virtually every aspect of social behavior.1PubMed Central. Social recovery by isolation-reared monkeys The isolated monkeys emerged fearful, socially incompetent, and prone to self-harming behaviors like biting their own limbs. They could not play normally, could not engage in social grooming, and showed little interest in other monkeys. When isolated females later became mothers themselves, many were neglectful or abusive toward their own infants.
The duration of isolation mattered enormously. Monkeys isolated for only the first three months showed some ability to recover socially after being returned to group living. Those isolated for six months or longer were far more impaired, and twelve months of isolation produced damage that appeared largely permanent using the standard rehabilitation methods available at the time.
The Pit of Despair
In the late 1960s and early 1970s, Harlow turned his attention to modeling depression. He and Stephen Suomi designed a device they called the “vertical chamber apparatus,” though it became widely known by the grimmer name Harlow himself gave it: the pit of despair. It was an inverted pyramid-shaped chamber, enclosed at the top, that prevented a monkey from climbing out. Monkeys placed in the chamber for weeks at a time became profoundly withdrawn. They stopped moving, huddled at the bottom, and showed behavioral patterns that Harlow compared to clinical depression in humans.
The work was controversial from the start, and not just on ethical grounds. Critics questioned whether the extreme conditions of the chamber told researchers anything meaningful about how depression develops naturally in humans or other primates. The behaviors produced looked more like learned helplessness under inescapable confinement than the kind of depression that arises from social loss or chronic stress. Still, the experiments contributed to an emerging scientific conversation about whether animal models could illuminate human mood disorders, and they remain some of the most frequently cited examples in discussions about the limits of animal research.
Recovery Through “Therapist” Monkeys
One of the more hopeful findings from Harlow’s lab came from rehabilitation attempts. Harlow and Suomi, later joined by Melinda Novak, discovered that socially devastated monkeys could be partly rehabilitated by pairing them with younger, socially healthy “therapist” monkeys. The approach was deliberate: younger monkeys were chosen because they were not yet old enough to engage in rough play or aggressive social behavior that might overwhelm a fragile, isolated monkey. Instead, the young therapists would approach the withdrawn monkey and simply cling to it, occasionally attempting gentle, elementary play.2PubMed Central. Harry Harlow’s pit of despair: Depression in monkeys and men
The depressed monkeys initially just huddled in a corner. But over time they began to reciprocate the clinging, then joined in play, and eventually started initiating play on their own. The formerly depressed monkeys appeared to behave relatively normally after extended contact with the therapists.3PubMed Central. Harry Harlow’s pit of despair: Depression in monkeys and men The recovery was not always complete, and it worked better for monkeys with shorter isolation histories. But the finding suggested that social damage, even extreme social damage, was not necessarily a one-way street. The right kind of gentle, persistent social contact could undo at least some of the harm.
How the Experiments Changed Thinking About Human Attachment
Harlow’s work arrived at a pivotal moment in developmental psychology. The British psychiatrist John Bowlby was already building a theory that children need a continuous, warm relationship with a caregiver for healthy psychological development, but his ideas were based mostly on clinical observations of orphaned and institutionalized children. Harlow’s controlled experiments with monkeys gave Bowlby something he could not get from human research: experimental evidence under conditions that would be unethical to impose on children. Bowlby profited highly from Harlow’s experimental work with rhesus monkeys in constructing his theory of the bond between children and their caregivers.4PubMed. “When strangers meet”: John Bowlby and Harry Harlow on attachment behavior
The influence flowed in both directions. René Spitz, a psychoanalyst who had documented the devastating effects of institutional care on human infants, what he called “hospitalism,” had inspired Harlow’s surrogate work in the first place. And Harlow’s monkey findings, in turn, gave new impetus to Spitz’s own research on the emotional development of deprived infants.5PubMed Central. Of monkeys and men: Spitz and Harlow on the consequences of maternal deprivation Bowlby used Harlow’s experimental findings to further develop his ideas on attachment, separation, and loss, which became the foundation of modern attachment theory.6PubMed Central. The nature of love: Harlow, Bowlby and Bettelheim on affectionless mothers
Before Harlow, the prevailing view in both psychoanalysis and behaviorism was that an infant’s bond with its mother was essentially a learned response to being fed. Freud’s followers framed it in terms of oral gratification. Behaviorists framed it in terms of conditioned reinforcement. Both assumed food was the underlying currency. The cloth-versus-wire experiment made that position difficult to maintain. Contact comfort was not a secondary bonus; it was the primary thing infant monkeys sought. The practical implication for human childcare was significant: children in hospitals and institutions needed more than adequate nutrition and hygiene. They needed to be held.
What Happens in the Body After Maternal Separation
Later researchers built on Harlow’s behavioral observations by measuring what happens physiologically when infant monkeys are separated from their mothers. One study examined rhesus monkeys that had been separated from their mothers early in life and raised instead with age-matched peers. Even after a year and a half of normal social living, these peer-reared monkeys still showed measurably different stress responses compared to monkeys raised by their mothers. Their baseline hair cortisol levels were significantly lower, and their cortisol response to acute stressors was substantially delayed.7PubMed Central. Maternal separation produces lasting changes in cortisol and behavior in rhesus monkeys
This is a telling finding because it suggests the effects of early maternal separation are not just behavioral quirks that fade with time. The stress-response system itself gets wired differently. A blunted or delayed cortisol response may look like calmness on the surface, but it actually reflects a dysregulated system: the body’s alarm does not fire at the right time or with the right intensity. In human children raised in institutional care, researchers have found strikingly similar patterns of altered stress physiology, which lends weight to the idea that Harlow’s monkey findings translate, at least in broad strokes, to humans.
Not All Macaque Species React the Same Way
One important nuance that often gets lost in popular retellings of Harlow’s work is that the severity of the response to maternal separation is not universal across primates, or even across closely related macaque species. Harlow worked primarily with rhesus macaques, which have a relatively rigid and hierarchical social structure. When researchers tested whether maternal separation would produce the same depressive behavior in Tonkean macaques, a species with more open and tolerant social relationships, they predicted the reaction would be less severe. The study confirmed that the infant’s reaction to mother loss is related to the social environment, not just the loss itself.8PubMed. Effects of six-day maternal separation on tonkean macaque infants
The implication is worth sitting with. Harlow’s experiments demonstrated something real and important about what happens when primates are deprived of social contact. But the specific intensity of the damage may reflect not just the deprivation itself but the kind of social world the infant was built for. A species whose social structure includes many potential caregivers, and whose members are more tolerant of unfamiliar individuals, produces infants that are more resilient to the loss of one specific caregiver. Rhesus macaques, with their tighter and more competitive social hierarchies, may represent a more vulnerable end of the spectrum.
The Ethics Debate That Followed
Harlow’s experiments, especially the isolation chambers, became flashpoints in the emerging animal rights movement. The Laboratory Animal Welfare Act had passed in the United States in 1966, but enforcement was limited and the broader cultural conversation about animal welfare in research was still in its early stages. As Harlow continued and escalated his deprivation experiments through the 1970s, criticism grew.9PubMed Central. Harry Harlow’s pit of despair: Depression in monkeys and men Publications like Peter Singer’s Animal Liberation in 1975 brought the argument into mainstream consciousness, and Harlow’s pit of despair became one of the most cited examples of what critics considered needlessly cruel animal experimentation.
The ethical criticism was not just about the suffering itself. Skeptics argued that the later isolation and depression experiments had diminishing scientific returns: the core insight about the importance of contact comfort and social bonds had already been established by the surrogate studies, and the increasingly extreme deprivation protocols were producing results that were predictable, if not obvious. Defenders countered that the severity data mattered for understanding the dose-response relationship between deprivation and damage, and that the therapist-monkey recovery work offered practical insights into rehabilitation. These arguments have never fully resolved, and Harlow’s legacy remains split: widely credited for transforming developmental psychology, and widely condemned for the methods he used to do it.
The Neuroscience of Why Touch Matters
Modern research has begun to fill in the biological mechanisms behind what Harlow demonstrated behaviorally. The opioid system and oxytocin are now known to be involved in social bonding and affect the timing and frequency of social touch in primates and other social mammals.10Neuroscience. The evolution of mytouch: Touch and sociality in nonhuman and human primates When a primate engages in grooming, nursing, or simply pressing against another warm body, these neurochemical systems activate and reinforce the behavior. The relationship between oxytocin release and social interactions is not perfectly predictable, but the general pattern is clear: affiliative physical contact triggers reward pathways that make the animal seek more of it.
This helps explain why Harlow’s cloth mother worked as well as it did, and also why it had its limits. The terrycloth surrogate provided enough tactile stimulation to engage some of the brain’s comfort circuitry, which is why the infants preferred it so strongly over wire. But a passive cloth form could not groom back, respond to distress signals, or initiate play. Monkeys raised exclusively with surrogates still developed social deficits, though far milder ones than monkeys raised in total isolation. The cloth mother was better than nothing, but it could not replace the dynamic, responsive contact that a living caregiver provides. That distinction between passive comfort and interactive social engagement turns out to be one of the most important takeaways from the entire body of Harlow’s research.
What the Experiments Mean for Human Childcare Policy
Harlow’s findings, combined with Bowlby’s theoretical framework and decades of subsequent human research, had direct consequences for how institutions treated young children. Before this body of work gained traction, orphanages and hospitals routinely minimized physical contact with infants, partly out of hygiene concerns and partly because the prevailing behaviorist philosophy warned against “spoiling” children with too much holding. Harlow’s monkeys helped shift that thinking. If a baby monkey, given a choice between food and comfort, chose comfort overwhelmingly, it became much harder to argue that human infants did not need the same.
Over the following decades, hospital policies changed to allow parents to stay with their children during admissions. Orphanage practices shifted toward smaller group settings and more caregiver contact, though the pace of reform varied dramatically by country. The movement toward foster care over institutional care for young children drew explicitly on attachment research that traces a direct line back to Harlow’s lab. None of this is to say Harlow single-handedly changed childcare policy. But his experiments gave the attachment movement something it desperately needed: hard evidence that was difficult to argue away, even for skeptics who dismissed clinical observations as anecdotal.
Surrogate Variations and What They Revealed
The wire-versus-cloth experiment is the version most people know, but Harlow’s lab ran many variations on the surrogate theme. Researchers experimented with surrogates that rocked, surrogates heated to body temperature, and surrogates that could be made to deliver blasts of air or retract spikes to “reject” the clinging infant. The heated surrogates were somewhat preferred over room-temperature ones, and rocking surrogates were preferred over stationary ones, but these features mattered less than the basic softness of the covering. Texture dominated temperature and motion as the key variable driving attachment behavior.
The “rejecting mother” surrogates were perhaps the most unsettling variation. When a surrogate was rigged to blast the infant with compressed air or push it away with a spring-loaded frame, the infant monkey did not abandon the surrogate. Instead, it clung harder. Harlow found this parallel to reports of abused human children, who often remain strongly attached to the parent who mistreats them. The finding underscored that attachment is not a rational cost-benefit calculation. It is a deep biological drive that persists even when the object of attachment is a source of pain. For researchers studying why abused children resist being separated from abusive parents, this observation from Harlow’s lab became a touchstone.